Wins Produced - Wages of Wins (Dan Rosenbaum, 2006)

Home for all your discussion of basketball statistical analysis.
Post Reply
Crow
Posts: 6146
Joined: Thu Apr 14, 2011 11:10 pm

Wins Produced - Wages of Wins (Dan Rosenbaum, 2006)

Post by Crow » Fri Apr 15, 2011 1:00 am

recovered page 1 of 19

Author Message
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina

PostPosted: Thu Jul 27, 2006 5:32 pm Post subject: Wins Produced - Wages of Wins (Berri, Schmidt, and Brook) Reply with quote
Dave Berri, Martin Schmidt, and Stacey Brook have created a measure called Wins Produced.

http://www.wagesofwins.com/

This measure is a variant of the many linear weights measures out there, but instead of using only theory to come up with the weights, they use both theory and regression techniques.

[This is clearly a working post, as I am making some changes as I reread parts of the book.]

(1) Their first step is to regress team wins onto offensive and defensive efficiency (or pace-adjusted points scored and points given up). What this does, in essence, is give them their weights for points scored, which is 0.033.

(2) Possessions = FGA + 0.44*FTA - OREB + TO = DREB + FGM + 0.44*FTM + TO. So from this and the fact that over their time period teams averaged 1.02 points per possession, they get their multiplier for possessions given up and acquired, which is 1.02 * 0.033 = 0.034. So rebounds, turnovers, and true shot attempts (FGA + 0.44*FTA) get weights of 0.034 or -0.034. Steals also get a multiplier of 0.034 because they are the opposite of a turnover.

(3) Next, they need to estimate the value of assists, blocks, and personal fouls. Let's start with blocks. They estimate that each block reduces opponents' two-point field goals made by 0.65 and since they have already estimated the value of a made two-point field goal (two points minus a field goal attempt) at 0.033, then blocks are worth 0.65*0.033 = 0.021.

(4) Next, they calculate, on average, how many opponents' free throws made are generated by a personal foul. This results in them weighting each personal foul by -0.018.

(5) Now onto assists. They admit that they almost wrote the book leaving assists out altogether. And one reason why they might have considered this is that accounting for assists will only hurt them when go back and see how well their measure predicts wins (they make an adjustment to fix this). But they found that when using wins produced from a previous season to predict wins, assists were a helpful predictor for what they could not explain. Thus, after some work they ended up with a weight of 0.022 for assists.

(6) Then they adjust their Wins Produced measure so that, on average, it is equal for every position. There is no basis for this adjustment in the numbers; they justify it with an argument that a team could not play all centers. (True, but it could be that centers should be paid more.) This adjustment likely will make it harder for them to predict team wins, so that is not a justification for this adjustment.

This adjustment is a big deal in their ratings, and given how critical they are in the book of the understanding basketball people have of the game, it is surprising to see them so comfortable relying on the same kind thinking to justify a pretty big adjustment. And I say this even though I agree with the need for this adjustment.

(7) Lastly, they adjust using team defensive measures. If we think back in terms of the original regression equation regressing wins onto points scored and points given up (per possession), when they aggregate up their Wins Produced by team, they already have accounted for points scored and (most of) possessions. But for points given up, they have only accounted for free throws made (through personal fouls) and part of two-point field goals made (through blocks). So if they account for the rest of two-point field goals made and for three-pointers made, they will have accounted for points given up. They make another adjustment that, in essence, does that. (It also cleans up possessions by accounting for team rebounds and opponents' non-steal turnovers.)

They say in the book that this adjustment barely affects the relative ratings of the players; the correlation between the ratings with and without the adjustment is 0.99.

But my guess is that it is huge in terms of helping them predict wins later on. My guess is that instead of predicting 95% of team wins like they do with this adjustment, they predict less than 70% without it. So that begs the question. If this adjustment barely budges their Wins Produced measure for individual players but is so critical in explaining team wins, it raises a big red flag as to the validity of using the prediction of team wins as a barometer of their methodology.

(8) So at the end of the day when they add up their Wins Produced team by team, what they are in essence doing is predicting points scored and points given up.

To make that point more clearly, I ran a regression of wins onto points scored and points given up. When I do so my predictions of team wins are, on average, 2.4 wins away from actual wins over the 1995-96 through 2004-05 seasons and 1.7 wins away in the 2003-04 season that the book highlights. (These are the same deviations they report in the book for their Wins Produced measure.)

Below I give how far off my predictions are (versus actual wins) and how far off the book's predictions are. (The book's are first, mine are second.) The correlation here is 0.96, so this it is clear that Berri and his co-authors have successfully produced a measure that aggregates up to predict offensive efficiency and defensive efficiency.

Code:
Team Book Mine
ATL 0.55 0.24
BOS 0.82 1.06
CHI 0.70 0.79
CLE 1.06 1.22
DAL 1.15 0.81
DEN 0.73 0.86
DET 2.75 3.67
GSW 2.10 1.92
HOU 0.90 0.92
IND 4.41 3.51
LAC 0.53 0.35
LAL 4.31 4.46
MEM 2.23 2.38
MIA 0.28 0.49
MIL 2.85 2.83
MIN 2.21 1.55
NJN 0.52 1.06
NOH 0.26 0.29
NYK 1.89 2.13
ORL 0.96 0.64
PHI 1.27 0.97
PHO 1.64 1.82
POR 3.10 3.69
SAC 0.38 0.54
SAS 3.44 4.40
SEA 2.31 2.15
TOR 0.09 0.59
UTA 4.41 4.53
WAS 0.68 0.74


But all of this still begs bigger questions.

(i) Does an analysis of how team statistics predict wins (which in essence is what Wins Produced does) tell us much about how to use statistics to apportion credit among players on a team?

One of the implications of this approach is that there is no room for credit to be given for shot creation. The authors do not arrive at this assumption empirically; it is simply something they assume given their approach.

In my opinion, understanding the value of creating shots is perhaps the most important aspect of analyzing basketball statistics. If, like in baseball, players each got a turn to take their shot, then this would not be an issue. But that is not the case, so I have a hard time making sense of an approach that assumes away what I consider to be a critical aspect of the game of basketball.

(ii) Remember that without the team adjustments (which have practically no effect on the relative ratings of players), Wins Produced likely does a terrible job predicting team wins. So what this says is that two versions of Wins Produced that are practically identical, one does a great job predicting team wins and the other does a pretty lousy job. What does this say about using the prediction of team wins as a barometer?

(iii) This is not really a criticism of Berri and his co-authors, but I have always felt that our box score statistics tell us more than we give them credit for. For the most part, this book follows the typical approach in logically relating the values of a point scored, field goal missed, rebound, turnover, etc. But I have always felt that these stats also tell us something about players in addition to the impact on the game at the time they occurred. Could guys who turn the ball over or a lot not be as good help defenders? Might the guy who gets steals do a better job keeping the floor spaced? Might the great rebounder do a better job catching tough pasess or picking up loose balls?

This is the logic I have used in relating my adjusted plus/minus ratings to points, rebounds, assists, steals, etc. And I have tended to find that the weights for these stats differ a lot from the logic-based approaches of Dean Oliver, John Hollinger, and Berri and his co-authors.

Last edited by Dan Rosenbaum on Fri Jul 28, 2006 10:05 pm; edited 2 times in total
Back to top
View user's profile Send private message Send e-mail Visit poster's website Yahoo Messenger
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina

PostPosted: Thu Jul 27, 2006 11:48 pm Post subject: Reply with quote
This is something that I suspect will spark some conversation. Except for the team adjustments (which shouldn't matter much for this), I approximated Wins Produced, Win Score, and lots of other measures and correlated them to my adjusted plus minus ratings. Here are the results.

[Note I had a typo in one of my formulas that I fixed, resulting in some new correlations relative to those I first reported.]

Wins Produced (no position adjustment): 0.3296
Wins Produced (position adjustment): 0.4545
Win Score (no position adjustment): 0.3079
Win Score (position adjustment): 0.4466
Old Win Score (no position adjustment): 0.1538
NBA Efficiency (no position adjustment): 0.3765
NBA Efficiency (position adjustment): 0.4460
PER (no position adjustment): 0.4345
PER (position adjustment): 0.4423
Offensive minus Defensive Rating (no position adjustment): 0.4137
Offensive minus Defensive Rating (position adjustment): 0.4481
Win Shares (no position adjustment): 0.4897
Win Shares (position adjustment): 0.4863
Win Shares per Minute (no position adjustment): 0.4327
Win Shares per Minute (position adjustment): 0.4420
My Statistical Adjusted Plus/Minus Rating: 0.5820

This is really interesting. Both Win Score and Wins Produced are both pretty terrible without position adjustments, but with the position adjustment they are not bad. It suggests that a lot of other methods have overvalued shot creation. But notice that the position adjustment is really important here - much more important than with any of the other metrics except for NBA Efficiency.

(Berri and co-authors have published at least one paper using what I call the Old Win Score, which ignored assists, blocks, and personal fouls. I also could not find any discussion of a position adjustment. That measure performs much worse than any other measure here.)

NBA Efficiency is is a lot like Wins Produced/Win Score - terrible without positon adjustments and good with them. In fact, if NBA efficiency is fixed so that similar efficiencies on two pointers and three pointers are counted the same, then NBA Efficiency with position adjustments is slightly better than Wins Produced/Win Score. This is contrary to what the authors argue in the book. Given how much of their argument in parts of their rests on their measure being better than NBA Efficiency (and not because of the position adjustments), these findings would significantly change their conclusions.

PER is much better than Wins Produced/Win Score without position adjustments, but once position adjustments are made, Wins Produced/Win Score does a little better than PER. But given that adjusting for position is much less important for PER, one might prefer PER on the grounds that positions are sometimes very difficult to determine for some players.

Dean Oliver's Offensive Rating minus Defensive Rating (at least I think this is Dean's) is worse than PER without position adjustments. But with position adjustments, it does a tiny bit better.

Win Shares comes out smelling like a rose. Without position adjustments it comes out better than any of the other measures so far. And unlike the other measure position adjustments actually make it worse, not better. Win Shares looks like it does not get as much credit as it should.

(Surprisingly, Win Shares per minute is correlated less with adjusted plus/minus ratings that regular Win Shares are.)

Finally, I included my statistical plus/minus metric, which does much better than any of the other measures. But that is not surprising, since it is designed to predict adjusted plus/minus ratings. Now I will not go into detail about how it is computed, but it does incorporate complicated adjustments for position and for other roles besides position. A really simple version of my statistical adjusted plus/minus measure without position adjustments has a correlation of 0.5380.

So we probably should not be so hard on Berri, Schmidt, and Brook as their measures do pretty well. There may be problems with the approach used to arrive at Wins Produced/Win Score, but once the position adjustments are included, they are pretty good metrics. These results are also a testament to Win Shares - doubly so since there is no need to position adjust with this measure.

Last edited by Dan Rosenbaum on Fri Jul 28, 2006 10:26 pm; edited 6 times in total
Back to top
View user's profile Send private message Send e-mail Visit poster's website Yahoo Messenger
deepak



Joined: 26 Apr 2006
Posts: 665


PostPosted: Fri Jul 28, 2006 12:43 am Post subject: Reply with quote
Thanks for this.

This is probably a silly question, but were you looking at Wins Produced, Win Scores, and Win Shares per possession? I know the rest are per-possession metrics, by definition.

Also, was the correlation to your adjusted +/- metric uniform across different types of players for each of the stats you looked at? For example, does PER do as good a job at estimating the effectiveness of guards as, say, big men?
Back to top
View user's profile Send private message
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina

PostPosted: Fri Jul 28, 2006 1:01 am Post subject: Reply with quote
deepak_e wrote:
Thanks for this.

This is probably a silly question, but were you looking at Wins Produced, Win Scores, and Win Shares per possession? I know the rest are per-possession metrics, by definition.

Also, was the correlation to your adjusted +/- metric uniform across different types of players for each of the stats you looked at? For example, does PER do as good a job at estimating the effectiveness of guards as, say, big men?

I put Wins Produced and Win Score into per 40 pace-adjusted minutes units, so that in essence makes them per possession. I will add a Win Share measure that is per minute, which is actually worse than the regular Win Share.

The correlations do vary by position. PER struggles the most relative to the other metrics with 3s, but does well with 1s and 2s. Wins Produced struggles the most with 4s, but does well with 5s. At this point I am not sure why.
Back to top
View user's profile Send private message Send e-mail Visit poster's website Yahoo Messenger
Mark



Joined: 20 Aug 2005
Posts: 807


PostPosted: Fri Jul 28, 2006 1:15 am Post subject: Reply with quote
Very interesting discussion you are opening here. After I read your first post I composed a response but your second post addresses one of the main thrusts I was going to make so I will try to adjust to that.

Regarding the second post:
I wonder if you could try to include protrade in the method comparison.
Understanding the variances from actual wins by each method is worth doing and perhaps looking at multiple years would also be helpful.

The main task is understanding what goes into offensive and defensive efficiency. The different methods present a choice: should we just count the traditional stats, or try to count more things (spacing, double teams, picks, blockouts, early ball movement, tips, saves, etc.), or focus on team productivity and give equal shares to those on the court as a team, or split credit to one, several or all players by rule and judgment of actual involvement in the play or the success or failure of the play based on play by play or tape? Should it be done at team level or by player? I think there is merit for both. Perhaps it would be possible to also lightly or vigorously test to find the best score with a blend of methods.

Would the authors of each method agree to a true prediction before the coming season?

Regarding the first post:
“Then they adjust their Wins Produced measure so that, on average, it is equal for every position. There really is no basis for this adjustment”

This does seem like an important point of discussion. How should players and positions be weighted? On offense if not equal then by their roles and role fulfillment? (As you apparently did) Team specific or league average? On defense perhaps it could be done by the average quality of players at that position (for example to address SF weakness)?

“Lastly, they adjust using team defensive measures.”
Ultimately an ideal player assessment would capture individual, team and help defense accurately rather than rely total on team roll-up equally for all.

“One of the implications of this approach is that there is no room for credit to be given for shot creation…In my opinion, understanding the value of creating shots is perhaps the most important aspect of analyzing basketball statistics. “

I agree it is important but shot creation and quality of shot creation are different things and to be clear quality of shot creation is different than shot results. Quality of shot creation could perhaps be measured in this new era with 82games and synergy data by a player’s past FG% for a set of zones, and shot types, coverages, and timing and perhaps even various aspects of matchup that play.

Viewing some stats are indicators of superior basketball ability and inferring to current uncounted activities that affect results may have some promise and some peril. For a speicifc player some insights can be posited about uncounted activities that might be affected by similar skills and physical abilities and checked by tape study and extreme charting for amount of correlation. Going to inferring from one skill to another for a set of players would needed to carefully checked as well.

” And I have tended to find that the weights for these stats differ a lot from the logic-based approaches of Dean Oliver, John Hollinger, and Berri and his co-authors.”

The correlation of these stats with winning may vary from logic but are they really more important than logic suggests? Can we prove that? I am asking for degree of confidence on this, not rejecting. Is it those actions specifically that is more important or are they in turn correlated with hidden uncounted factors? Should we just weight the indicator stats more heavily or continue to uncover the hidden uncounted and score them separately?

From your adjusted +/- explanation: "points, rebounds, and assists, the marginal effect of an additional point, assist, offensive rebound, and defensive rebound is 1.08, 1.14, 0.78, and 0.11 points per 100 possessions, respectively. “

Compared to Berri it would appear that in your adjusted +/- work you found assists much more correlated with success than Berri weights and defensive rebounding less correlated than Berri weights. Can further discussion push any further on these longstanding topics of debate?

“Players who attempt lots of 3 points and free throws appear to be more valuable than players who specialize in two point field goal attempts. “

Makes sense but how much should TS% be adjusted from its straightforward logical weighting? Could teams high or low on these and their overal FG% somehow be used to guess at the extra weights for these type of shooting strength? Or is that stretching too far? Context matters and varies by team, details of the player mix and lineups used on the floor and opponent.

“…the results suggest that holding all of these other game statistics constant, players who play more minutes tend to help their team point differential. This result would be expected if coaches observe and reward contributions not picked up in game statistics (e.g. good defense) by playing those players more minutes. Note, however, that the coefficient is not huge. Holding the other game statistics constant, the difference between a 20 minutes per game player and a 40 minutes per game player is only 2.16 points per 100 possessions – about the same as an extra steal per 40 minutes.”

“In the future I hope to add height and age/experience to these regressions. It appears to me that looking at the results in many of the tables that young, inexperienced players tend to have lower pure adjusted plus/minus ratings than their game statistics would suggest. It seems that the young, inexperienced players may not contribute as much to their teams in ways not picked up by game statistics.”

Thanks for sharing this data and the impression and look forward to perhaps hearing more in the future. Validates on average the coaches eye. Indeed it would be interesting to see more splits: by position for a given coach, perhaps by different coaches of same team over several years, and perhaps also league average by position, role, age, body type, contract.

"So we probably should be spending more time talking about why Wins Produced does a better job than PER. "

It seems pretty straightforward that wins produced captures more of defense than PER and that may be a main reason it performs better.

Adjusted +/- appears to win because it seeks to capture the traditional stats and the hidden uncounted, a broader sweep than any of these others. And applies a lot of statistically sophisicated and basketball smart work in the analysis of the information.

I still wonder how protrade's variable to player credit system based on rule from play by play or an even further enhanced credit assignment system based on tape could do compared to adjusted +/-.

Last edited by Mark on Sat Jul 29, 2006 10:47 am; edited 10 times in total
Back to top
View user's profile Send private message
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina

PostPosted: Fri Jul 28, 2006 1:32 am Post subject: Reply with quote
Thanks for all of the great comments, Mark. I cannot promise that I will get to all of them, but over time (and keeping in mind that I can't reveal all of my tricks), I will see what I can get to.

Last edited by Dan Rosenbaum on Fri Jul 28, 2006 10:27 pm; edited 2 times in total
Back to top
View user's profile Send private message Send e-mail Visit poster's website Yahoo Messenger
Mark



Joined: 20 Aug 2005
Posts: 807


PostPosted: Fri Jul 28, 2006 1:39 am Post subject: Reply with quote
10-4.
Back to top
View user's profile Send private message
deepak



Joined: 26 Apr 2006
Posts: 665


PostPosted: Fri Jul 28, 2006 2:36 am Post subject: Reply with quote
I don't understand how per-minute Win Shares could do worse than raw Win Shares. Seems to go against the fundamental APBRmetrics principle that per-minute (or per-possession) stats are superior. Any thoughts on this?
Back to top
View user's profile Send private message
Mark



Joined: 20 Aug 2005
Posts: 807


PostPosted: Fri Jul 28, 2006 2:43 am Post subject: Reply with quote
I assume most or all of these methods can be made to fit tighter with the win-loss pattern if they included some summary measure of variabilty. It would be interesting to see how much better they get after this tweak and to compare the amount of variability in the results of the methods and the amount of impact of the variability factor in explaining results. GMs could by this method perhaps determine a little more exactly what type of adjustment to make?

(I assume in the method result comparison list all of the methods are fit to win-loss based on season averages.)

Last edited by Mark on Fri Jul 28, 2006 7:24 pm; edited 1 time in total
Back to top
View user's profile Send private message
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina

PostPosted: Fri Jul 28, 2006 9:43 am Post subject: Reply with quote
deepak_e wrote:
I don't understand how per-minute Win Shares could do worse than raw Win Shares. Seems to go against the fundamental APBRmetrics principle that per-minute (or per-possession) stats are superior. Any thoughts on this?

Maybe Justin when he gets back from a three-day road trip will have some thoughts on this. I am hardly an expert on Win Shares, and my intution follows yours.
Back to top
View user's profile Send private message Send e-mail Visit poster's website Yahoo Messenger
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina

PostPosted: Fri Jul 28, 2006 9:56 am Post subject: Reply with quote
I took a look at how well the different measures correlate with offensive and defensive adjusted plus/minus ratings and those results were pretty interesting too.

The problem with PER is that it explains almost none of defense. It's correlation with offensive adjusted plus/minus is higher than my statistical plus/minus measure - although some of that is due to my measure being more than a measure of offense.

Wins Produced/Win Score - even without the team adjustments for defense - do a pretty good job on defense, but this is where my statistical plus/minus measure gains a lot of traction versus all of the measures. It is hard to assess defense with box score stats, but it is possible to do quite a bit better than the logic-based approaches do. I think what is going on is I am able to pick up a lot of the "uncounted" contributions on the defensive side that are correlated with box score stats.
Back to top
View user's profile Send private message Send e-mail Visit poster's website Yahoo Messenger
Mark



Joined: 20 Aug 2005
Posts: 807


PostPosted: Fri Jul 28, 2006 11:52 am Post subject: Reply with quote
The biggest uncounted in many methods is shot defense. Offensive rating - Defensive Rating and Win Shares also have it in their roll-up. Protrade's credit/blame system also covers it. (So does defensive tendex and cumulative tendex incorporating it. EWins doesnt but could still be added to the method comparison.)

Protrade could in my view be tweaked and improved- currently opponent made shot blame and missed shot credit surprisingly gets split equally. I missed this point before; this is similar to adjusted +/-'s equal handling of shot defense credit and blame. I can understand the philosophy that defense is a team endeavor that would be used to explain this equal scoring choice but I personally might go 2/3 credit or blame to direct defender and 1/3 to rest of team but maybe 50-50 would be ok too. I dont see why shotdefense shouldnt be fairly similar to scoring shooting- protrades only gives 70-75% of the credit or blame to the shooter not 100%. (Ideally if based on tape it would cover defensive switches accurately and to allow blaming the guy whose breakdown somewhere lead to the evental shot- some or a lot).

With regards to short creation, with game shot charts by player it would be possible if they were scored against 82games chart of player scoring by distance or exact zone or synergy's list by zone and type of shot to calculate the quality of shots taken based for each shot on the average historical FG% of that player from that spot with heavy contest, contest or no contest and then compiled the FG% and find the average expected FG%. (i.e. a corner 3pt shot has a player historical FG%, as does a wing 3pt shot-a different one- as does 17 footer, 10 footer, a layup, etc.)

You could then by comparing the shot creation average expected FG% to actual FG% add specificity to the old adage he or we "got good shots but didnt make them" or more fully appreciate good nights making a high rate of tough shots.

Adding fouls and free throws you could make the comparison shot creation expected TS% vs actual TS%. Or simply label shots taken good, fair and poor ones and be evaluated on that and the actual returns.

Dan, in general terms does the adjustment work you do on the +/- data by player give scores produced by this method awareness of game to game volatility and scoreboard/win impact and serve as an advantage in the comparison to the other methods which do not have that as scored? Or not? You have an edge over some by fully covering defense and arguably over all in capturing the impact of other uncounted activity associated with a player on/off by player but I wondered if this might be another source of your edge?

Last edited by Mark on Fri Jul 28, 2006 7:58 pm; edited 1 time in total
Back to top
View user's profile Send private message
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina

PostPosted: Fri Jul 28, 2006 3:37 pm Post subject: Reply with quote
Mark wrote:
Dan, in general terms does the adjustment work you do on the +/- data by player give scores produced by this method awareness of game to game volatility and scoreboard/win impact and serve as an advantage in the comparison to the other methods which do not have that as scored? Or not? You have an edge over some by fully covering defense and arguably over all in capturing the impact of other uncounted activity associated with a player on/off by player but I wondered if this might be another source of your edge?

I am not sure what you mean by "game-to-game volatility." Of course, player productivity is volatile, but what our metrics capture is the average of that volatile productivity.

By using adjusted plus/minus ratings as a barometer, I am not saying that it necessarily is the best way to evaluate individual players. In lots of cases sample sizes are too small to get accurate adjusted plus/minus ratings. (Among players playing less than 500 minutes per season, the correlation between adjusted plus/minus ratings from year to year is practically zero.) But in theory the adjusted plus/minus rating should capture practically all of the contributions a players makes towards winning - not just those we happen to record in a box score.

So if we think of the adjusted plus/minus rating as being made up of the true impact of a player and a random component that isn't related to anything, then it should be a very good barometer of which of these box score stat based metrics does the best job measuring that impact. Because of the random component, the correlation will never be one, but we can still compare the relative correlations.

One last point. Adjusted plus/minus ratings are context specific. But so are box score statistics. Put Steve Kerr is a situation where he had to create most of his team's shots and his efficiency would likely fall like a rock - just like his adjusted plus/minus rating.
Back to top
View user's profile Send private message Send e-mail Visit poster's website Yahoo Messenger
Mark



Joined: 20 Aug 2005
Posts: 807


PostPosted: Fri Jul 28, 2006 6:08 pm Post subject: Reply with quote
I struggled with how to phrase the question a little but basically I was seeking reassurance that the method comparison was apples to apples, that is all averages and not adjusted for game to game volatility.

"Adjusted plus/minus ratings are context specific." I am trying to understand how much.

I wondered if the analytic process for adjusted +/- provided an opportunity to capture and adjust the player's result +/- (different from the raw input +/-) for game to game volatility in a way that the other methods hadnt gone thru but could with some form of tacked on volatility measure. I wondered if it is was fair or not to say that the adjustment process was capturing not just the average +/- but more about the impact of the "when" of the data: the strong and weak stretches of play, and how frequent they were, and how they were distributed over games and perhaps how they impacted win-loss.

I'll make one more try at explaining what I meant with an example:
Would three players otherwise exactly identical on team +/- season data score the same under adjusted +/- if one had a really outstanding game every 5 nights and 4 games modestly below the resulting average, one played exactly the same every time right to that same average, and the last one played 3 games modestly above the resulting average but 2 games further below the average?

Somewhat separate from the above talk, your adj. +/- method I believe gives extra weight to clutch time actions, which would be a perfectly legitimate, smart competitive advantage over almost all the other methods. It is a small numbers of plays but could have big impact in matching to win-loss depending how it is done. If another comparison was made for adjusted +/- with/without clutch weighting how much of the competitive advantage comes from that feature vs. capturing the hidden uncounted? If the other methods built in clutch weighting (as protrade also does) how much ground could they make up?

I offer these comments in the spirit of understanding more about the results of the method comparison and looking for best practices and urging wider adoption of them.

Last edited by Mark on Sat Jul 29, 2006 10:50 am; edited 1 time in total
Back to top
View user's profile Send private message
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina

PostPosted: Fri Jul 28, 2006 9:02 pm Post subject: Reply with quote
Mark wrote:
"Adjusted plus/minus ratings are context specific." I am trying to understand how much.

I wondered if it is was fair or not to say that the adjustment process was capturing not just the average +/- but more about the impact of the "when" of the data: the strong and weak stretches of play, and how frequent they were, and how they were distributed over games and perhaps how they impacted win-loss.

I'll make one more try at explaining what I meant with an example:
Would three players otherwise exactly identical on team +/- season data score the same under adjusted +/- if one had a really outstanding game every 5 nights and 4 games modestly below the resulting average, one played exactly the same every time right to that same average, and the last one played 3 games modestly above the resulting average but 2 games further below the average?

These three players would get the same adjusted plus/minus rating. The only "when" I account for is that I weight each possession by the probability that an extra point would change the outcome of the game with playoffs counting more than the regular season.

Quote:
Somewhat separate from the above talk, your adj. +/- method I believe gives extra weight to clutch time actions, a perfectly legitimate, smart competitive advantage over almost all the other methods. It is a small numbers of plays but could have big impact in matching to win-loss. If another comparison was made for adjusted +/- with/without clutch weighting how much of the competitive advantage comes from that feature vs. capturing the hidden uncounted? If the other methods built in clutch weighting (as protrade also does) how much ground could they make up?

I am not sure how to answer the question, but the weighting does matter.
Last edited by Crow on Thu May 12, 2011 1:55 am, edited 1 time in total.

Crow
Posts: 6146
Joined: Thu Apr 14, 2011 11:10 pm

Re: Wins Produced - Wages of Wins (Berri, Schmidt, and Brook

Post by Crow » Fri Apr 15, 2011 7:11 pm

page 2

Author Message
Mark



Joined: 20 Aug 2005
Posts: 807


PostPosted: Fri Jul 28, 2006 9:38 pm Post subject: Reply with quote
Ok thanks for setting aside both these questions. I didnt realize the clutch adjustment was so limited in application. Appreciate the sharing of the method results.
Back to top
View user's profile Send private message
deepak



Joined: 26 Apr 2006
Posts: 664


PostPosted: Sat Jul 29, 2006 2:02 am Post subject: Reply with quote
Dan, have you looked at how raw +/- publicly available at 82games correlates to your adjusted +/-? Also, the Roland Ratings?

I'd expect that the Roland Ratings do a better job than PER. Still not sure about win shares.
Back to top
View user's profile Send private message
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina

PostPosted: Sat Jul 29, 2006 9:42 am Post subject: Reply with quote
deepak_e wrote:
Dan, have you looked at how raw +/- publicly available at 82games correlates to your adjusted +/-? Also, the Roland Ratings?

I'd expect that the Roland Ratings do a better job than PER. Still not sure about win shares.

Unfortunately, I have not put together that data.
Back to top
View user's profile Send private message Send e-mail Visit poster's website Yahoo Messenger
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina

PostPosted: Sat Jul 29, 2006 10:37 am Post subject: Reply with quote
One of the main points of Wages of Wins is that according to the evidence presented in the book, it appears that NBA decision-makers are irrational. They write: "It is not that people in the NBA are lazy or stupid. It is just that the tools at their disposal do not allow them to see the value of the various actions players take on the court."

The argument that supports this conclusion is they show that Wins Produced explains wins much better than NBA Efficiency, but NBA Efficiency is much more closely tied to salaries and All-Rookie Team voting (done by coaches). Thus, according to their evidence, it is irrational for teams to not be using something like Wins Produced to make their decisions.

OK, but the difficulty with this argument is that it hinges on Wins Produced better explaining wins than NBA Efficiency. Wins Produced does a great job explaining wins because of their team defense adjustment, but the authors admit that this adjustment has very little effect on their relative rankings of players. So if it doesn't matter much for the relative rankings of players, I just don't see how it can be used as a justification for the methodology. To me, that whole exercise raises a big red flag about the validity of using the prediction of team wins as a barometer for a metric for individual players.

But then if we move to another barometer - adjusted plus/minus ratings - we see that Wins Produced only performs better than NBA Efficiency if position adjustments are used for Wins Produced but not for NBA Efficiency. That significantly changes the story of much of their book. Instead of a story about NBA teams overvaluing scorers, their story becomes one that NBA decision-makers are irrational because they don't properly position adjust.

Moreover, the authors provide little justification for their position adjusting, especially in relation to how important it is to their metric. They argue that big players would have difficulty filling the roles of guards; i.e. a team could not play all centers. But if centers truly are worth more than guards as their unadjusted Wins Produced suggests, this would not be the only reaction of NBA decision-makers. Instead of playing centers at guard, what would happen would be that they would pay centers more than guards - which is precisely what does happen. So rather than proving conventional wisdom wrong, maybe the authors have provided justification for conventional wisdom.

Now I am not necessarily trying to defend NBA decision-makers as being super-rational. Lots of points made in Wages of Wins are points I agree with wholeheartedly. And I highly recommend that everyone in this APBRmetric community read this book. You will find lots that you agree with in this book and lots that will force you to think more deeply about things.

But when we go to cast stones at NBA decision-makers, we need to be sure that our own house is in order. And that is really my biggest complaint about Wages of Wins. My experience has been that NBA people often do a much better job than we give them credit for. I work for Cleveland, and last year when we traded for Flip Murray I was against it. Flip was the lowest rated two guard in my system. (And I am sure he would not be rated too highly in Wins Produced.)

But you know what? Flip did not play too badly for us. He was not a star or even a good starter, but he made some changes to how he played in Seattle and he contributed to us doing well down the stretch and in the playoffs. So in that situation if Danny had listened to me (or probably consulted Wins Produced), the team would have won fewer games.

There is a lot about this game that we in this community know well, but there is a lot we don't know well. I strongly believe that good stats work can play an important role in a well-run organization. But I vehemently disagree with those who are ready to start calling NBA people irrational because of some results from a possibly mistaken empirical analysis. We all can work on being better than that.
Back to top
View user's profile Send private message Send e-mail Visit poster's website Yahoo Messenger
Analyze This



Joined: 17 May 2005
Posts: 364


PostPosted: Sat Jul 29, 2006 11:07 am Post subject: Reply with quote
I find it strange that your posting doesn't get more replies from people who are mentioned in your post or other leading apbr metricians. (only Marc is heavenly responding) The fact that you are sometimes answering your own posts shows how important you find the remarks that you made in the original post of this topic. I've read Berri his book, just like I've done with Oliver his work. (and Hollinger). The problem I have with remarks like “And I have tended to find that the weights for these stats differ a lot from the logic-based approaches of Dean Oliver, John Hollinger, and Berri and his co-authors.” or the other judgements you make about different methods is that I can’t give a value to your work because you are not making public how it works. If you decide to make something for the money and don't give comment about " your tricks" you are not helping the stat community as a whole, something that Oliver, Berri and others did, but you are just helping the Cavs and are basically just cashing in. Don't get this the wrong way. I've read from different people that you are a really nice guy. So you probably are. But don’t start telling how much bettter your results are without making your method public. How can the other analysts respond if they don't know how your method works exactly.
Back to top
View user's profile Send private message
Mike G



Joined: 14 Jan 2005
Posts: 3558
Location: Hendersonville, NC

PostPosted: Sat Jul 29, 2006 11:12 am Post subject: Reply with quote
Dan Rosenbaum wrote:
... My experience has been that NBA people often do a much better job than we give them credit for. .


Thanks for taking the high middle ground, Dan. While some among us fit the description, others (I consider myself one of these) presume that competency is the norm in such a megadollar business as NBA ownership.

The various (and radically different) forms of analysis seen in this forum are proof positive that there is not a single "Us" position. (See any thread involving Antoine Walker.) I suppose the evolution of all efforts will someday provide such things as salary that's commensurate with value, formulaic team-building, etc.

There really should be a good number of dead-end routes per each new breakthrough, or else we aren't turning every stone.

Meanwhile, I've been doing eWins, which seems to be unique in one aspect: individual eWins are not meant to add up to team Wins. It seems the sum of individual departure from average is only half the team departure: A squad of players totalling 51 eWins (41+10) will win 61 games (41+20).

W = eW*2 - 41

My multiplier seems to be 2, but it might really be 2.03, or 1.98. Or something more complex. Other systems might find 1.4 or 2.2 to be appropriate. In any event, since a scoring advantage of 10% over the course of the season, between team A and team B, will surely produce much more than 10% difference in their W/L%, doesn't it seem reasonable that individual Wins Contributed (or whatever name one chooses) should not add up to Team Wins?
_________________
`
36% of all statistics are wrong

Last edited by Mike G on Sat Jul 29, 2006 11:48 am; edited 1 time in total
Back to top
View user's profile Send private message Send e-mail
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina

PostPosted: Sat Jul 29, 2006 11:23 am Post subject: Reply with quote
Analyze This wrote:
I find it strange that your posting doesn't get more replies from people who are mentioned in your post or other leading apbr metricians. (only Marc is heavenly responding) The fact that you are sometimes answering your own posts shows how important you find the remarks that you made in the original post of this topic. I've read Berri his book, just like I've done with Oliver his work. (and Hollinger). The problem I have with remarks like “And I have tended to find that the weights for these stats differ a lot from the logic-based approaches of Dean Oliver, John Hollinger, and Berri and his co-authors.” or the other judgements you make about different methods is that I can’t give a value to your work because you are not making public how it works. If you decide to make something for the money and don't give comment about " your tricks" you are not helping the stat community as a whole, something that Oliver, Berri and others did, but you are just helping the Cavs and are basically just cashing in. Don't get this the wrong way. I've read from different people that you are a really nice guy. So you probably are. But don’t start telling how much bettter your results are without making your method public. How can the other analysts respond if they don't know how your method works exactly.

Point well taken, but my piece at 82games http://www.82games.com/comm30.htm lays out the basic points of my methodology in more detail than most books or academic papers would allow - certainly moreso than Wages of Wins. Also, I do not answer every question about my methodology, but I have spent a lot of time in a lot of different forums talking about details about my work. Now I cannot give out my data or my results, because they are not mine to give out (they never have been even before I worked for the Cavs). But I agree that without others replicating my work, it is hard to put it into perspective. But I have laid out the methodology in a lot of detail; I can't be responsible for others not replicating my work.
Back to top
View user's profile Send private message Send e-mail Visit poster's website Yahoo Messenger
Mark



Joined: 20 Aug 2005
Posts: 807


PostPosted: Sat Jul 29, 2006 11:44 am Post subject: Reply with quote
I agree that listening to, looking for the intelligence of others is wise as well as trying to find the faults. Refining one's own knowledge both thru self-directed research and awareness of and absorption of what is useful from the research and observations of others.

With regards to Flip Murray both sides had arguments before the acquisition and then afterwards both sides could point to time periods and parts of the game where the choice worked (regular season, with Snow, including clutch shooting regular season) and where it didnt (playoffs). Snow PG / Murray SG combo worked pretty well regular season. Flip had third highest win % on court but that can be read two ways, and both be part of the story, Flip working and Flip put in good context to be safe.

According to 82 games Cavs went from playing Flip at point almost not at all regular season to 20% in playoffs, to try to capture his historical ability to really score on PGs but he didnt have it going and their quickness and craftiness burned him back threefold on points per 48 minutes (including 10 FTAs per 48 at PG compared to only half that at SG). Murray's poor performance in limited floortime in 2004-5 playoffs could have read different ways (his offense game doesnt work as well against playoff's higher level of defense vs he didnt get enough run to judge) but it was a key showdown dataset on Murray. Acquiring him has to be put in context of immediate regular season need, who else they could get, and prospects for getting Hughes back for playoffs and in what condition. Perhaps they could have spent more to get someone else, proven in playoffs.


I agree with deepak that it would be useful to see unadjusted +/- and roland ratings added to the comparison mix. The Heroes of the Hardwood book too (and protrade) if time and interest was there for the fullest possible comparison. Might you publish in an academic journal about this type of method performance comparison and the surrounding context?


With regards to replication of adjusted +/-, others previously talked about attempting to do so and bring the method into the public view. I hope they will do so but it would be a huge task to replicate the database and then navigate all the analysis so I doubt it will happen.

The other methods all seem to bump up against a limit on how much they can explain, a limit on correlation of results. This may show that using the same dataset of the commonly tracked, the seen has a max regardless of method.

Dan's method capturing team play and the hidden uncounted contributions to it has the lead and should get more study from us as best we are able to gather from him or independently. Available unadjusted player pair data might allow some questions to be asked, answered approximating insights available thru adjusted +/- datasets.

I assume Dan's comment about not being free to release the adjusted +/- data even before the Cavs work is because the data came from 82games under conditions limiting how much could be revealed? Although it is easy and understandable to want revelation, I accept business need for confidentiality that will withhold much of what teams might pay for. Really we are very fortunate 82games shares so much, so thanks again for that and thanks Dan for sharing as much as you feel you can about method and implications. Although I ask a lot of questions, reply to and use as you will. I ultimately respect your limits and trying to be a leading, considerate participant in really 3 worlds.


I remain interested in Mike's Ewins because of the different approach he notes. It is more traditional use of the tracked stats and individual based. More knowledge about transference of players and transferability of their win producing value from one team to another needs more work to aid trade and free agent analysis and even the draft.. Methods that are good at capturing contributions to team play and individual based methods are both important to consult.

Identifying different roles, different successes levels in those roles of the same player on same team also needs more precise data sorting and analysis. All the methods cited in this thread could with unlimited time and resources be broken into data subsets to evaluate players by different positions and roles. Teams certainly should.

Mike, do you plan any further enhancements to eWins in near future? To more fully capture defense, including one on one defense?

Would you want to give Dan the information necessary to score your method the same way, against the others if he were inclined? Or Dan is there anything more you could say that would allow others to pursue accurate comparable scoring of other methods similar to your scoring if you arent inclined to score more or all possible method contenders and variations? Rating the methods and putting light on the best elements and the weaknesses seems like a very timely and important thing to do, to help kick off the next era in player and team evaluation.

What should be done? Capture team and individual contributions. Defense as fully as offense. Use many methods, refine them but to some extent preserve their uniqueness. Consider a best practices grand theory or blended meta-ratings. Count as much as possible, beyond traditional stats using tape and also see the game wholistically. Compare players by position and role but be very flexible about what you do with position and role given the talents of the players you have, the mix and the potential of various lineups against the range of challenges by opponents. Try to measure and rate the coaches input more. It can be an endless search to know more and use that knowledge...

Last edited by Mark on Fri Aug 04, 2006 11:07 pm; edited 5 times in total
Back to top
View user's profile Send private message
HoopStudies



Joined: 30 Dec 2004
Posts: 705
Location: Near Philadelphia, PA

PostPosted: Sat Jul 29, 2006 1:02 pm Post subject: Reply with quote
Dan Rosenbaum wrote:
...
There is a lot about this game that we in this community know well, but there is a lot we don't know well. I strongly believe that good stats work can play an important role in a well-run organization. But I vehemently disagree with those who are ready to start calling NBA people irrational because of some results from a possibly mistaken empirical analysis. We all can work on being better than that.


As Dan knows, the concept of rationality is an important issue in economics. There is an assumption of rational decision makers and economists are often looking for evidence of that rationality, if not using the assumption to show something else. The Malcolm Gladwell book, Blink, discusses how experts can come to a more rational conclusion in the first two seconds of thought than others can with some analysis. An expert's "gut feel" may not be easily explained or easily justified, but it is often shown to be justifiable after very extensive follow up. The simple analysis often used to justify a decision when experts are saying it "seems wrong" can very definitely be wrong.

We are here to do thoughtful analysis, not to do quick and dirty things that fly in the face of an expert's blink.

My phrasing of this concept was as follows about a year ago:

hoopstudies wrote:

My null hypothesis is usually traditional coaching or management wisdom. So a hot hand exists, defense wins championships, and statistics are irrelevant until I prove otherwise (which I think I've done in many cases). Others may choose a different null hypothesis, but I think mine makes sense because I work with coaches and management and I'm not [in the position of] Billy Beane -- it is my burden to prove things, not theirs.

Bottom line, though, is that it does go both ways and I feel a significant burden of proof for anything that strays from traditional wisdom.


That being said, there is irrationality in sports (people in sports have said that to ME). There is irrationality within every decision-making group I have ever seen. There is irrationality among the discussions here. Any GROUP will have its irrational decisions because of competing interests and significant inefficiencies in negotiation to reach a group decision.

(Or at least apparently irrational. There are reasonable justifications for some apparently excessive salaries, for instance. It is an interesting economic phenomenon associated with the limited supply of players and extreme non-commoditized market - rational short and moderate term thoughts can make for an apparently irrational long-term or steady-state market. Highly skewed distributions of value makes it much easier to have mistakes, too. Are mistakes irrational? Not necessarily.)

Wages of Wins definitely makes you think about this kind of stuff, whether you agree with the specifics of the methodology or not.
_________________
Dean Oliver
Author, Basketball on Paper
The postings are my own & don't necess represent positions, strategies or opinions of employers.
Back to top
View user's profile Send private message Visit poster's website
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina

PostPosted: Sat Jul 29, 2006 1:02 pm Post subject: Reply with quote
Mark wrote:
With regards to replication of adjusted +/-, I hope the others who previously talked about attempting to do so and bring the method into the public view will do so.

I want to address this topic, because it came up with Analyze This, as well. Replicating adjusted plus/minus ratings, at least doing it the way I do it, is never going to be something that can be done on a wide scale.

Most of this community is comfortable working with datasets numbering in the thousands in something like MS Excel. The dataset that I use for my adjusted plus/minus ratings numbers in the hundreds of thousands and it would be bigger if I had not condensed it.

So that means computing adjusted plus/minus ratings requires a person to learn a programming language like SAS or Stata (or something more basic). That is something that most people do not have the time to do.

It also requires becoming very proficient at working with complex datasets. This is what I teach and while I am NOT a very good economist, I am probably in the upper echelon of economists in terms of putting together complex datasets.

After putting the data together and getting the model to run, understanding how to make sense of the results requires very good intuitive econometrics skills. Again, this is one of my strengths relative to other economists, since I do a lot of teaching of econometrics at the undergraduate, MA, and Ph.D levels.

So the point here is not that I am sort of genius. I certainly am not. It just so happens that my skill set is pretty rare (even for economists) and just so happens to be a good fit for computing adjusted plus/minus ratings. It is really just a right place at the right time kind of thing.

To put this into perspective, even with my skill set, taking my programs and converting them from one program to another (SAS to Stata) took me several weeks this Spring and Summer. And that is with the help of a graduate student of mine - one who has been trained by me to do this kind of work and who is very smart and very good at this kind of stuff. I think he could attest to how difficult this was. And this was only taking a set of programs that I had already written and converting them into another program.

I don't write this to scare people off from replicating my work. I encourage others to do so, and Winston and Sagarin are not the only folks to compute adjusted plus/minus ratings. So it can be done. But I thought this was as appropriate of a place as any to warn folks to be prepared that doing is no small feat.

That said, hopefully someone can figure out how to do what I do but in a much simpler fashion. Obviously I am not smart enough to do that.

So the point here is that it is going to be very difficult to bring adjusted plus/minus ratings into the mainstream in the same way as PER or Dean's work. And I understand the frustration and skepticism that results from this.
Back to top
View user's profile Send private message Send e-mail Visit poster's website Yahoo Messenger
Mark



Joined: 20 Aug 2005
Posts: 807


PostPosted: Sat Jul 29, 2006 3:33 pm Post subject: Reply with quote
I revised my remark about hoping for replication by following it with skepticism it will be done, but perhaps someone passionate about one team might try it for 1/30th of the full league. That could be a way to start and have data in public and further method discussion. Probably some pitfalls compared to doing all 30 but probably still enough benefits to look at it.
Back to top
View user's profile Send private message
Ben



Joined: 13 Jan 2005
Posts: 264
Location: Iowa City

PostPosted: Mon Jul 31, 2006 1:40 pm Post subject: Reply with quote
Mark wrote:
I revised my remark about hoping for replication by following it with skepticism it will be done, but perhaps someone passionate about one team might try it for 1/30th of the full league. That could be a way to start and have data in public and further method discussion. Probably some pitfalls compared to doing all 30 but probably still enough benefits to look at it.


I still think the big problem is getting the data (you need the full league's data). There are enough people out there with the skills in applied statistics and the popular statistics programs that I'm confident the ratings would get done.
Back to top
View user's profile Send private message
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina

PostPosted: Mon Jul 31, 2006 4:56 pm Post subject: Reply with quote
I like to tout this group as a pseudo-academic community, so I would love to hear any thoughts folks have about Wages of Wins. I suspect that if a larger swath of our community chimes in about the book, there would be a real chance of getting the authors to maybe respond. If it is just me and a couple others, I doubt they will.

It would be really great to see this group interacting with the sports economists that are interested in the same topics we are. There would be a great benefit for all groups if there was more interaction.
Back to top
View user's profile Send private message Send e-mail Visit poster's website Yahoo Messenger
Mark



Joined: 20 Aug 2005
Posts: 807


PostPosted: Mon Jul 31, 2006 7:03 pm Post subject: Reply with quote
It has only been out 2 months. It is summer. Apparently half or somewhat more on basketball- perhaps if 100% basketball it might have move even faster into this market. The reviews, the news articles and the blog may have taken some of the mystery out of it. On Berri's blog I see he has responded to some questions and comments. Further discussion could be beneficial.

Last edited by Mark on Fri Aug 04, 2006 11:08 pm; edited 5 times in total
Back to top
View user's profile Send private message
Ben



Joined: 13 Jan 2005
Posts: 264
Location: Iowa City

PostPosted: Tue Aug 01, 2006 11:25 am Post subject: Reply with quote
Dan Rosenbaum wrote:
I like to tout this group as a pseudo-academic community, so I would love to hear any thoughts folks have about Wages of Wins. I suspect that if a larger swath of our community chimes in about the book, there would be a real chance of getting the authors to maybe respond. If it is just me and a couple others, I doubt they will.

It would be really great to see this group interacting with the sports economists that are interested in the same topics we are. There would be a great benefit for all groups if there was more interaction.


It could be that most haven't read the book, but I also think disagreements provide the most active threads. My guess is most people find your analysis convincing.

Crow
Posts: 6146
Joined: Thu Apr 14, 2011 11:10 pm

Re: Wins Produced - Wages of Wins (Berri, Schmidt, and Brook

Post by Crow » Fri Apr 15, 2011 7:55 pm

page 3

Author Message
Mark



Joined: 20 Aug 2005
Posts: 807


PostPosted: Tue Aug 01, 2006 12:28 pm Post subject: Reply with quote
There are other reviews arriving. In addition to Dan's, you can reach one by Roland thru his site, and I see David Lewin has a review of it up there now too, and there is a note saying Kevin will have one there soon.
Back to top
View user's profile Send private message
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina

PostPosted: Tue Aug 01, 2006 4:28 pm Post subject: Reply with quote
Here are the reviews. There is free acess to Roland's; you just have to give your e-mail address.

Roland Beech
David Lewin
Kevin Pelton
Back to top
View user's profile Send private message Send e-mail Visit poster's website Yahoo Messenger
Mark



Joined: 20 Aug 2005
Posts: 807


PostPosted: Tue Aug 01, 2006 6:56 pm Post subject: replacement level FG% and shot creation value Reply with quote
What is the replacement level / no real shot creation value added FG%?

If you simply take a typical contested 2 point shot? Ed Peterson's extreme charting provides one estimate at between 35-40%. http://www.82games.com/saccon.htm

Team by team data for shots in crunch last 3 second of shotclock also seems to center roughly around 40% (I didnt actually calculate the average). http://www.82games.com/clockx.htm

Perhaps players should get credit in some fashion (as a separate stat or in larger formulas) for varying level of shot creation value if they get a shot with expected FG% better than this new baseline level of a shot instead of using the perhaps too high standard of 50% or even league average (which have shot creation value in them)? Of course actual success / failure would still be scored same as value of basket (or scoring play) / no basket.

It is another way to view things, maybe too lenient but I wanted to try to find a way to show credit for shot creation.


(Average actual FG% of players by position within a certain distance of replacement level on overall player quality might be worth looking at too as an additional reference, but because it can be affected by shot creating impact of others, or reflect low man of the totum pole, and against subs and garbage time I'd prefer to stick the approaches mentioned above.)

Last edited by Mark on Tue Aug 01, 2006 9:27 pm; edited 1 time in total
Back to top
View user's profile Send private message
Mark



Joined: 20 Aug 2005
Posts: 807


PostPosted: Tue Aug 01, 2006 9:25 pm Post subject: Search for how best to distribute the shots Reply with quote
Has anyone created or seen some sort of game theory model where for a specif team matchup and each the five positions on the floor you compared own TS% and TS% allowed and established some help defense rules that would dampen FG% where applied at some cost elsewhere and determined (by simulation?) the optimal shot distribution?

Bob C., does your simulation need to take a shot distribution or can it find the optimal one too? Quickly or would be require a really long process? Need to apply much more computing horsepower?

Dan R. would it be possible to work with the massive dataset you used for adjusted +/- to chronicle actual shot distribution game to game by position. role, player quality with the counterpart offense /defense FG% matchup data for each play summed up and see at least at a surface level how consistent actual shot distribution was compared to what a strategic analysis might suggest was optimal? And if there were any common themes in which overall game shot patterns succeeded above expected or at some simplified level (backcourt/frontcourt, inside/outside/3, star/role player) which play by play sequences succeeded above expected?

Last edited by Mark on Wed Aug 02, 2006 12:56 pm; edited 1 time in total
Back to top
View user's profile Send private message
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina

PostPosted: Tue Aug 01, 2006 11:21 pm Post subject: Reply with quote
Mark, game log databases can be used in a lot of ways, and what you suggest may be possible.

Back to a question above, I was able to regress a player's offensive adjusted plus/minus rating onto their true shot attempts per 40 minutes. I was able to hold constant a player's true shooting percentage, how many "extra" points they scored relative to if they had shot 50 percent, their turnover rate, and their rebound rate.

The things I am holding constant should account for practically everything that the given player contributes to offensive efficiency directly through their shooting, turnovers, and rebounding. What the coefficient on true shot attempts per 40 minutes tells us is how much scorers help offensive efficiency of their teammates. And the upshot is that increasing true shot attempts by 4 (one standard deviation) increases offensive efficiency by about 1 point per 40 minutes.

Take away Allen Iverson's true shot attempts and replace him with a player who shoots an average amount of shots (but is otherwise the same) and we are looking at offensive efficiency dropping by about 3 points per 40 minutes.

Moreover, only about 1/3 of this effect is due to teammates' shooting efficiency increasing. The bulk of the effect comes from teams with scorers turning the ball over less.

These effects are highly statistically significant (t-stat=9.12), so this is strong evidence that scorers do, on average, positively affect their teammates. The surprise is that the effect is mostly through teammates commiting fewer turnovers rather than shooting better.
Back to top
View user's profile Send private message Send e-mail Visit poster's website Yahoo Messenger
Nikos



Joined: 16 Jan 2005
Posts: 346


PostPosted: Tue Aug 01, 2006 11:56 pm Post subject: Reply with quote
Awesome work Dan. I never really thought about it in terms of teamattes not commiting turnovers, but intuitively I always felt that even Iverson is helping his team if his offensive efficiency is below his teams average by a decent amount. His high Usage does indeed help the offense. Only question is would his efficiency rise a lot with less Usage and more offensive help? AKA what would Iverson's Dean Oliver skill curve be like with a good quality offensive supporting cast?

It seems on the Wages board you overvalued the role players (or maybe it was Dave Berri) IMO. Does this new finding of yours give you a little more respect for Iverson, or do you still feel the role players were much more important than given credit for? What is your subjective opinion based on these objective stats about that 2001 Philly team? Personally I felt the defense (Ratliff then Dikemebe helped a lot -- along with having excellent defensive role players). But without Iverson's high Usage and high quality play, those role players might have been able to play such a high level defense, nor be as offensively efficient as they were.

Your thoughts?
Back to top
View user's profile Send private message
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina

PostPosted: Wed Aug 02, 2006 7:29 am Post subject: Reply with quote
I still do not find Iverson to be particularly effective - he rates a little above average. His relatively poor shooting does hurt and he is a pretty poor defender. The evidence seems to suggest that with all of the energy he expends on offense, he takes a lot of plays off on defense. So the positive effects he has on reducing the team turnovers rate (he is one of the best in the league at this) are offset by his defense.
Back to top
View user's profile Send private message Send e-mail Visit poster's website Yahoo Messenger
Nikos



Joined: 16 Jan 2005
Posts: 346


PostPosted: Wed Aug 02, 2006 8:43 am Post subject: Reply with quote
So your saying Iverson in 2001 was just a good player, or above average but not by much?

Could it be a coincidence that his defense might be perceived as bad, because he has to play so hard on offense? Doesn't his offense allow those role players to have meaningful roles to play to begin with?

Subjectively, don't you think that he would fit in on a team like the Spurs? Who have excellent defense and good offensive players (Duncan etc...) and could make his numbers and thus his +/- stats more impressive? You mentioned your +/- aren't neccesarily conclusive unless they span over multiple years -- but even then team dynamics and player health/age can change dramatically. Some players could have drastically different +/- yet still produce similiar stats. Doesn't mean they aren't of a similiar calibur player.

I guess I just don't see how the 01 Sixers can win 56 games with a bunch of average role players. What teams in the past have ever done that?
Back to top
View user's profile Send private message
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina

PostPosted: Wed Aug 02, 2006 9:03 am Post subject: Reply with quote
Nikos wrote:
So your saying Iverson in 2001 was just a good player, or above average but not by much?

Could it be a coincidence that his defense might be perceived as bad, because he has to play so hard on offense? Doesn't his offense allow those role players to have meaningful roles to play to begin with?

Subjectively, don't you think that he would fit in on a team like the Spurs? Who have excellent defense and good offensive players (Duncan etc...) and could make his numbers and thus his +/- stats more impressive? You mentioned your +/- aren't neccesarily conclusive unless they span over multiple years -- but even then team dynamics and player health/age can change dramatically. Some players could have drastically different +/- yet still produce similiar stats. Doesn't mean they aren't of a similiar calibur player.

I guess I just don't see how the 01 Sixers can win 56 games with a bunch of average role players. What teams in the past have ever done that?

I suspect that Iverson was a little better in 2000-01; his stats suggest that his adjusted plus/minus ratings (I don't have data that far back) would have been better that year. And on the Wages of Wins blog, I made a case for how he could be an MVP in a situation like what transpired in 2000-01. I am not quite sure that he was that good, but I do believe that he provides a team a fairly unique opportunity to load up on really good role players, which is what Philly did in 2000-01.
Back to top
View user's profile Send private message Send e-mail Visit poster's website Yahoo Messenger
deepak



Joined: 26 Apr 2006
Posts: 664


PostPosted: Wed Aug 02, 2006 10:53 am Post subject: Reply with quote
Dan Rosenbaum wrote:

I suspect that Iverson was a little better in 2000-01; his stats suggest that his adjusted plus/minus ratings (I don't have data that far back) would have been better that year. And on the Wages of Wins blog, I made a case for how he could be an MVP in a situation like what transpired in 2000-01. I am not quite sure that he was that good, but I do believe that he provides a team a fairly unique opportunity to load up on really good role players, which is what Philly did in 2000-01.


Which of his stats suggest that his adj +/- in 2001 would have been better? Is it that he rebounded slightly better and got more steals?

Because, from a PER-perspective, his last season was his career-best. He had a much higher assist-rate, his turnover-rate was about the same, and his TS% was a career high.
Back to top
View user's profile Send private message
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina

PostPosted: Wed Aug 02, 2006 11:02 am Post subject: Reply with quote
deepak_e wrote:
Which of his stats suggest that his adj +/- in 2001 would have been better? Is it that he rebounded slightly better and got more steals?

Because, from a PER-perspective, his last season was his career-best. He had a much higher assist-rate, his turnover-rate was about the same, and his TS% was a career high.

The good role players around him made him more valuable that season. They made the other team less efficient on offense and that allowed Iverson's efficiency relative to the opposition to rise. In essence, what happened was the Philly role players changed the game into where Iverson's relatively inefficient shots were good shots rather than mediocre shots. This is another case where context matters.
Back to top
View user's profile Send private message Send e-mail Visit poster's website Yahoo Messenger
94by50



Joined: 01 Jan 2006
Posts: 499
Location: Phoenix

PostPosted: Wed Aug 02, 2006 11:05 am Post subject: Reply with quote
Dan Rosenbaum wrote:
These effects are highly statistically significant (t-stat=9.12), so this is strong evidence that scorers do, on average, positively affect their teammates. The surprise is that the effect is mostly through teammates commiting fewer turnovers rather than shooting better.

However, if I had to make an argument supporting the idea that great players make their teammates better, these would just about be the first two effects I would look for, as follows:

1) Decrease in turnovers - results from the primary scorer being able to or asked to create his own shot, thus lessening the need for other players to handle the ball.

2) Increase in shooting efficiency from the field - results, obviously, from the defense being concentrated on the primary scorer, giving the role players easier, more open shots.
Back to top
View user's profile Send private message
Ben



Joined: 13 Jan 2005
Posts: 264
Location: Iowa City

PostPosted: Wed Aug 02, 2006 11:12 am Post subject: Reply with quote
Dan Rosenbaum wrote:

This is really interesting. Both Win Score and Wins Produced are both pretty terrible without position adjustments, but with the position adjustment they are not bad. It suggests that a lot of other methods have overvalued shot creation.


Dan, would you mind expanding on this last sentence here?
Back to top
View user's profile Send private message
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina

PostPosted: Wed Aug 02, 2006 11:27 am Post subject: Reply with quote
Ben wrote:
Dan Rosenbaum wrote:

This is really interesting. Both Win Score and Wins Produced are both pretty terrible without position adjustments, but with the position adjustment they are not bad. It suggests that a lot of other methods have overvalued shot creation.

Dan, would you mind expanding on this last sentence here?

Without getting into specifics, I am treading a fine line here in that I am arguing that the value of shot creation is greater than what Wages of Wins assumes (i.e. zero), but less than what is implicitly assumed by metrics such as NBA Efficiency and PER. Hopefully that is helpful.
Back to top
View user's profile Send private message Send e-mail Visit poster's website Yahoo Messenger
Ben



Joined: 13 Jan 2005
Posts: 264
Location: Iowa City

PostPosted: Wed Aug 02, 2006 11:42 am Post subject: Reply with quote
Dan Rosenbaum wrote:
Ben wrote:
Dan Rosenbaum wrote:

This is really interesting. Both Win Score and Wins Produced are both pretty terrible without position adjustments, but with the position adjustment they are not bad. It suggests that a lot of other methods have overvalued shot creation.

Dan, would you mind expanding on this last sentence here?

Without getting into specifics, I am treading a fine line here in that I am arguing that the value of shot creation is greater than what Wages of Wins assumes (i.e. zero), but less than what is implicitly assumed by metrics such as NBA Efficiency and PER. Hopefully that is helpful.


Yes, thanks. I guess you would put it closer to where Orating,Drating, Win Shares does? In between that and PER?

[edit to add questions.]

Last edited by Ben on Wed Aug 02, 2006 11:44 am; edited 1 time in total

Crow
Posts: 6146
Joined: Thu Apr 14, 2011 11:10 pm

Re: Wins Produced - Wages of Wins (Berri, Schmidt, and Brook

Post by Crow » Fri Apr 15, 2011 8:05 pm

page 4

Author Message
deepak



Joined: 26 Apr 2006
Posts: 664


PostPosted: Wed Aug 02, 2006 11:43 am Post subject: Reply with quote
Dan Rosenbaum wrote:

The good role players around him made him more valuable that season. They made the other team less efficient on offense and that allowed Iverson's efficiency relative to the opposition to rise. In essence, what happened was the Philly role players changed the game into where Iverson's relatively inefficient shots were good shots rather than mediocre shots. This is another case where context matters.


Thanks. I suppose you'd disagree with the claim of some that player's don't make eachother better.

But you said Iverson's stats suggested that his adjusted +/- would have been better in 2001 (if I didn't misread what you meant). I'm wondering what you meant by that statement. Is it his stats relative to what the opposing team did?
Back to top
View user's profile Send private message
Kevin Pelton
Site Admin


Joined: 30 Dec 2004
Posts: 978
Location: Seattle

PostPosted: Wed Aug 02, 2006 11:55 am Post subject: Reply with quote
Ben wrote:
Yes, thanks. I guess you would put it closer to where Orating,Drating, Win Shares does? In between that and PER?

Unless I'm mistaken, Dean's Offensive Rating also assumes no value to shot creation. (Which is not to say that Dean sees no value in shot creation, it's just not done in that particular metric.)
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina

PostPosted: Wed Aug 02, 2006 11:56 am Post subject: Reply with quote
deepak_e wrote:
Thanks. I suppose you'd disagree with the claim of some that player's don't make eachother better.

But you said Iverson's stats suggested that his adjusted +/- would have been better in 2001 (if I didn't misread what you meant). I'm wondering what you meant by that statement. Is it his stats relative to what the opposing team did?

The evidence seems to suggest that players who shoot a lot, on average, increase the offensive efficiency of their teammates. But whether I agree or disagree with that statement depends on how someone defines "better." Adding a player who shoots a lot will likely reduce his teammates' scoring.

I probably should have said that the combination of Iverson's stats and his teams' stats suggested that his adjusted plus/minus would have been better in 2000-01.
Back to top
View user's profile Send private message Send e-mail Visit poster's website Yahoo Messenger
Ben



Joined: 13 Jan 2005
Posts: 264
Location: Iowa City

PostPosted: Wed Aug 02, 2006 11:59 am Post subject: Reply with quote
admin wrote:
Ben wrote:
Yes, thanks. I guess you would put it closer to where Orating,Drating, Win Shares does? In between that and PER?

Unless I'm mistaken, Dean's Offensive Rating also assumes no value to shot creation. (Which is not to say that Dean sees no value in shot creation, it's just not done in that particular metric.)


That's what I thought too, but then they did pretty well in the correlations with plus-minus. Is that because they incorporate team defense?
Back to top
View user's profile Send private message
Mark



Joined: 20 Aug 2005
Posts: 807


PostPosted: Wed Aug 02, 2006 1:47 pm Post subject: Reply with quote
Instead of Iverson I looked at Tim Duncan and Spur players with and without him and someone else in. Without some see a small turnover rate increase (+0.1 to 0.3 per 40 minutes) increase but others dont and the FG% changes seem fairly small too (0-3% on FG%) though these things add up and small can still be significant. The type and quality fo role player matters. Some are more self-sufficient, able to create on their own, shoot efficently and more often without the star and San Antonio seems to have put those type players around Duncan and on the court when he is not on. The offensive ratings of the main players with and without Duncan seem to be within 1 point. Duncan's full on/off is only 2.6 pts on team offense. Duncan and Iverson are different positions, different star types, maybe the polar opposites in some ways. It is not just them (and not all one way), it is the nature of the role players too- fit with and fit without the star are both important and maybe this last part was better executed in San Antonio, though Philly doesnt play without Iverson as much (less than half as much in regular season, playoffs probably would be a good deal closer). Team FG% on/off with Iverson +3.4%, Duncan -0.3%. Iverson shooting rate is about 30% higher and Dan's information suggests that would be missed more. Perhaps the quality of replacement varies in these cases too (Ollie and Salmons lower weighted average offensive PER than Horry). Complicated story to sort out.

Last edited by Mark on Wed Aug 02, 2006 2:59 pm; edited 4 times in total
Back to top
View user's profile Send private message
94by50



Joined: 01 Jan 2006
Posts: 499
Location: Phoenix

PostPosted: Wed Aug 02, 2006 2:50 pm Post subject: Reply with quote
admin wrote:
Ben wrote:
Yes, thanks. I guess you would put it closer to where Orating,Drating, Win Shares does? In between that and PER?

Unless I'm mistaken, Dean's Offensive Rating also assumes no value to shot creation.

Not explicitly, no. He suggests adding one point to the ORtg for every 1% increase in possession percentage. Presumably 20% is the base percentage; therefore, a player with a 114 ORtg using 24% of his team's possessions would have a "true" ORtg of 118.
Back to top
View user's profile Send private message
deepak



Joined: 26 Apr 2006
Posts: 664


PostPosted: Wed Aug 02, 2006 2:57 pm Post subject: Reply with quote
94by50 wrote:
admin wrote:
Ben wrote:
Yes, thanks. I guess you would put it closer to where Orating,Drating, Win Shares does? In between that and PER?

Unless I'm mistaken, Dean's Offensive Rating also assumes no value to shot creation.

Not explicitly, no. He suggests adding one point to the ORtg for every 1% increase in possession percentage. Presumably 20% is the base percentage; therefore, a player with a 114 ORtg using 24% of his team's possessions would have a "true" ORtg of 118.


Does he also say to subtract points from ORtg for players below 20%?
Back to top
View user's profile Send private message
Mark



Joined: 20 Aug 2005
Posts: 807


PostPosted: Wed Aug 02, 2006 3:01 pm Post subject: Reply with quote
Dan, thinking about Iverson vs Duncan I wondered was your work on impact of high shooting rate guys paced adjusted? If it is, then a lot of this may not be on point.

But could a lot of Iverson's impact by as a major pace changer (while Duncan is not) instead of making things easier for teammates in halfcourt, with running presumably leading to higher FG%? I see Philly last season 6th fastest pace overall but in 2001 it was 19th. Was the pace change a mistake? Has it been better for Iverson than the team as a whole? It isnt the only factor as their success rate changed before the pace did but the faster pace the last two seasons doesnt seem the answer, so far with these players and coach. Anyone know Philly's pace and offensive efficiency without Iverson over the recent period, any big difference from glory years to now for these?They could be computed from b-r.com data (perhaps later). I see the basic Iverson on/off team scoring differential in 2005-6 is twice what it was in 2002-3 from 82games.

Last edited by Mark on Wed Aug 02, 2006 3:31 pm; edited 4 times in total
Back to top
View user's profile Send private message
94by50



Joined: 01 Jan 2006
Posts: 499
Location: Phoenix

PostPosted: Wed Aug 02, 2006 3:15 pm Post subject: Reply with quote
deepak_e wrote:
94by50 wrote:
admin wrote:
Ben wrote:
Yes, thanks. I guess you would put it closer to where Orating,Drating, Win Shares does? In between that and PER?

Unless I'm mistaken, Dean's Offensive Rating also assumes no value to shot creation.

Not explicitly, no. He suggests adding one point to the ORtg for every 1% increase in possession percentage. Presumably 20% is the base percentage; therefore, a player with a 114 ORtg using 24% of his team's possessions would have a "true" ORtg of 118.


Does he also say to subtract points from ORtg for players below 20%?

Well, yes, but I thought that would be understood.
Back to top
View user's profile Send private message
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina

PostPosted: Wed Aug 02, 2006 4:33 pm Post subject: Reply with quote
Offensive minus defensive rating does better if that possession rate adjustment is added. It is right up their with Win Shares.

Another thing. I can build a prediction model using only minutes played and team wins and come up with individual ratings that predict adjusted plus/minus ratings as well as PER and Wins Produced. And I don't even need to position adjust.

And if I aggregate this up to the team level, I would predict team wins almost perfectly.
Back to top
View user's profile Send private message Send e-mail Visit poster's website Yahoo Messenger
deepak



Joined: 26 Apr 2006
Posts: 664


PostPosted: Wed Aug 02, 2006 4:47 pm Post subject: Reply with quote
Dan Rosenbaum wrote:
Another thing. I can build a prediction model using only minutes played and team wins and come up with individual ratings that predict adjusted plus/minus ratings as well as PER and Wins Produced. And I don't even need to position adjust.


And what would that formula be?
Back to top
View user's profile Send private message
Mark



Joined: 20 Aug 2005
Posts: 807


PostPosted: Wed Aug 02, 2006 5:33 pm Post subject: Reply with quote
Dan would you be willing to score a meta-rating?

Perhaps

30% Statistical Adjusted Plus/Minus Rating +
20% Offensive minus Defensive Rating (position adjustment) +
30% Wins Produced (position adjustment) +
20% Win Shares

that would seem 50% global based and 50% local

or whatever weighting you prefer

( I leave some others out, including roland rating or just counterpart points or PER and protrade in these cases for lack of scores)


Also is there anything notable about the comparative ability of the methods you've scored to accurately grade the top 4 teams in the league by regular season record or appearance in conference finals? Average correlation is a main thing but accuracy with strongest teams is also a valuable feature.

Last edited by Mark on Fri Aug 04, 2006 11:11 pm; edited 4 times in total
Back to top
View user's profile Send private message
ziller



Joined: 30 Jun 2005
Posts: 126
Location: Sac Metro

PostPosted: Wed Aug 02, 2006 5:41 pm Post subject: Reply with quote
Dan Rosenbaum wrote:
Another thing. I can build a prediction model using only minutes played and team wins and come up with individual ratings that predict adjusted plus/minus ratings as well as PER and Wins Produced. And I don't even need to position adjust.

And if I aggregate this up to the team level, I would predict team wins almost perfectly.


So is this a way of saying that PER and WP are really poor predictors of your adjusted plus/minus ratings, or that minutes played and team wins are extremely vital in accurately rating players?

Your statement makes sense - team quality and minutes played would tend to lead to some accuracy in +/-. But I guess I don't understand the usefulness of this.

Also, as a general comment on this entire thread... All we (and by we, I mean Dan) seem to be doing is talking about which of these other rating measures we use or have been introduced to (WP, WS, PER, OEff-DEff, etc.) help predict Dan's ratings. We probably established as a group a year ago that Dan's ratings were pretty darn good, with a very solid methodology and sensible results. But I guess I'm missing the value of judging these correlations instead of judging the success and merits of the ratings systems on their own?

In other words, how do we know Dan's ratings are the best? If we don't know that, why are we judging other systems by how closely they predict Dan's? How else can we judge the systems without buying a basketball team?

On the whole, I agree that Wins Produced is lacking, mostly because it undervalues production in favor of overvaluing efficiency. But I'm not particularly convinced any measure we've seen is close to vastly superior to the others. Because basketball is so nuanced (especially in comparison with baseball), it might take seeing the implementation of certain rating systems in a predictive sense to get tons further on parsing the quality of those systems. (Whereas in baseball, the theory pretty definitively explains why OPS and VORP are important.) I think I fit in with the +/- side of the equation moreso than any others, and I wanted to stand up and clap when Dan started talking about using correlations to judge which stats tell us about other player qualities (that is really important), but there's a lot of noise everywhere.
_________________
SactownRoyalty.com
tziller@gmail.com
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Analyze This



Joined: 17 May 2005
Posts: 364


PostPosted: Wed Aug 02, 2006 5:47 pm Post subject: Reply with quote
Dan Rosenbaum wrote:
Another thing. I can build a prediction model using only minutes played and team wins and come up with individual ratings that predict adjusted plus/minus ratings as well as PER and Wins Produced. And I don't even need to position adjust.


That would be very interesting. How would you do that exactly?
Back to top
View user's profile Send private message
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina

PostPosted: Wed Aug 02, 2006 9:55 pm Post subject: Reply with quote
Ziller, that was a very good post. I don't think any of this relies upon adjusted plus/minus ratings being the gold standard for rating players in all situations. In particular, I have repeatedly noted that adjusted plus/minus ratings are not always the best thing to use for rookies or players who have not played a lot of minutes. Also, for partial season analysis, adjusted plus/minus ratings are probably not that useful.

But I think it is easier to argue that if we could observe a player for tens of thousands of minutes in lots of different contexts, adjusted plus/minus ratings would tend to get things right; they would be "unbiased."

That is all that I am relying upon in this analysis, since by looking at the correlations of nearly 1200 player-seasons, that is in essence what I am doing. So I think this is a valid way to analyze various metrics - although the fact that my statistical rating comes out the best is probably not good evidence that it is better than the others. Unlike the other measures, it is designed to predict adjusted plus/minus ratings so it almost has to do better.

The rating based upon wins and minutes was nothing that I was advocating; it was more a way of putting this conversation in perspective.

And finally I apologize to those posts I have not answered, including Mark's request for a meta-rating.

Crow
Posts: 6146
Joined: Thu Apr 14, 2011 11:10 pm

Re: Wins Produced - Wages of Wins (Berri, Schmidt, and Brook

Post by Crow » Fri Apr 15, 2011 8:06 pm

Page 5

Author Message Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina
Posted: Thu Aug 03, 2006 9:10 am Post subject:

Bringing this back to the original topic of Wages of Wins, I just don't think we are ever going to see the authors participate in this forum. I guess this is their argument. http://dberri.wordpress.com/2006/07/31/ ... -of-blogs/ Quote:
Academic journals also serve as gate keepers. The journals at the top of the food chain reject most of the papers that are submitted. The material that is accepted is generally the best work in the field. The difficulty of the process and the demand placed on the writer forces one to bring your best game. Furthermore, the peer review exposes your work to the expertise of others and also, and this is perhaps most important, prevents one from making statements and claims that are not supported by the evidence. Those who try and avoid this process – and the people who do this know who they are – invariably produce work of lower quality, making mistakes reviewers would easily spot. In the blogosphere, though, these mistakes can pass unnoticed by an audience that does not understand the methods the person is employing or the basics of economic theory.
If this post is referring to me, I plead guilty as charged. I am a veteran of the peer review process (mostly through serving as a reviewer at many of the journals at the "top of the food chain"), but not with my sports economics research. I am NOT a respected sports economist. (My other research has gotten a fair bit of attention in the peer review process - see Google Scholar.) I don't really disagree with what Dave says here, but I think we could also say. Quote:
Those who try and avoid this process – and the people who do this know who they are – invariably produce work of lower quality, making mistakes those with lots of institutional detail would easily spot. In academia, though, these mistakes can pass unnoticed by an audience that does not understand the details of the institutions the academic is studying.
Well, even if the authors of Wages of Wins do not participate in this forum, the book has prompted an interesting discussion. And that is a good thing. And, as I have said many times before, I highly recommend that folks in this community read this book.
Back to top

ziller



Joined: 30 Jun 2005
Posts: 126
Location: Sac Metro
Posted: Thu Aug 03, 2006 11:01 am Post subject:

Dan, Thanks for taking some time and thought in responding to my concerns. Like I said, I agree that your measure is the best I've seen. The problems with it? You've talked about the limitations as far as the complexity of calculating or reproducing it (not something even Joe Apbrmetrician can do; strictly for the very experienced statistical analysts out there) and the amount of data that must be used to get good, fairly clean readings (multiple seasons, which negates in-season use of it). In this way, your measure - as solid, well-balanced, and intuitive as it is - is inaccessible. So I definitely see why you're discussing measures that correlate well or marginally with it - if some metric that were easy to formulate gave us good indications of a player's relative adjusted +/-, then we would definitely give that metric a hand up. But it appears no current metrics correlate strongly with your ratings, with the strongest correlation being about .48. Apbrmetrics is so difficult because the balance between theory and results. PER has great results - the best players in our eyes are the best players in PER - but a couple of theoretical problems (defense, clearly). Wins Produced looks nice on paper, but doesn't give us results even partially resembling reality. In a month or two, when Wages is fully canonized in sports analysis circles as "Moneyball for hoops" (which it's not, and probably not trying to be), then PER and WP will be the measures at the forefront of our "movement." They're flawed and they end in terribly different results. That's a problem in getting our work to catch on. (I use 'our' very loosely here.) But you are absolutely correct that stuff like Wages is highly important because it gets us thinking and it gets us talking and it gets us doing. I mean, hell, if nothing else, Wages brought Dan Rosenbaum back to the APBR board! That's a victory itself. Myself, though: I've almost given up on collection-based rating systems. I'm almost fine with offensive PER and I'm waiting for better defensive stats to address a "defensive PER." Because roles are so integral to building basketball teams, I'd rather focus on stuff we can measure now and adapting that into describing players in a statistical sense, much like Ed's fantastic work with clusters. To me, when talking about building teams or picking 20-year-olds in a draft, it's more valuable to say Player X shoots 60% from inside, gets 6% of offensive rebound opportunities and blocks 4% of all inside opponent shot attempts than saying Player X is the third best center available. As we've seen in Team USA's 2004 and 2006 effort, as well as the Lakers' 2004 season, having the best players doesn't mean having the best team. (And yes, all of this is making my inner struggle between fit-a-need Kenny Thomas and way-more-talented Shareef Abdur-Rahim rage.)_________________SactownRoyalty.com tziller@gmail.com
Back to top

ziller



Joined: 30 Jun 2005
Posts: 126
Location: Sac Metro
Posted: Thu Aug 03, 2006 11:09 am Post subject:

Oh, I forgot to respond to Berri's discussion on academic journals vs. blogs. Nothing will ever get done if all serious discussion on quantative analysis in sports is limited to academia. Yes, the quality will be consistently better in journals than on the internet, but that's true of every field. Yet, people seem to find the internet valuable. I've learned more from KevinP's columns, Roland's data, Ed's charts, Dan's web-published papers, and Hollinger and DeanO's books than I probably ever will from an academic journal on the subject. The wealth of intelligent quantitative analysis on this board and at related sites is astounding, and because of the people involved and the self-vetting we all freely undertake with each others' work, I am more than comfortable with the quality._________________SactownRoyalty.com tziller@gmail.com
Back to top

Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina
Posted: Thu Aug 03, 2006 11:18 am Post subject:

One small point I should add to Ziller's post is that we should think about the adjusted plus/minus rating like this. Adjusted plus/minus rating = BS + OC + e, where BS is what can be measured by box score stats, OC measures the other contributions made by a player not measured (or correlated with) box score stats, and e is random variation (i.e. error) The reason we can't get high correlations with adjusted plus/minus ratings is because that random error is pretty big. And there is no way to construct a measure that correlates with it. And since the true impact of a player - and least what is likely to translate from season to season - is measured by BS + OC, we don't want to try to correlate with e. So the point here is that the low correlations are not really a big concern. In this context a correlation of 0.5 might be pretty good.
Back to top

Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina
Posted: Thu Aug 03, 2006 11:28 am Post subject:

Ziller's last post made me want to say that the biggest travesty in Wages of Wins is that KevinP's writings are not mentioned; I am not sure it is possible to be an "expert" in this area without reading KevinP's stuff.
Back to top

Mark



Joined: 20 Aug 2005
Posts: 807
Posted: Thu Aug 03, 2006 11:34 am Post subject:

Sports economics research is probably still in the second quarter of the game so there is time for publication / more academic recognition. Perhaps recent publication expands demand, deepens search, increases journal placements. Someone will write another book that summarizes and builds upon more of the work known and used in this community. The index of that book can or should have many mentions of names and ideas "published" here and at other internet and news sites. I am willing to read other academic journals for new insights but the sum of mini-studies and on going studies, technical and philosophical discussions here and at the previous site is impressive and folks can benefit from it or be unaware at their own loss. I assume we will someday see a book from Kevin. Perhaps one from Dan or another one from Dean if confidentiality agreements don't completely block. Or perhaps from another tapped into this pool of thinkers. As much as Wages of Win may be the current focus, there have been other stages / other books and there will be more.Last edited by Mark on Thu Aug 17, 2006 12:23 am; edited 2 times in total
Back to top

Ben



Joined: 13 Jan 2005
Posts: 264
Location: Iowa City
Posted: Thu Aug 03, 2006 11:46 am Post subject:

Dan, have you considered writing up some of your critiques for a journal?
Back to top

Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina
Posted: Thu Aug 03, 2006 11:56 am Post subject:

Ben wrote:
Dan, have you considered writing up some of your critiques for a journal?
Yes, that is part of the reason why I am presenting this work here. It is very useful feedback.
Back to top

ziller



Joined: 30 Jun 2005
Posts: 126
Location: Sac Metro
Posted: Thu Aug 03, 2006 12:04 pm Post subject:

I would buy a Kevin Pelton book. Hell, I've already purchased two Dean Oliver books, and he's only published one! (And since I'm getting ready to move again, I'll invariably lose my copy of BoP again, which will necessitate a third. Damn you, Dean Oliver.)_________________SactownRoyalty.com tziller@gmail.com
Back to top

HoopStudies



Joined: 30 Dec 2004
Posts: 705
Location: Near Philadelphia, PA
Posted: Thu Aug 03, 2006 12:23 pm Post subject:

Dan Rosenbaum wrote:
Bringing this back to the original topic of Wages of Wins, I just don't think we are ever going to see the authors participate in this forum. I guess this is their argument. http://dberri.wordpress.com/2006/07/31/ ... -of-blogs/ Quote:
Academic journals also serve as gate keepers. The journals at the top of the food chain reject most of the papers that are submitted. The material that is accepted is generally the best work in the field. The difficulty of the process and the demand placed on the writer forces one to bring your best game. Furthermore, the peer review exposes your work to the expertise of others and also, and this is perhaps most important, prevents one from making statements and claims that are not supported by the evidence. Those who try and avoid this process – and the people who do this know who they are – invariably produce work of lower quality, making mistakes reviewers would easily spot. In the blogosphere, though, these mistakes can pass unnoticed by an audience that does not understand the methods the person is employing or the basics of economic theory.

For most of us, the peer review process is, unfortunately, a negative thing to get involved in. I fully believe that academia often does the best work in a field and that peer review is important to doing so. However, when I wrote my book and in the 7 years before then when I did the Journal of Basketball Studies, there was really no academic track at all relating to what I was doing. I couldn't generate what I did without just throwing it up on the web and getting a non-peer but fully public review. I did get some people who thought I was crazy, but in the thousands of comments I got, it was less than 1% who didn't like what I did. In an academically fleshed out subject, that wouldn't replace peer review, but quantitative sports analysis hasn't been that until recently. (Further, note that Bill James and Pete Palmer did work that is now accepted by academics in sports econ. I'd like to think that what we started in basketball without academic peer review will stand the test of time and ultimately academia.) But the reason peer-review is now something negative for many of us is that it is positive only for academics, who get promoted and security from publishing. For no one else do such rewards get lavished, with the possible exception of the out-of-the-blue non-academic who both does good work and happens to stumble on JQAS before they stumble onto APBRmetrics. For all else, it is a cost of time and potential job secrets. That being said, some pecking order in terms of quality of work is necessary and actually does get established to some degree here in this forum. Dan's work was first posted here and he has an NBA job. Ed K's work was on display almost exclusively through this forum. He will get a job in the NBA at some point because of it. As long as there is good quality research -- on top of quality discussion -- getting posted, people with influence will read this site and open doors. So, in that way, this site should also serve as a gate-keeper -- not to the world of academia, but hopefully to the world of the NBA._________________Dean Oliver Author, Basketball on Paper The postings are my own & don't necess represent positions, strategies or opinions of employers.
Back to top

HoopStudies



Joined: 30 Dec 2004
Posts: 705
Location: Near Philadelphia, PA
Posted: Thu Aug 03, 2006 12:32 pm Post subject:

ziller wrote:
I would buy a Kevin Pelton book. Hell, I've already purchased two Dean Oliver books, and he's only published one! (And since I'm getting ready to move again, I'll invariably lose my copy of BoP again, which will necessitate a third. Damn you, Dean Oliver.)
One of the silly mistakes I made in BoP was not explicitly acknowledging Kevin Pelton in the Acknowledgements. I have hopefully since made up for that by both giving him a signed book and constantly pointing out to him and others that he is the best writer/spokesman for what we are doing. So, yes, hopefully there will be Kevin Pelton book in the future for you to lose and buy over and over again. And then you can curse him._________________Dean Oliver Author, Basketball on Paper The postings are my own & don't necess represent positions, strategies or opinions of employers.
Back to top

Mark



Joined: 20 Aug 2005
Posts: 807
Posted: Thu Aug 03, 2006 6:08 pm Post subject:

New research is vital as is quality discussion. Discussion often draws more out of the research than originally envisioned and delivered. A quick check of the memberlist post count at the current site shows almost 25 members with 100 posts and another 65 with at least 10 and another 150 with at least 1 and 200 more registered members and an unknown numbers of guests. And years of activity at the prior yahoo site. Strong individual efforts aided in ways by feedback. Thanks also to Kevin for hosting and adminstration.Last edited by Mark on Thu Aug 17, 2006 2:31 am; edited 6 times in total
Back to top

Kevin Pelton
Site Admin


Joined: 30 Dec 2004
Posts: 978
Location: Seattle
Posted: Thu Aug 03, 2006 6:16 pm Post subject:

HoopStudies wrote:
I have hopefully since made up for that by both giving him a signed book and constantly pointing out to him and others that he is the best writer/spokesman for what we are doing.
That's nice and all, but what really makes up for it is you buying me dinner on several occasions.
Back to top

deepak



Joined: 26 Apr 2006
Posts: 664
Posted: Fri Aug 04, 2006 11:45 am Post subject:

Dan, I wonder if the random component in your Adjusted +/- Ratings are throwing off some of these comparisons. You found that raw Win Shares correlated better than Win Shares per minutes, which was unexpected. You also were able to come up with a model using only TmWn% an MP that correlated better than PER. It seems like you're assuming that all the ratings your evaluating would be uniformly impacted by the "randomness" in your Adj +/-. But if that's not the case, then it would be beneficial to try to first reduce that randomness. To that end, perhaps it's worthwhile to limit the players used in the comparison to those who's Adjusted +/- rating has a relatively low error (i.e. very confident). I wonder if you'd get similar results.
Back to top

Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina
Posted: Fri Aug 04, 2006 11:58 am Post subject:

That randomness is stuff like a player's teammates making a lot of lucky shots while he is in the game or rebounds that magically find the one offensive guy near the boards, or the should-be turnover that bounces back to the offensive team. All of these things can make how the scoreboard moves in a short stretch of a game not very reflective of how well the players are playing. This stuff is pretty important when estimating adjusted plus/minus ratings and causes it not to be so good for low minutes players, but this randomness - at least the way I am trying to define it above - should not be correlated with box score stats or the true impact of a player. To the extent that players put themselves in a position to get these lucky breaks, that should be picked up by BC or OC. It is unlikely that this randomness is more correlated with one metric than another. If I limit the correlations to high minutes players, the correlations go up by about the same amount for each of the metrics.

Crow
Posts: 6146
Joined: Thu Apr 14, 2011 11:10 pm

Re: Wins Produced - Wages of Wins (Berri, Schmidt, and Brook

Post by Crow » Fri Apr 15, 2011 8:14 pm

pages 6-10 missing

Page 11

Author Message basketballvalue



Joined: 07 Mar 2006
Posts: 208
Posted: Wed Nov 15, 2006 12:37 am Post subject:

Dan Rosenbaum wrote:
And yes, WINVAL and my adjusted plus/minus ratings do differ (I think Aaron would tell you that his metric is not yet a full adjusted plus/minus measure), but that is only because we are describing different things, i.e. we differ in how much we value clutch/non-clutch time and/or we differ in how we want to handle changes over time in player value. .
Wow, this is a nice long, detailed thread. I'm having trouble keeping up with the flurry of posts. Dan, you definitely characterized correctly that I'm still working on having a coherent, fully adjusted plus/minus rating. I'm hoping to have the same data I have available for 2005-2006 up for 2006-2007 ASAP, then I'll start to release some of the building blocks of a "back-to-basics" adjusted +/- that doesn't use regression. I think there is not enough data generated in a single season to make the adjustments as I had originally planned, but I still think there are adjustments that can be made. As far as some of the philosophy in this thread, I agree with what I think is the general theme that even if adjusted +/- correctly valued everyone's contribution on a unit in one year(say, 2005-2006), it would still be unlikely to predict with 100% accuracy what would happen if you randomly chose 5 players to make up a unit the next year(say, this season). There's still intangibles of chemistry that can't be predicted. Thanks, Aaron
Back to top

Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina
Posted: Fri Nov 17, 2006 2:35 pm Post subject:

Dave Berri has a very nice discussion of PER, using a "model" to construct player ratings, and the use (or lack thereof) of statistical analysis in NBA decision-making. http://dberri.wordpress.com/2006/11/17/ ... cy-rating/ Quote:
Looking at the specific weights Hollinger chooses we see another problem. In discussing the NBA Efficiency metric – which the NBA presents at its website – I argued that this measure fails to penalize inefficient shooting. The regression of wins on offensive and defensive efficiency reveals that shooting efficiency impacts outcomes in basketball. The ball does indeed have to go through the hoop for a team to be successful. The same critique offered for NBA Efficiency also applies to Hollinger’s PERs, except the problem is even worse. Hollinger argues that each two point field goal made is worth about 1.65 points. A three point field goal made is worth 2.65 points. A missed field goal, though, costs a team 0.72 points. Given these values, with a bit of math we can show that a player will break even on his two point field goal attempts if he hits on 30.4% of these shots. On three pointers the break-even point is 21.4%. If a player exceeds these thresholds, and virtually every NBA played does so with respect to two-point shots, the more he shoots the higher his value in PERs. So a player can be an inefficient scorer and simply inflate his value by taking a large number of shots.
Quote:
Less than 15% of wins in the NBA are explained by payroll. Regressions are nice, but not always understood by everyone. So to further illustrate the lack of association between pay and wins I took another approach. Specifically I ranked the teams in the NBA last year in terms of payroll and then divided this ranking into five equal segments. The results revealed that the teams in the top 20% spent an average of about $78 million on players and won –on average – 35.7 games. The next 20% spent $61 million and won 42.5 games. In the middle we see teams that spent only $54 million and won 39.7 games. When we look at the last two groupings – the teams that spent the least – we see clearly the very weak link between pay and wins in basketball. The 20% of teams ranked just below the middle in payroll won 47.7 games while spending $47 million on players. And the teams at the very bottom of the payroll rankings spent less than $38 million on its players and won 39.5 games. Yes, the teams at the bottom spent less than half what the teams spent at the top and actually won more games. Okay, pay and wins do not have a strong link. What does this tell us about player evaluation? In football payroll explains less than 5% of wins. But in football we also see very little consistency in player performance. So decision-makers cannot easily know how to spend money to ensure success in the future. A similar problem – though to a lesser extent – exists in baseball. In basketball, though, players are much more consistent across time. The correlation between a player’s per-minute Win Score this season and last season is 0.84. As we detail in The Wages of Wins, the consistency we observe in basketball exceeds what we observe in either baseball or football. Despite this consistency, though, payroll is still not strongly linked to wins. In sum, decision-makers have a greater ability to predict the future in the NBA, yet the payroll-wins relationship still remains very weak. When we look at what determines salary we see the problem. The primary player characteristic that dictates wages in the NBA is scoring. Shooting efficiency, rebounds, turnovers, and steals – factors that all impact outcomes – are not strongly linked to player pay. Given this evidence, we think players are evaluated incorrectly in the NBA. Too much emphasis is placed on scoring, and not enough on all the other factors that impact outcomes.
I would like to reiteratre that despite my many criticisms of Wages of Wins, there are a lot of good ideas in the book and at the blog.
Back to top

kjb



Joined: 03 Jan 2005
Posts: 865
Location: Washington, DC
Posted: Fri Nov 17, 2006 3:12 pm Post subject:

Berri & company have raised an interesting point, and I'm kinda torn on it. Yes, shooting efficiency is extremely important -- there's no question. But, someone has to take the shots. Does treating all shot attempts as negative (which I think their metric does) reflect reality, though? Like I say, I'm torn on the subject. I know that shooting efficiency is valuable, but at the same time, does it really hurt a team when Player A misses and his team gets the offensive rebound? Dan, didn't you have a number showing a very low shooting percentage as a sort of benchmark for positive adjusted +/-?
Back to top

Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina
Posted: Fri Nov 17, 2006 3:37 pm Post subject:

The problem is that Wins Produced doesn't really use regression to answer this question. It assumes that the average points per possession throughout the league tells us all we need to know about the breakeven shooting percentage for the marginal shot. It assumes players convert shots on the margin at the same rate as their average shot. And so if we remove those players who shoot a lot (and draw a lot of the focus of the defense), the shooting percentages of remaining players will not change. Wins Produced is very similar to Bob Chaikin's simulator along those lines. Coming from a sports economics literature dominated by baseball, it is very natural to make this assumption, since this issue does not come up in a sport where every player gets his turn at bat. When I look at this issue empirically using adjusted plus/minus statistics, I tend to find that the breakeven shooting percentage is somewhere between what PER/NBA Efficiency uses and what Wins Produced uses. Wins Produced rightly addresses a problem with PER/NBA Efficiency, but in my opinion it so overcorrects that the resulting metric does a poor job explaining how players impact winning.Last edited by Dan Rosenbaum on Fri Nov 17, 2006 5:09 pm; edited 2 times in total
Back to top

Neil Paine



Joined: 13 Oct 2005
Posts: 774
Location: Atlanta, GA
Posted: Fri Nov 17, 2006 4:06 pm Post subject:

I found this to be one of Berri's most well-reasoned posts at the WoW Journal. I've said it before -- PER doesn't have the "power of language", because it lacks units. You can say that it's kind of like Points Produced per 40 minutes, but it really isn't, because it mixes apples and oranges (defensive and offensive stats) right and left, the weights are sometimes totally subjective (Why is an assist worth .67 "points"? And frankly, saying that, "on an assist, the passer does one thing and the scorer does two things, so the passer must get 1/3 credit," just doesn't cut it), the league average is forced to be 15, etc. So, yeah, Berri has some legit gripes about NBAEff and PER, because Wins Produced is at least derived pseudo-scientifically. If he's lurking, I wonder how John would respond to Berri's post (probably wishful thinking on my part, but still)...
Back to top

Ben



Joined: 13 Jan 2005
Posts: 266
Location: Iowa City
Posted: Fri Nov 17, 2006 5:06 pm Post subject:

Dan Rosenbaum wrote:
The problem is that Wins Produced doesn't really use regression to answer this question. It assumes that the average points per possession throughout the league tells us all we need to know about the breakeven shooting percentage for the marginal shot. It assumes that a player converts shots on the margin at the same rate as their average shot. And so if we remove those players who shoot a lot (and draw a lot of the focus of the defense), the shooting percentages of remaining players will not change. Wins Produced is very similar to Bob Chaikin's simulator along those lines. Coming from a sports economics literature dominated by baseball, it is very natural to make this assumption, since this issue does not come up in a sport where every player gets his turn at bat.
I'm not an economist, but I've studied economics a little bit and it seems to me that they would be the most likely to see this distinction (between marginal and average shot). In fact, I'm surprised that an economist could miss it.
Back to top

bchaikin



Joined: 27 Jan 2005
Posts: 689
Location: cleveland, ohio
Posted: Fri Nov 17, 2006 6:21 pm Post subject:

It assumes players convert shots on the margin at the same rate as their average shot. And so if we remove those players who shoot a lot (and draw a lot of the focus of the defense), the shooting percentages of remaining players will not change. Wins Produced is very similar to Bob Chaikin's simulator along those lines......it is very natural to make this assumption... might you care to expound on what this assumption is that the simulator uses?...
Back to top

asimpkins



Joined: 30 Apr 2006
Posts: 245
Location: Pleasanton, CA
Posted: Sat Nov 18, 2006 2:31 pm Post subject:

davis21wylie2121 wrote:
(Why is an assist worth .67 "points"? And frankly, saying that, "on an assist, the passer does one thing and the scorer does two things, so the passer must get 1/3 credit," just doesn't cut it)
What is the appropriate way to weight assists? It's never satisfying to just pick a number, but what other choice do you have? How do other rating systems solve it?
Back to top

deepak



Joined: 26 Apr 2006
Posts: 665
Posted: Sat Nov 18, 2006 5:03 pm Post subject:

Another thing that bothered me about crediting assists: Does it really make sense for the passer and shooter to share credit on an assisted score, but for the shooter to be exclusively penalized on the miss?
Back to top

Neil Paine



Joined: 13 Oct 2005
Posts: 774
Location: Atlanta, GA
Posted: Sat Nov 18, 2006 8:36 pm Post subject:

Basketball is troublesome in the way that it fails to lend itself to regression. In baseball, if I wanted to find the various weights (in runs) for all of the offensive events that lead to run scoring, I would simply regress team singles, doubles, walks, home runs, etc. against team runs scored, and come out with something like this. If I regressed the relevant offensive stats in basketball against points, however, I would always come out with this formula: Pts = (2*FG) + 3FG + FT. In other words, assigning partial "points" credit to assists and offensive rebounds is simply not intuitive, and pretty much any method that does this is going to have to fudge on the "value" of an assist. But if you regress on wins, you don't have the problem of points being totally dependent on a few of the variables that you want to regress. I presume that this is why Berri based his research on "wins produced" and not "points produced" -- linear regression is feasible. asimpkins wrote:
davis21wylie2121 wrote:
(Why is an assist worth .67 "points"? And frankly, saying that, "on an assist, the passer does one thing and the scorer does two things, so the passer must get 1/3 credit," just doesn't cut it)
What is the appropriate way to weight assists? It's never satisfying to just pick a number, but what other choice do you have? How do other rating systems solve it?
That's just it -- they don't do it any better than JH. Whether it be Bellotti's "Points Created", Tendex, NBA Efficiency, etc., they all arbitrarily pick a weight for assists and go with it. At least Hollinger subtracts out the assistant's credit from the assistee's field goals -- I would venture to say that most of these linear weights methods do not. However, while I agree with many of its results (and I used to be a huge PER devotee), I have grown wary of a lot of the logic behind PER. The formula is as arbitrary as that used by the much-maligned IBM Award, but Hollinger justifies his assertion that the PER is superior by saying that it confirms his pre-existing conceptions of how players should rank -- after all, those other methods must be hogwash, because they don't have Shaq as the number one center, or they have David Robinson ranking ahead of Michael Jordan (I've got news for you, John, PER has been known to do the same thing). It's called confirmation bias, and Berri's right -- that's not how science works. I believe DeanO's Points Produced uses probability theory, which is a huge step in the right direction. Similarly, Wins Produced, while flawed, is at least an empirical approach (until they employ their fudge factor at the end of the calculation, that is). We all laughed when Berri came out with Rodman as the 1998 NBA MVP, but the important thing is that he's trying to derive results scientifically. I don't like the way he's cut himself off from this group, or the way he shuns all criticism, or the way he treats Wins Produced as the be-all and end-all of stats. That attitude can only be bad for APBRmetrics. But I welcome the attempt to do better than simply "picking a number".
Back to top

Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina
Posted: Sat Nov 18, 2006 9:53 pm Post subject:

First point, Wins Produced is designed to predict net team efficiency (offensive minus defensive efficiency) and if only one year was used in the prediction, it would perfectly predict net team efficiency. Their results tell us nothing more than that net team efficiency does a pretty good job predicting wins. Second, Berri often argues that he uses regression to come up with the relative value of points, rebounds, field goal attempts, etc., but in general that is not true. For points, field goal and free throw attempts, rebounds, steals, and turnovers, Berri simply uses the leaguewide average of points per possession to arrive at the relative value of the linear weights. I guess that is emprical, but since it is so far removed from anything at the individual level, I would argue that it is only marginally more emprical than PER or other linear weights methods. Berri runs a regression, but he did not need to, so I think it is misleading to argue that this is really regression-based. For blocks and personal fouls, he does use empirical work, but not regression, to arrive at the appropriate weights. Only for assists does he really use regression analysis. At the end of the day, what we really want to know is how much better does a team, on average, do with a guy who is more of an assister or a scorer or a rebounder. And it only the value of the assister that Berri really uses emprical work to ask the appropriate question. So I guess in sum I would argue that Wins Produced is much closer to just "picking a number" than it really appears.
Back to top

Kevin Pelton
Site Admin


Joined: 30 Dec 2004
Posts: 979
Location: Seattle
Posted: Sun Nov 19, 2006 4:40 am Post subject:

Maybe this is ultimately a trivial point, but I strongly disagree with Berri's disdain for "the laugh test". Confirmation bias is selectively looking for evidence to fit the test hypothesis, but conventional wisdom is not the test hypothesis, it is (or at least should be, in my mind) the null hypothesis. The null hypothesis is supposed to get the benefit of the doubt. Surprising results shouldn't cause us to reject the method behind them, but they should cause us to scrutinize it and understand why they occur. In the case of Berri's previous effort, the rating for Dennis Rodman revealed some serious flaws - those flaws, in this case, being tied to the effort to determine the value of individual statistics through regression at the team level. If we literally just laughed, that would not have been useful, but neither would have been accepting the results because they were "scientific." There's also a pragmatic argument. For the most part, we aren't dealing with scientists, so the scientific method is not always appropriate. There is no question that fans and front office personnel are using the laugh test. Should we just ignore that?
Back to top

Mike G



Joined: 14 Jan 2005
Posts: 3605
Location: Hendersonville, NC
Posted: Sun Nov 19, 2006 9:09 am Post subject:

davis21wylie2121 wrote:
...they all arbitrarily pick a weight for assists and go with it. At least Hollinger subtracts out the assistant's credit from the assistee's field goals -- I would venture to say that most of these linear weights methods do not. .
Not all points are equally 'assisted'. 82games tracks assisted% of FG for players, though (as far as I know) this info isn't available before season's end. This makes it possible to scale down scoring by use of the unassisted% of players' points. The remainder of credit can then go to the assist man. This year, I'm experimenting with estimating player unAst%, and also with scaling actual assist rates to some power of (TmPPG/OppPPG). Since I'm correlating player eWins to team pythagorean-expected (pW), I get an exponent that boosts assist-men's eW appropriately when they assist for winning teams. Thus far, it seems by multiplying the assist rate by (TmPPG/OppPPG)^4 , I get a best fit. This is a pretty large exponent, and it exaggerates the difference between good assists for good teams, and for bad teams. Set the exponent to zero (no good-team credit), and Utah's Deron Williams' per-36 assist rate is 7.5, while Indiana's Jamaal Tinsley is at 7.7. With the ^4 distinction, the Utah PG soars to an 'effective' rate of 10.3, while Tinsley sags to 6.6 . Of course, this early in the season, PPG-diff is hardly the last word on 'good/bad' teams. So it's an ongoing experiment. BobC has referred to 'assists that lead to wins' (or not), which he determines thru his sim. I'm proposing there may be other ways of guessing at a 'sliding scale' assist weight._________________` 36% of all statistics are wrong
Back to top

tomverve



Joined: 24 Feb 2005
Posts: 17
Posted: Sun Nov 19, 2006 1:55 pm Post subject:

admin wrote:
Maybe this is ultimately a trivial point, but I strongly disagree with Berri's disdain for "the laugh test". Confirmation bias is selectively looking for evidence to fit the test hypothesis, but conventional wisdom is not the test hypothesis, it is (or at least should be, in my mind) the null hypothesis. The null hypothesis is supposed to get the benefit of the doubt.
The idea behind the laugh test is that if the results of some analysis are strongly out-of-whack with one's pre-existing notions of what the results should look like, then there must be something wrong with the analysis and it should be restructured until its results ultimately do achieve some harmony with expectations. If that isn't confirmation bias I don't know what is. Still, I'd agree that the laugh test is to some extent a useful methodological tool, for these reasons. 1: Presumably, the threshold for confirmation bias endorsed by the laugh test is rather high-- the results of an analysis have to be so far from expectation as to induce (or at least suggest the image of) laughter. So we only enact the "bias" when our two sources of judgment-- statistical analyses and direct observation and social discourse-- are heavily dissonant. This ties into 2: We should regard the "common sense" set of judgments, evaluations and expectations of the general basketball-watching community to have some degree of value, even if they are not infallible. This is not in the least part due to the fact that these judgments take into account observations that do not factor into any given statistical analysis (i.e. there is useful information to be obtained from watching games that is opaque to the numbers extracted for a given kind of analysis, and of course vice versa). Ultimately the laugh test could be seen as enforcing a kind of consistency of judgment across two disparate but useful sources of information and evaluation of NBA games and players. Conventional wisdom about basketball can serve as a useful counterbalance to statistical analysis, and arguably such a counterbalance is needed since analysis of basketball is so difficult and murky of a prospect for all the well-known reasons. To say the laugh test has no value is essentially to say that judgments formed from watching NBA games have no value (as opposed to, say, limited value), which seems a rather bold claim.
Back to top

Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina
Posted: Sun Nov 19, 2006 3:59 pm Post subject:

Berri and co-authors in their previous work came up with a metric that did not include assists, blocks, or fouls. None of those statistics are directly tied to points scored or possessions, so it is possible (with a team adjustment) to come up with a metric that perfectly predicts team efficiency even without using assists, blocks, or fouls. So why did Wages of Wins move away from that metric? It passed through the peer-review process, and if they had used a team adjustment (I am not sure they did), it would have predicted team wins just as well as the current Wins Produced. I think they moved from that metric through lots and lots of e-mails from an APBRmetric member (not me) that convinced them that this older metric did not pass the laugh test. That old metric has almost no correlation with adjusted plus/minus statistics (not that they knew this or would have cared). It led to perverse results in several published papers. So how can the scientific method and peer-review process fail so badly? Models can be wildly off the mark and so if they are only judged on some form of internal consistency, they can produce perverse results. Calibrating a model to perfectly predict subjective evaluations may not be scientific, but completely ignoring (and belittling like they do in the book) all non-statistical analysis is NOT how good empirical work is done in other areas of economics. Good empirical work is always a combination of theory and evidence that incorporates as much subjective knowledge as is practically possible. We may risk being non-scientific if we overfit the data to predict our subjective evaluations. But we also run the risk of being irrelevant if we completely ignore subjective evaluations.


Page 12

Author Message asimpkins



Joined: 30 Apr 2006
Posts: 245
Location: Pleasanton, CA
Posted: Mon Nov 20, 2006 11:54 am Post subject:

davis21wylie2121 wrote:
but Hollinger justifies his assertion that the PER is superior by saying that it confirms his pre-existing conceptions of how players should rank -- after all, those other methods must be hogwash, because they don't have Shaq as the number one center, or they have David Robinson ranking ahead of Michael Jordan (I've got news for you, John, PER has been known to do the same thing). It's called confirmation bias, and Berri's right -- that's not how science works.
I see what you're saying. Two comments though: I never really think of PER as a scientific attempt to rate players. I think you're right that it doesn't compare to some of these other projects. For me, it's always just been a summary tool. It gathers all these box score numbers from all over the place -- more than I could ever keep inside my head -- and gives me a single number summary of what a player has achieved statistically. And for that, it's extremely useful. Going from statistical dominance to true value is another jump, but that doesn't mean that there is no merit to trying to map out the former. Secondly, I don't think the primary appeal of PER is that it confirms pre-existing conceptions. You yourself pointed out how often it doesn't. I think its strength is that all the weights for each stat seem very well crafted and thought out. Assists, of course, stand out as being particularly arbitrary, but there's not much that can be done there within the scope of this project. The 2/3 - 1/3 split seems like a reasonable compromise.
Back to top

Eli W



Joined: 01 Feb 2005
Posts: 402
Posted: Mon Nov 20, 2006 4:01 pm Post subject:

A lot more from the Wages of Wins blog: http://dberri.wordpress.com/2006/11/19/ ... on-a-rock/
Back to top

94by50



Joined: 01 Jan 2006
Posts: 499
Location: Phoenix
Posted: Mon Nov 20, 2006 5:00 pm Post subject:

From the post linked above: "If you take ten shots at a below average rate you hurt your team. If you take twenty shots at a below average rate you hurt your team even more." 1) Is this statement true? 2) If it is true, how do we know that? If it is not true, how do we know that? 3) Is this universally true, or is it context-dependent? I still have trouble believing that efficiency should be judged independently, as opposed to being judged in the context of shooting frequency. I also have been trained to think that using the average as the baseline of acceptable performance is wrong. But if there's no measurable correlation between efficiency and shot volume... what's correct?
Back to top

Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina
Posted: Mon Nov 20, 2006 5:20 pm Post subject:

By the way, somebody who at least lurks here at APBRmetrics is posting comments at the Wages of Wins blog relating some of the arguments of mine in this thread and others. Please be a little careful in those posts; in a couple cases they have overstated arguments that I have made. It is good, I guess, that these comments have resulted in the closest thing to a conversation about Wins Produced that we have to date. It is frustrating that it is the ONLY way we can have a conversation in anywhere near real time. I apologize for prolonging this, but I still think this conversation - despite how strained it is - is productive and interesting. Some of the digressions have been particularly interesting. And, as I have said before, I have learned a lot from the conversation in this thread.
Back to top

Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina
Posted: Mon Nov 20, 2006 5:48 pm Post subject:

94by50 wrote:
From the post linked above: "If you take ten shots at a below average rate you hurt your team. If you take twenty shots at a below average rate you hurt your team even more." 1) Is this statement true? 2) If it is true, how do we know that? If it is not true, how do we know that? 3) Is this universally true, or is it context-dependent? I still have trouble believing that efficiency should be judged independently, as opposed to being judged in the context of shooting frequency. I also have been trained to think that using the average as the baseline of acceptable performance is wrong. But if there's no measurable correlation between efficiency and shot volume... what's correct?
I actually sent Dave an e-mail on this point just last night. It can be shown that replacing a player with an otherwise identical player, except that the number of true shot attempts per 40 minutes is one standard deviation higher, results in about 3 more wins a season. This is holding the number of "extra" points constant (approximately points minus true shot attempts), so it corresponds to the thought experiment above. This result is statistically significant with a p-value of less that 0.0000 (which means there is less than in a 1 in 10,000 chance that this is due to random chance). The 95% confidence interval rules out that this effect is smaller than 2 wins. I generally provide Dave with far more detail about the nature of this evidence than I can provide publicly, but still no response. After probably more than a dozen attempts publicly and privately to provide evidence to bear on the claims that Wages of Wins makes, the only response has been that such evidence is "hogwash." I think he actually has used the word "silly," but I think the point is the same.
Back to top

kjb



Joined: 03 Jan 2005
Posts: 865
Location: Washington, DC
Posted: Mon Nov 20, 2006 5:50 pm Post subject:

It's interesting that Berri's crew says there's no relationship between usage and efficiency. This is a bitch of a subject to study, and their results differ from the results that Dean reported here. I don't think the methodology they describe really gets at the issue effectively because of their focus on single game performances. If a guy is shooting well on a given night, odds are, he's going to shoot more. The phrase that relationship differently -- saying repeatedly, "...the more he shoots, the better he shoots." But, they're addressing an argument that I don't think Conventional Wisdom is actually making. Rather, CW says that changing a player's role from low usage to higher usage will often bring about lower efficiency. And vice versa -- going from higher usage to lower usage will often bring about higher efficiency. Looking at game-to-game fluctuations in shot attempts and shooting percentage don't address this issue. As I recall, Dean's findings do address this issue, though I don't know what methods Dean used to reach the findings. (I recall from Dean's book that his skills curves rely heavily on the notion that efficiency tends to decline as usage increases. Was Dean merely finding fiction in the numbers?)Last edited by kjb on Tue Nov 21, 2006 9:40 am; edited 1 time in total
Back to top

Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina
Posted: Mon Nov 20, 2006 5:53 pm Post subject:

Improving the shooting percentage of teammates is not the only way that scorers can affect their teammates. In fact, it may not even be the primary way. Difficulty in creating a shot often leads not to bad shots, but to no shots at all due to increased turnovers.
Back to top

94by50



Joined: 01 Jan 2006
Posts: 499
Location: Phoenix
Posted: Mon Nov 20, 2006 7:09 pm Post subject:

Dan Rosenbaum wrote:
I actually sent Dave an e-mail on this point just last night. It can be shown that replacing a player with an otherwise identical player, except that the number of true shot attempts per 40 minutes is one standard deviation higher, results in about 3 more wins a season. This is holding the number of "extra" points constant (approximately points minus true shot attempts), so it corresponds to the thought experiment above. This result is statistically significant with a p-value of less that 0.0000 (which means there is less than in a 1 in 10,000 chance that this is due to random chance). The 95% confidence interval rules out that this effect is smaller than 2 wins.
At some point, I'd love to see the details of the study that gives this conclusion. For some time now, I've tried to figure a reasonable method to "credit" increased efficiency to players who are responsible for more than their share of their team's offense (and vice versa), but none of it would be empirically or statistically valid - it's just guesswork, so it's been more or less fruitless. How many true shot attempts per 40 minutes is one standard deviation? Dan Rosenbaum wrote:
Difficulty in creating a shot often leads not to bad shots, but to no shots at all due to increased turnovers.
Absolutely. As long as there's a shot clock, anyone who can create shots should be more valuable than a player who can't, all other things equal. Any missed shot attempt may still be rebounded by the offense. A turnover can't be reversed in any way. A missed field goal attempt has to be "worth more" to a team than a turnover. I don't see how it can be otherwise.
Back to top

Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina
Posted: Mon Nov 20, 2006 8:47 pm Post subject:

94by50 wrote:
Dan Rosenbaum wrote:
Difficulty in creating a shot often leads not to bad shots, but to no shots at all due to increased turnovers.
Absolutely. As long as there's a shot clock, anyone who can create shots should be more valuable than a player who can't, all other things equal. Any missed shot attempt may still be rebounded by the offense. A turnover can't be reversed in any way. A missed field goal attempt has to be "worth more" to a team than a turnover. I don't see how it can be otherwise.
First off, a standard deviation in true shot attempts is about 4. Second, here is another way to think about the formula for a possession (ignoring the 0.96 multiplier that I use to account for team rebounds). POSS = TO + TSM + alpha*(TSA - TSM - OREB) + (1-alpha)*oppDREB, where TO is turnovers, TSM is true shots made (field goals made + 0.44*free throws made) TSA is true shot attempts OREB is offensive rebounds oppDREB is opponent's defensive rebounds This equation works for any value of alpha. Berri and his co-authors actually talk about this formula using two different values of alpha, i.e. when it equals one and zero. Here is the formula when alpha equals one. POSS = TO + TSA - OREB Here is the formula when alpha equals zero. POSS = TO + TSM + oppDREB But if Berri and his co-authors are going to use these two formulas, they need to add them up to get the proper weights. 2*POSS = 2*TO + TSA + TSM - OREB + oppDREB This simplifies to the following. POSS = TO + TSA - 0.5*TSMissed - 0.5*OREB + 0.5*oppDREB, where TSMissed = FGMissed + 0.44*FTMissed Note that this will result in rebounds being worth less than in Wins Produced. Also, it changes the relative value of points versus shot attempts, because it accounts for the fact that missed shots may be rebounded. (That is why TSMissed ultimately comes in as a positive.) In other words, a proper application of the Wages of Wins logic results in a model very much at odds with what Wages of Wins claims. Rebounders are worth less. Shot creators are worth more. But the bigger problem is that any value of alpha will make this possession formula work. In other words, it needs to be estimated. In essence, what they have done is take an empirical question (the relative weight for offensive rebounds relative to defensive rebounds) and assumed it away. For no good reason (although I must admit that I had not thought it through like this until now). In econometric terms what they have done is take an equation that is not identified and through an arbitrary assumption simply assumed away the identification problem. (And made a mistake in doing so.) Sometimes we HAVE to do that in economics to make a model tractable. But this is not one of those cases. I am sorry for not noticing this until now. Noticing this several months ago would have saved a lot of trouble. I really was not expecting to find something like this, but some of the posts here and at the Wages of Wins blog got me thinking about this differently. It is that type of collaboration that makes a forum like this so valuable. P.S. Let me make something clear that may not be so. If you are only wanting to calculate the number of possessions, it doesn't really matter much what value you use for alpha. But if you want to use these possession equations to derive the relative value of rebounds, then it is critical to estimate alpha. So for what most of us use these possession formulas for, this is not an issue. Not so for Wins Produced, where this is the foundation of their whole methodology.Last edited by Dan Rosenbaum on Tue Nov 21, 2006 2:02 am; edited 1 time in total
Back to top

Analyze This



Joined: 17 May 2005
Posts: 364
Posted: Tue Nov 21, 2006 1:32 am Post subject:

WizardsKev wrote:
It's interesting that Berri's crew [url=http://dberri.wordpress.com/2006/06/11/ ... e-nba}says there's no relationship between usage and efficiency.[/url] This is a bitch of a subject to study, and their results differ from the results that Dean reported here. I don't think the methodology they describe really gets at the issue effectively because of their focus on single game performances. If a guy is shooting well on a given night, odds are, he's going to shoot more. The phrase that relationship differently -- saying repeatedly, "...the more he shoots, the better he shoots." But, they're addressing an argument that I don't think Conventional Wisdom is actually making. Rather, CW says that changing a player's role from low usage to higher usage will often bring about lower efficiency. And vice versa -- going from higher usage to lower usage will often bring about higher efficiency. Looking at game-to-game fluctuations in shot attempts and shooting percentage don't address this issue. As I recall, Dean's findings do address this issue, though I don't know what methods Dean used to reach the findings. (I recall from Dean's book that his skills curves rely heavily on the notion that efficiency tends to decline as usage increases. Was Dean merely finding fiction in the numbers?)
I did skill curves for all Belgium players (game by game). What I found out was that some players (an important %) actually have a higher efficiency when they use a high amount of the possessions (between 20 and 25%) than when they use a low/medium amount of the total possessions. But with most the efficiency goes down when the % of possessions goes up. This was just a small remark, now carry on with the discussion.
Back to top

Harold Almonte



Joined: 04 Aug 2006
Posts: 616
Posted: Tue Nov 21, 2006 8:20 am Post subject:

Who were that 25%, scorers, long distance shooters, some role players?
Back to top

kjb



Joined: 03 Jan 2005
Posts: 865
Location: Washington, DC
Posted: Tue Nov 21, 2006 9:44 am Post subject:

Analyze This wrote:
WizardsKev wrote:
It's interesting that Berri's crew says there's no relationship between usage and efficiency. This is a bitch of a subject to study, and their results differ from the results that Dean reported here. I don't think the methodology they describe really gets at the issue effectively because of their focus on single game performances. If a guy is shooting well on a given night, odds are, he's going to shoot more. The phrase that relationship differently -- saying repeatedly, "...the more he shoots, the better he shoots." But, they're addressing an argument that I don't think Conventional Wisdom is actually making. Rather, CW says that changing a player's role from low usage to higher usage will often bring about lower efficiency. And vice versa -- going from higher usage to lower usage will often bring about higher efficiency. Looking at game-to-game fluctuations in shot attempts and shooting percentage don't address this issue. As I recall, Dean's findings do address this issue, though I don't know what methods Dean used to reach the findings. (I recall from Dean's book that his skills curves rely heavily on the notion that efficiency tends to decline as usage increases. Was Dean merely finding fiction in the numbers?)
I did skill curves for all Belgium players (game by game). What I found out was that some players (an important %) actually have a higher efficiency when they use a high amount of the possessions (between 20 and 25%) than when they use a low/medium amount of the total possessions. But with most the efficiency goes down when the % of possessions goes up. This was just a small remark, now carry on with the discussion.
It's a good "small remark". I think we can all agree that some players would get more efficient with more possessions -- to a point -- while others would see a drop off in efficiency. Dean posted on the boards a couple times an average effect, which was a decline in efficiency as usage increases. Do you see something similar using players from Belgium?
Back to top

Analyze This



Joined: 17 May 2005
Posts: 364
Posted: Wed Nov 22, 2006 1:37 am Post subject:

@wizardskev. Yes, I can see that. @Almonte. It's mixed.
Back to top

kjb



Joined: 03 Jan 2005
Posts: 865
Location: Washington, DC
Posted: Wed Nov 22, 2006 11:49 am Post subject:

What does Malcolm Gladwell say about PER in his latest blog entry? http://gladwell.typepad.com/gladwellcom I'd read it myself, but my workplace has a firewall blocking access to many blogs. Gladwell's unfortunately is one of them.
Back to top

Analyze This



Joined: 17 May 2005
Posts: 364
Posted: Wed Nov 22, 2006 12:28 pm Post subject:

WizardsKev wrote:
What does Malcolm Gladwell say about PER in his latest blog entry? http://gladwell.typepad.com/gladwellcom I'd read it myself, but my workplace has a firewall blocking access to many blogs. Gladwell's unfortunately is one of them.
"I've long been a fan of John Hollinger, who writes about basketball for espn.com, in large part because of Hollinger's statistical system for analyzing NBA players. Hollinger calls it PERs, and I like it chiefly because I'm in favor of any system that tries to improve on what I think are our woefully inadequate intuitive judgments of basketball ability. Now David Berri--whose book, "The Wages of Wins", I wrote about a few months back--has critiqued Hollinger's methodology. Berri's argument is quite simple. As those of you who have read "Wages of Wins" know, Berri's big problem with the way we judge pro basketball players is that we over-rate the importance of how many points a player scores, and vastly under-rate the importance of things like turnovers, rebounds, and shooting percentage. Now comes Berri's critique of Hollinger: he says that Hollinger makes the same mistake. Here's the critical section: In discussing the NBA Efficiency metric – which the NBA presents at its website – I argued that this measure fails to penalize inefficient shooting. The regression of wins on offensive and defensive efficiency reveals that shooting efficiency impacts outcomes in basketball. The ball does indeed have to go through the hoop for a team to be successful. The same critique offered for NBA Efficiency also applies to Hollinger’s PERs, except the problem is even worse. Hollinger argues that each two point field goal made is worth about 1.65 points. A three point field goal made is worth 2.65 points. A missed field goal, though, costs a team 0.72 points. Given these values, with a bit of math we can show that a player will break even on his two point field goal attempts if he hits on 30.4% of these shots. On three pointers the break-even point is 21.4%. If a player exceeds these thresholds, and virtually every NBA played does so with respect to two-point shots, the more he shoots the higher his value in PERs. So a player can be an inefficient scorer and simply inflate his value by taking a large number of shots. I'd be interested to see how Hollinger replies to this. As I recall from the last time I posted on Berri, some readers have a problem with Berri's conclusions, mostly because his system ends up highly valuing players like Ben Wallace and Dennis Rodman and Kevin Garnett and dismissing the value of players like Allen Iverson. But the more Berri's fleshes out in arguments, the more convinced I become. If you're a skeptic, I urge you to start reading Berri's blog. One more point: one of the fascinating things about this argument is how similar it is to the argument currently going on in medicine about "clinical" versus "acturial" decision-making. One study after another has demonstrated that in a number of critical diagnostic situations, the unaided judgment of most doctors is substantially inferior to a diagnosis made with the assistance of some kind of algorithm or decision-rule. Doctors don't like to admit this. But it happens to be true. A lot of the huffing and puffing about Berri's ideas, it strikes me, is just basketball's version of the same defensiveness and close-mindedness."

Crow
Posts: 6146
Joined: Thu Apr 14, 2011 11:10 pm

Re: Wins Produced - Wages of Wins (Berri, Schmidt, and Brook

Post by Crow » Fri Apr 15, 2011 8:14 pm

Page 14
Author Message ziller



Joined: 30 Jun 2005
Posts: 126
Location: Sac Metro
Posted: Sun Nov 26, 2006 10:11 pm Post subject:

From Berri's discussion of Dan's critique, as linked by Dan above: Quote:
We never claimed that the Wins Produced metric by itself was “important.” I do believe this measure is a good representation of what a player does on a basketball court. Consequently, with this representation in hand, we can investigate various topics in economics. And I think some of our work in economics can be thought of as “important” (at least to us).
Is this the problem? Speaking for myself, I think looking at the rating systems is pretty damn important. If a measure purports to assess the quality of a basketball player's contribution and gets the level of mainstream attention Berri's work has, I think it's important. I know Berri et al think the same, so this comes off as posturing. If Wins Produced were just a system used to help explore economic topics in basketball and the NBA, well then to take those economic discussions seriously, one would have to consider the system pretty airtight, I would think._________________SactownRoyalty.com tziller@gmail.com
Back to top

Kevin Pelton
Site Admin


Joined: 30 Dec 2004
Posts: 976
Location: Seattle
Posted: Sun Nov 26, 2006 10:20 pm Post subject:

That was a very good post by Gladwell and I'm glad TrueHoop got a plug, but I think this comment: Malcolm Gladwell wrote:
every little tweak raises the quality of decision-making in the middle part of the curve just a little bit higher
reveals the real problem with this whole thing. How do we know the quality of decision-making is getting higher? What is the measuring stick for the measurements? I'm not sure Berri has made his answer clear, but explaining team performance is clearly a factor; explaining future team performance is as well. Dan would argue it's explaining adjusted plus-minus. John, I would think, would argue that it's about fitting the theoretical value of each action of the court. As has been argued before, none of these are wrong, they're all just different. DeanO's pointed out that even if we had an unequivoally perfect measure of performance, that wouldn't be a measure of ability, because performance is affected by a lot more than just ability. But it appears -- and I hate to speak for others, as much as it may appear otherwise -- that Berri disagrees based on the strong correlation between box-score performance in successive seasons. So that is another part of this issue.Last edited by Kevin Pelton on Sun Nov 26, 2006 10:28 pm; edited 1 time in total
Back to top

Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina
Posted: Sun Nov 26, 2006 10:25 pm Post subject:

ziller wrote:
From Berri's discussion of Dan's critique, as linked by Dan above: Quote:
We never claimed that the Wins Produced metric by itself was “important.” I do believe this measure is a good representation of what a player does on a basketball court. Consequently, with this representation in hand, we can investigate various topics in economics. And I think some of our work in economics can be thought of as “important” (at least to us).
Is this the problem? Speaking for myself, I think looking at the rating systems is pretty damn important. If a measure purports to assess the quality of a basketball player's contribution and gets the level of mainstream attention Berri's work has, I think it's important. I know Berri et al think the same, so this comes off as posturing. If Wins Produced were just a system used to help explore economic topics in basketball and the NBA, well then to take those economic discussions seriously, one would have to consider the system pretty airtight, I would think.
I don't think this is posturing. Berri is mostly interested in telling some economics-related story. To do so, he needs a metric (1) good enough to tell the story and (2) not so complicated that it crowds out the story he wants to tell. That is perfectly legitimate and something a lot of us should keep in mind. Simplicity is important. I just am not convinced that Wins Produced meets criterion (1).
Back to top

Analyze This



Joined: 17 May 2005
Posts: 364
Posted: Mon Nov 27, 2006 2:04 am Post subject:

Mister Rosenbaum, can you respond to what mister Berri wrote here: To point (d) – this statement refers to an approach Rosenbaum offered several months ago to test our model. What he did is regress the player rankings from his replication of Winston-Sagarin (described below) on player rankings derived from Wins Produced and other models. In other words, he evaluated how well each model predicted his model. We did not know that the “Holy Grail” (a term often used in the APBRmetrics community) of player performance models was the one that explained Rosenbaum’s player rankings the best. If that is the standard, though, then I guess we are going to have to live with a model that comes up short. I actually remember this post you made on this forum to which he refers and I thought the same thing when I first read that post (perhaps I was wrong). Could you also explain what the difference is between these 4+/- ratings: the Sagarin&Winston approach the Dan Rosenbaum approach (Danval?) the basketballvalue ratings the 82 games ratings. If I'm not mistaking the 82 games ratings are an unadjusted +/- rating, and the first two (or 3?) are adjusted. But what is the difference between Danval and Winston&Sagarin val? And why is your +/- approach more accurate? Would it also be possible to give a sample of last season of Danval?_________________Where There's a WilT There's a Way
Back to top

Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina
Posted: Mon Nov 27, 2006 3:05 am Post subject:

Analyze This wrote:
Mister Rosenbaum, can you respond to what mister Berri wrote here: To point (d) – this statement refers to an approach Rosenbaum offered several months ago to test our model. What he did is regress the player rankings from his replication of Winston-Sagarin (described below) on player rankings derived from Wins Produced and other models. In other words, he evaluated how well each model predicted his model. We did not know that the “Holy Grail” (a term often used in the APBRmetrics community) of player performance models was the one that explained Rosenbaum’s player rankings the best. If that is the standard, though, then I guess we are going to have to live with a model that comes up short. I actually remember this post you made on this forum to which he refers and I thought the same thing when I first read that post (perhaps I was wrong). Could you also explain what the difference is between these 4+/- ratings: the Sagarin&Winston approach the Dan Rosenbaum approach (Danval?) the basketballvalue ratings the 82 games ratings. If I'm not mistaking the 82 games ratings are an unadjusted +/- rating, and the first two (or 3?) are adjusted. But what is the difference between Danval and Winston&Sagarin val? And why is your +/- approach more accurate? Would it also be possible to give a sample of last season of Danval?
The 82 games plus/minus ratings just calculate net points per possession when a player is on the court minus net points per possession when a player is off the court. Basketballvalue (I think) is like the 82 games plus/minus ratings, but it makes an adjustment for the players a player plays with. I don't think it makes an adjustment for opposing players, but I could be wrong about that. Adjusted plus/minus statistics use a regression approach that calculates values for players - values that do the best job of explaining how teams do when given players are on the court, holding constant the effect of the other players on the court. For the purposes of examining the importance of offensive rebounders versus defensive rebounders or scorers versus assisters or one rating system versus another, there should not be much difference between different adjusted plus/minus statistic systems. The key differences would lie in how much importance was placed on clutch/garbage time play or how narrow the time periods were over which adjusted plus/minus statistics were calculated. But even those differences are unlikely to matter that much in debates like the one in this thread. (What is more likely to cause differences is if mistakes in the data (even in official NBA game logs) are not handled properly. There are circumstances under which this can affect the results fairly significantly.) Where these systems differ is going to be how they move from adjusted plus/minus statistics to a rating system, i.e. how they go about trying to predict future player value. And there the differences can be quite significant and the different systems can rate players very differently. But this is largely centers around how the analyst trades off variance vs. bias and how much (if at all) they incorporate box score statistics. This is something that is not really relevant to this thread, because the point here is not that adjusted plus/minus rating systems are necessarily the preferred rating system. So when Berri claims that I "evaluate how well each model predicts my model," that really misses the point. Is it imporant for a model to be able to explain (or predict) how well a team plays when a particular player is in the game? It is hard to imagine the conditions under which this is unimportant. Berri's argument would imply that if Big Bird used team wins as his measure of player value, that would invalidate using team wins (either in the present or the future) as a barometer for ANY other metric. Anyone could then argue that it doesn't make sense to make an argument for a model just because it does a good job explaining Big Bird's model. An argument needs to be made for why it is unimportant to explain (or predict) how a team does when a particular player is in the game. Berri has yet to make such an argument.
Back to top

Nikos



Joined: 16 Jan 2005
Posts: 346
Posted: Mon Nov 27, 2006 7:53 am Post subject:

davis21wylie2121 wrote:
I ran once a regression on PER and the other "Hollinger Stats" (TS%, AstR, ToR, Usg, RebR); here's how it went down: Code:
Coefficients Intercept -24.01548294 TS% 0.440228848 AstR 0.154933814 ToR -0.306420306 Usage 0.605165609 RebR 0.537938584 Or: PER = (0.440228848*TS%)+(0.154933814*AstR)+(-0.306420306*ToR)+(0.605165609*Usg)+(0.537938584*RebR)-24.01548294
It works to within +/- 0.97 PER. Not sure if that will help you answer your question, Kevin, but it does tell you how many units of Usage are needed to offset a drop in TS%, for instance.
Can you explain your last statement here? For some reason I can't make sense of how many Usage Units are needed to offset a disparity in TS%? (If that is what you are referring to). Thanks
Back to top

HoopStudies



Joined: 30 Dec 2004
Posts: 705
Location: Near Philadelphia, PA
Posted: Mon Nov 27, 2006 8:01 am Post subject:

admin wrote:
That was a very good post by Gladwell and I'm glad TrueHoop got a plug, but I think this comment: Malcolm Gladwell wrote:
every little tweak raises the quality of decision-making in the middle part of the curve just a little bit higher
reveals the real problem with this whole thing. How do we know the quality of decision-making is getting higher? What is the measuring stick for the measurements?
Much of Gladwell's commentary these days is related to his book, Blink. The book proclaims to be about an expert's first blink about a subject, the first two seconds in which they have to see something and form an opinion. He says we're good at it. But the book is really about decision-making, with an unsubtle emphasis on experts' decision-making, which is different than non-experts. As with Berri's book, I highly recommend Blink, but also have problems with some of the details. Ignoring those details, what makes a good book for me is one that stimulates thought and questions -- both Blink and Wages do that. I hope, for what we have here, that people discussing here really have read Wages. Back to KevinP's question -- how we determine what is good decision-making? This is actually one of my qualms with Blink -- it doesn't make clear what is good decision-making. But that is, in part, because it is hard to define. "Good" is subjective and to make something studyable or more objective, you have to give it an objective. There can be good trades for both teams if you define the objectives for the two teams the right way. Was drafting Hakeem Olajuwon over Michael Jordan a good decision? We might say it was in light of the info available, which then says that both an objective and a state of knowledge are needed to define a good decision. I'm not putting forth much of an answer, I admit, but I am trying to acknowledge it as a good question._________________Dean Oliver Author, Basketball on Paper The postings are my own & don't necess represent positions, strategies or opinions of employers.
Back to top

Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina
Posted: Mon Nov 27, 2006 3:45 pm Post subject:

At this point, probably the main argument that Berri states in defense of his claim that "people tend to exaggerate [interaction] effects is that player performance is pretty consistent across seasons." Here is his argument. http://dberri.wordpress.com/2006/11/19/ ... on-a-rock/ Quote:
And to conclude this essay, let me address the question, do players take rebounds from their teammates? The answer is “absolutely.” Of course they also take shot attempts from their teammates. In fact, we show in The Wages of Wins that the more productive your teammates the less productive you will be. But people tend to exaggerate this effect. We found the interaction effect to exist but it appears to be rather small. To see this point, consider the consistency in player performance across time. If your performance depended entirely on your teammates, player performance would fluctuate dramatically across a player’s career. This is indeed what we see in football. But in basketball we see much more consistency, suggesting the size of the interaction effects have been exaggerated.
Most players do not change teams and thus we would expect interaction effects to be pretty similar for those players from year to year. Even some players who move to a new team may inherit teammates much like the ones they left. But we can still see how the year-to-year correlation varies for players who stay with or change teams. Year to Year Correlation of Wins Produced for players who stay on same team and played 1,000 minutes in both current season and past season = 0.80. Year to Year Correlation of Wins Produced for players who change teams and played 1,000 minutes in both current season and past season = 0.66. So this means that for players who stay on the same team, last year's Wins Produced explains around 64% of the variation in this year's Wins Produced. For players who switch teams that drops to 44%. To me that is a pretty big drop. Berri makes a big deal of the differences between Wins Produced and NBA Efficiency and the correlation between those two metrics is 0.77 - larger than the season-to-season correlation for players who switch teams. So the very evidence that Berri points to but does not actually present seems to suggest that interaction effects may be quite large. If we do the same exercise with PER, the drop in correlation is 0.83 to 0.76. So PER correlations are higher to begin with and fall less when players switch teams. This suggests that Wins Produced is more susceptible to interaction effects than PER is.Last edited by Dan Rosenbaum on Mon Nov 27, 2006 4:40 pm; edited 1 time in total
Back to top

Eli W



Joined: 01 Feb 2005
Posts: 401
Posted: Mon Nov 27, 2006 4:13 pm Post subject:

Interesting stuff Dan. Would you be willing to share what the year-to-year correlations are for adjusted plus-minus and statistical plus-minus?
Back to top

Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina
Posted: Mon Nov 27, 2006 4:49 pm Post subject:

Here are the year-to-year correlations for players that play 1,000 minutes in both the current and previous season. For my statistical plus/minus measure, the correlation is 0.83 for players that change teams and 0.88 for players that don't. For my noisiest adjusted plus/minus measure (one that allocate the full error term to every player), the correlation is 0.52 for players that change teams and 0.70 for players that don't. For a less noisy adjusted plus/minus measure (one that allocates just 1/10th of the error to every player), the correlation is 0.98 for both players that change or do not change teams.
Back to top

Neil Paine



Joined: 13 Oct 2005
Posts: 774
Location: Atlanta, GA
Posted: Mon Nov 27, 2006 5:53 pm Post subject:

Nikos wrote:
davis21wylie2121 wrote:
I ran once a regression on PER and the other "Hollinger Stats" (TS%, AstR, ToR, Usg, RebR); here's how it went down: Code:
Coefficients Intercept -24.01548294 TS% 0.440228848 AstR 0.154933814 ToR -0.306420306 Usage 0.605165609 RebR 0.537938584 Or: PER = (0.440228848*TS%)+(0.154933814*AstR)+(-0.306420306*ToR)+(0.605165609*Usg)+(0.537938584*RebR)-24.01548294
It works to within +/- 0.97 PER. Not sure if that will help you answer your question, Kevin, but it does tell you how many units of Usage are needed to offset a drop in TS%, for instance.
Can you explain your last statement here? For some reason I can't make sense of how many Usage Units are needed to offset a disparity in TS%? (If that is what you are referring to). Thanks
It just means that every point of TS% is worth .44 PER points when plugged into the regression formula; likewise, each point of usage is worth .61 PER points. This doesn't tell you much in and of itself because, as Kevin points out, the values aren't standardized -- they aren't on the same scale. However, it's cool for tinkering around with the values: say I've got a player with the league average in Assist Ratio, Turnover Ratio, etc., but he's got a pitiful TS%. So how much usage would I have give to him for him to come out with a 15.00 PER? That's just one example. You can play around with any of the numbers, to get a sense for how much each "Hollinger stat" impacts PER.
Back to top

Kevin Pelton
Site Admin


Joined: 30 Dec 2004
Posts: 976
Location: Seattle
Posted: Mon Nov 27, 2006 6:23 pm Post subject:

Dan, those correlations -- are they including the team adjustment? If so, then I don't think they tell us much about interaction effects, just that players changing teams have different team adjustments. That is valuable information because Berri has argued that the team adjustment has little impact, but doesn't really speak as much to the bigger issue.
Back to top

Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina
Posted: Mon Nov 27, 2006 6:58 pm Post subject:

admin wrote:
Dan, those correlations -- are they including the team adjustment? If so, then I don't think they tell us much about interaction effects, just that players changing teams have different team adjustments. That is valuable information because Berri has argued that the team adjustment has little impact, but doesn't really speak as much to the bigger issue.
These are without the team adjustment. Like you, I figured it would be more fair to Berri to do this without the team adjustment.
Back to top

Analyze This



Joined: 17 May 2005
Posts: 364
Posted: Wed Nov 29, 2006 9:00 am Post subject:

A lot of the people who attacked Dan Rosenbaum used the following argument: (c) Publish your ideas in a respected academic journal and we will pay attention to them. Mister Rosenbaum says why wait x months, why can't we discuss it right away on the internet. I agree with that. But my question is; why don't you do both? You can discuss stuff on the net and you can publish your method in an academic journal. If you do that, nobody can ignore you anymore, like some do now._________________Where There's a WilT There's a Way
Back to top

Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina
Posted: Wed Nov 29, 2006 10:56 am Post subject:

Analyze This wrote:
A lot of the people who attacked Dan Rosenbaum used the following argument: (c) Publish your ideas in a respected academic journal and we will pay attention to them. Mister Rosenbaum says why wait x months, why can't we discuss it right away on the internet. I agree with that. But my question is; why don't you do both? You can discuss stuff on the net and you can publish your method in an academic journal. If you do that, nobody can ignore you anymore, like some do now.
First, I will be doing both, but it may be six months, a year, two years, etc. before we see the fruits of any of that. And by that time, the discussion likely will have moved on. Second, academic economists have very different interests than folks here at APBRmetrics do. In general, they don't care about how well a model evaluates players; they just want to use a "good enough" model to tell stories that have something to do with economics. So they spend most of their time thinking about issues that are very different than the issues that this community thinks about. So if I wrote an adjusted plus/minus rating-based article and sent it anywhere other than JQAS, most likely the editor and referees will be folks who will be thinking about adjusted plus/minus ratings for the first time when they read my paper. It will be interesting to hear their reactions, but it is asking a lot to expect those reviewers to catch up on what has become a long and comlex conversation through reading just a few pages from a paper. Especially since there are so many thoughts and ideas that are second nature at a forum like this that have to be explained for an audience of academic economists. Maybe Wages of Wins has increased the appetite for such a discussion among academic economists, but I am not sure. Even Dave Berri does not really seem to have much interest in this discussion. And if he doesn't (when it is a discussion of his work), it seems overly optimistic to think that anyone else would really care that much. And one other point, for most academic economists, I am not sure that JQAS counts as a respected journal. The truth of the matter is that among academic economists, sports economics is an afterthought. Two papers of mine (co-authored with my advisor at Northwestern) each individually have more google scholar citations than the sum total of all of the work of all three of the authors of Wages of Wins. That does not make me a respected academic economist (I am sure my department would trade my vita for Dave Berri's in a heartbeat), but it does give an indication of where sports economics sits in the pecking order. And within sports economics, the number of folks who spend much time thinking about the NBA is just a small fraction. So it is a very small audience and one that doesn't have a lot of interest in the issues that we discuss in this forum.
Back to top

Crow
Posts: 6146
Joined: Thu Apr 14, 2011 11:10 pm

Re: Wins Produced - Wages of Wins (Berri, Schmidt, and Brook

Post by Crow » Fri Apr 15, 2011 8:15 pm

Page 15

Author Message Analyze This



Joined: 17 May 2005
Posts: 364
Posted: Wed Nov 29, 2006 2:04 pm Post subject:

I agree with you that Berri created his metric as a tool to do economic research. But his book gets a lot of attention in the mainstream media. And that is not because he talked about the competitive balance in the nba. It is because of the nba metric. Dean Oliver his book has sold more than a few copies. So perhaps a bigger part of the population than you think has interest in this kind of work. Perhaps an academic journal is not the best way to get a lot of attention. But it's a start. And why not publish your work in a book? I would buy a copy _________________Where There's a WilT There's a Way
Back to top

Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina
Posted: Wed Nov 29, 2006 2:14 pm Post subject:

Analyze This wrote:
I agree with you that Berri created his metric as a tool to do economic research. But his book gets a lot of attention in the mainstream media. And that is not because he talked about the competitive balance in the nba. It is because of the nba metric. Dean Oliver his book has sold more than a few copies. So perhaps a bigger part of the population than you think has interest in this kind of work. Perhaps an academic journal is not the best way to get a lot of attention. But it's a start. And why not publish your work in a book? I would buy a copy
Oh, I agree there is interest in this. But that does not mean there is interest in this in academic journals. And it might be quite expensive in time and effort to find out. And since I like actually helping make things happen rather than just analyzing what happens, a book is tough.
Back to top

Analyze This



Joined: 17 May 2005
Posts: 364
Posted: Wed Nov 29, 2006 2:23 pm Post subject:

Yeah but a board is -how do you say that in English- " it passes faster". It's not as permanent as a book. And your ideas are spread over heaps of topics, there is not so much structure. You will also reach less people. And you can twist and turn as much as you want but for a lot of people not publishing but only writing on the net, will give you less authority/ make you less important and so your impact will be smaller. Not with the happy few, but outside the apbrmetrcis community this will be the case. You can replace a book by a academic journal and make the same remarks (exept the reaching more people part) . If you really want to start a big debate, or give more than "the happy few" a better understanding of basketball you need to publish in my opinion. If you just want to sell your stuff to an nba team and work behind the scenes for an nba team, than of course you don't need to publish. That's my opinion._________________Where There's a WilT There's a Way
Back to top

ziller



Joined: 30 Jun 2005
Posts: 126
Location: Sac Metro
Posted: Thu Nov 30, 2006 4:34 pm Post subject:

Another guy with an audience addresses the subject: King Kaufman at Salon. Kaufman did pan the book previously. He bases his arguments not on math necessarily, but the fact Dennis Rodman was never more valuable than Michael Jordan. EDIT: I should note this very thread is linked in the piece, thus completing the cycle._________________SactownRoyalty.com tziller@gmail.com
Back to top

kjb



Joined: 03 Jan 2005
Posts: 860
Location: Washington, DC
Posted: Fri Dec 01, 2006 10:22 am Post subject:

FFSBasketball wrote:
Gladwell continues the discourse.
Can you cut & paste and email what Gladwell wrote? My job's firewall keeps me off Gladwell's blog.
Back to top

Neil Paine



Joined: 13 Oct 2005
Posts: 774
Location: Atlanta, GA
Posted: Fri Dec 01, 2006 10:33 am Post subject:

Quote:
The Perfect and the Good I wrote a piece for the The New Yorker a few weeks ago about a group of people who have created a neural network that predicts (or tries to predict) the box office of movies from their scripts. (It's not up on my site yet, but will be soon). The piece drew all kinds of interesting responses, a handful of which pointed out obvious imperfections in the system. Those criticisms were entirely accurate. But they were also, I think, in some way beside the point, because no decision rule or algorithm or prediction system is ever perfect. The test of these kinds of decision aids is simply whether--in most cases for most people--they improve the quality of decision-making. They can't be perfect. But they can be good. In "Blink," for instance, I wrote about the use of a decision tree at Cook County Hospital in Chicago to help diagnose chest pain. Lee Goldman, the physican who devised the chest pain decision rule, says very clearly that he thinks that there are individual doctors here and there who can make better decisions without it. But nonetheless Goldman's work has saved lots and lot of lives and millions and miillions of dollars because it improves the quality of the average decision. Is the average movie executive better off with a neural network for analyzing scripts than without it? My guess is yes. That's why I wrote the piece. I think that one of the most important changes we're going to see in lots of professions over the next few years is the emergence of tools that close the gap between the middle and the top--that allow the decision-making who is merely competent to avoid his errors to be reach the level of good. I think the same perspective should be applied to the basketball algorithms I've been writing about. It is easy to point out the ways in which either Hollinger's system or Berri's system fail to completely reflect the reality of what happens on the basketball court. But of course they are imperfect: neither Berri or Hollinger would ever claim that they are not. The issue is--are we better off using them to assist decision-making that we are making entirely judgements about basketball players using conventional metrics? Here I think the answer is a resounding yes. (Keep in mind that I live in New York City and have had to watch Mr. Thomas bungled his way toward disaster. I would think that.) And the reason that lots of smart people, like Berri and Hollinger and others, spend so much time arguing back and forth about different variations on these algorithms, is that every little tweak raises the quality of decision-making in the middle part of the curve just a little bit higher. That's a pretty noble goal. That said, here are the latest updates on the Hollinger-Berri back and forth. And remember. I don't think this is a question of one of them being wrong and the other right. They are both right. It's just that one of them may be a little more right than the other. Here we go. First Hollinger's response, courtesy of truehoop.com, (an excellent site by the way.) November 26, 2006

Back to top

Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina
Posted: Fri Dec 01, 2006 11:55 am Post subject:

At Wages of Wins, Quote:
kjb // Dec 1st 2006 at 8:21 am How undramatic is the team adjustment? I’ve read Dan’s critique, in which he inserted some absurdly radical values for blocked shots (for example), and once the team adjustment was added, it had no effect on the overall ratings.
I am not sure what you mean by "overall ratings," so this might not be what I have argued. Radical values for blocked shots (or personal fouls or assists) will change the player ratings, but they will not affect how well Wins Produced explains team wins.
Back to top

kjb



Joined: 03 Jan 2005
Posts: 860
Location: Washington, DC
Posted: Fri Dec 01, 2006 1:17 pm Post subject:

Dan Rosenbaum wrote:
At Wages of Wins, Quote:
kjb // Dec 1st 2006 at 8:21 am How undramatic is the team adjustment? I’ve read Dan’s critique, in which he inserted some absurdly radical values for blocked shots (for example), and once the team adjustment was added, it had no effect on the overall ratings.
I am not sure what you mean by "overall ratings," so this might not be what I have argued. Radical values for blocked shots (or personal fouls or assists) will change the player ratings, but they will not affect how well Wins Produced explains team wins.
That's exactly what I meant. I guess I wasn't as clear as I'd hoped to be.

Crow
Posts: 6146
Joined: Thu Apr 14, 2011 11:10 pm

Re: Wins Produced - Wages of Wins (Berri, Schmidt, and Brook

Post by Crow » Fri Apr 15, 2011 8:18 pm

Back to top

tpryan



Joined: 11 Feb 2005
Posts: 96
Posted: Mon Dec 04, 2006 4:44 am Post subject:

Dan Rosenbaum wrote:
Analyze This wrote:
I agree with you that Berri created his metric as a tool to do economic research. But his book gets a lot of attention in the mainstream media. And that is not because he talked about the competitive balance in the nba. It is because of the nba metric. Dean Oliver his book has sold more than a few copies. So perhaps a bigger part of the population than you think has interest in this kind of work. Perhaps an academic journal is not the best way to get a lot of attention. But it's a start. And why not publish your work in a book? I would buy a copy
Oh, I agree there is interest in this. But that does not mean there is interest in this in academic journals. And it might be quite expensive in time and effort to find out. And since I like actually helping make things happen rather than just analyzing what happens, a book is tough.
Dan, As you may know, there is a Statistics in Sports Section of the American Statistical Association (ASA) and there is considerable interest in sports among members of ASA. The Section's leading members (Scott Berry, Jim Albert, Jay Bennett, Carl Morris, etc.) have acquired some stature in various ways (e.g., Albert has written a couple of baseball books, Morris, a star academic statistician at Harvard, has served as a consultant to the Boston Red Sox, etc.). The Journal of the American Statistical Association (JASA), a prestigious journal, occasionally publishes sports applications papers, as does Chance, which is published by ASA. (JASA is at a considerable higher level of mathematical sophistication than Chance.) Tom
Back to top

Ben F.



Joined: 07 Mar 2005
Posts: 391
Posted: Thu Dec 07, 2006 5:01 pm Post subject:

More on the whole subject, this from TrueHoop which links to I think a very soundly reasoned post.
Back to top

Ben F.



Joined: 07 Mar 2005
Posts: 391
Posted: Thu Dec 07, 2006 7:35 pm Post subject:

Matthew Yglesias weighs in as well. I found this quote particularly interesting in light of the usage vs. efficiency debate, and it seems to espouse the conventional wisdom in a simple and articulate way: Matthew Yglesias wrote:
Looking at WoW I think something almost all basketball fans have trouble with is the seeming implication that if you put five high-efficiency, low-volume shooters who were good at rebounding and avoiding turnovers on the floor simultaneously that you'd have a really effective team. This seems wrong to most of us; it seems as if in that situation the team would either see turnovers skyrocket (shot clock violations) or else shooting efficiency decline. And it would be good to have a formula that took that sort of thing into account. To construct a formula like that, though, you'd either need to just guess what would happen, or else you'd need much more data out of which to try and build an empirically grounded non-linear analysis. But since as best I can tell no coaches actually field lineups like that (they, like most people, are just assuming it wouldn't work) there isn't much to be done.

Back to top

bchaikin



Joined: 27 Jan 2005
Posts: 678
Location: cleveland, ohio
Posted: Thu Dec 07, 2006 9:25 pm Post subject:

I found this quote particularly interesting in light of the usage vs. efficiency debate, and it seems to espouse the conventional wisdom in a simple and articulate way: Matthew Yglesias wrote: Looking at WoW I think something almost all basketball fans have trouble with is the seeming implication that if you put five high-efficiency, low-volume shooters who were good at rebounding and avoiding turnovers on the floor simultaneously that you'd have a really effective team. why wouldn't you have a really effective team? as long as one was a PG, another an SG, and one a SF, a PF, and a C, why wouldn't the team be really effective?... This seems wrong to most of us; it seems as if in that situation the team would either see turnovers skyrocket (shot clock violations) or else shooting efficiency decline. not sure why one would think this - perhaps if this individual listed some of these players he had in mind it might help. i mean, if right now in 06-07 you fielded a team of antonio daniels, josh childress, quinton ross, shane battier, and ben wallace, would you seriously expect this team to have their turnovers skyrocket or their shooting efficiency dramatically decline from their current turnover rates and shooting efficiencies?... And it would be good to have a formula that took that sort of thing into account. To construct a formula like that, though, you'd either need to just guess what would happen, or else you'd need much more data out of which to try and build an empirically grounded non-linear analysis. turnovers have been kept since 1977-78. that's 29 years of data, 758 team seasons (i.e. the celtics have had 29 team seasons during this time), close to 14,800,000 total minutes of player regular season data in all. how much more data do you need?... But since as best I can tell no coaches actually field lineups like that (they, like most people, are just assuming it wouldn't work) there isn't much to be done. does this statement make any sense? any team a coach puts out for a full season is going to have similar team possessions as their opponents (within 4 for every game), and somewhat similar total touches or scoring opportunities over the span of a season. you field a team that previously had players/starters from different teams at each position that had say 20% less scoring attempts per minute than the average player at their positions in the league, then put them all together on the same team they are obviously going to increase their overall scoring attempts, with the certain amount of each increase being dependent on each player's touches/min and what they did per touch.... as for whether one thinks their turnovers would increase instead, there is plenty of data to show how players have performed with an increase (or decrease) in scoring attempts per minute over their careers, on the same team and also on different teams...
Back to top

deepak



Joined: 26 Apr 2006
Posts: 664
Posted: Thu Dec 07, 2006 11:41 pm Post subject:

bchaikin wrote:
why wouldn't you have a really effective team? as long as one was a PG, another an SG, and one a SF, a PF, and a C, why wouldn't the team be really effective?...
You might have an effective team, but probably not. The reason is if all the players are low usage players, then there's a good chance they don't have the requisite skills for creating shots for themselves or teammates effectively (ball handling, general offensive versatility, etc.). If you don't have such players on the court, they will have a tough time getting good shots against a decent NBA defense. I think that's the intuition most people have.
Back to top

bchaikin



Joined: 27 Jan 2005
Posts: 678
Location: cleveland, ohio
Posted: Thu Dec 07, 2006 11:54 pm Post subject:

You might have an effective team, but probably not. The reason is if all the players are low usage players, then there's a good chance they don't have the requisite skills for creating shots for themselves or teammates effectively (ball handling, general offensive versatility, etc.). If you don't have such players on the court, they will have a tough time getting good shots against a decent NBA defense. I think that's the intuition most people have. what about the five i mentioned - daniels, childress, ross, battier, and wallace? is this a good example or a bad example of a team that would not be effective? let's forget the generalizations here - can anyone list a couple of examples of 5 players, one at each position, that are high efficiency, low volume players, maybe good rebounders or if not then good defenders, but also good at avoiding turnovers, that if playing together would have their turnovers skyrocket and/or their shooting efficiency drop?...
Back to top

Analyze This



Joined: 17 May 2005
Posts: 364
Posted: Fri Dec 08, 2006 2:54 am Post subject:

bchaikin wrote:
This seems wrong to most of us; it seems as if in that situation the team would either see turnovers skyrocket (shot clock violations) or else shooting efficiency decline. not sure why one would think this - perhaps if this individual listed some of these players he had in mind it might help. i mean, if right now in 06-07 you fielded a team of antonio daniels, josh childress, quinton ross, shane battier, and ben wallace, would you seriously expect this team to have their turnovers skyrocket or their shooting efficiency dramatically decline from their current turnover rates and shooting efficiencies?...
I think that. The team that you mention has only scoring role players. That team can't score enough to win games if each player did what he did today (the same role). Some of those players should need to take very different roles. If some of them try to score much more, they probably would try something they can't do (namely stay effective when using a much bigger part of the scoring possessions), they would force it (higher to ratio), and their ef fg% would go down (because when they use a big part of the scoring possessions they are not able to stay as effective as when they used a small part of the scoring possessions)_________________Where There's a WilT There's a Way
Back to top



Page 16

Author Message royce.toyfu



Joined: 11 Jul 2006
Posts: 19
Posted: Fri Dec 08, 2006 11:20 am Post subject:

bchaikin wrote:
This seems wrong to most of us; it seems as if in that situation the team would either see turnovers skyrocket (shot clock violations) or else shooting efficiency decline. not sure why one would think this - perhaps if this individual listed some of these players he had in mind it might help. i mean, if right now in 06-07 you fielded a team of antonio daniels, josh childress, quinton ross, shane battier, and ben wallace, would you seriously expect this team to have their turnovers skyrocket or their shooting efficiency dramatically decline from their current turnover rates and shooting efficiencies?...
Do you think that if Ben Wallace doubled his shot attempts he would double his points? How many low-volume shooters do you think there are whose shooting percentages aren't helped by limiting their attempts? Just the five you mentioned? bchaikin wrote:
And it would be good to have a formula that took that sort of thing into account. To construct a formula like that, though, you'd either need to just guess what would happen, or else you'd need much more data out of which to try and build an empirically grounded non-linear analysis. turnovers have been kept since 1977-78. that's 29 years of data, 758 team seasons (i.e. the celtics have had 29 team seasons during this time), close to 14,800,000 total minutes of player regular season data in all. how much more data do you need?...
I think his point isn't that there haven't been turnovers, its that there aren't teams made up of players that are high efficiency low volume shooters. This may not be true, such a team has probably been assembled at some time or another. I don't know of any off hand, but my knowledge is limited.
Back to top

kjb



Joined: 03 Jan 2005
Posts: 865
Location: Washington, DC
Posted: Fri Dec 08, 2006 11:39 am Post subject:

If you think about it, it's literally impossible to field a team comprised only of low volume, high efficiency players. Because once you put 5 guys of that type on the floor, SOMEONE has to take the shots. They're not going to just keep passing the ball around looking for an open shot to the point of getting repeated shot clock violations. At some point, someone on the team will pull the trigger, even if it's not a shot he'd normally attempt. And, I suspect at that point, we'd see that player's shooting efficiency begin to decline. Because, instead of catch & shoot open jumpers with his feet set, he might have to put the ball on the floor, make a move, or shoot off balance with a hand in his face.
Back to top

bchaikin



Joined: 27 Jan 2005
Posts: 689
Location: cleveland, ohio
Posted: Fri Dec 08, 2006 1:15 pm Post subject:

If you think about it, it's literally impossible to field a team comprised only of low volume, high efficiency players. Because once you put 5 guys of that type on the floor, SOMEONE has to take the shots. They're not going to just keep passing the ball around looking for an open shot to the point of getting repeated shot clock violations. At some point, someone on the team will pull the trigger, even if it's not a shot he'd normally attempt. And, I suspect at that point, we'd see that player's shooting efficiency begin to decline. Because, instead of catch & shoot open jumpers with his feet set, he might have to put the ball on the floor, make a move, or shoot off balance with a hand in his face. if tomorrow gilbert arenas, antawn jamison, and caron butler all slip on the same bar of soap in the shower, and are all out for the next month, and your starting five for that time is antonio daniels, deshawn stevenson, jarvis hayes, etan thomas, and brendan haywood - what would happen? not record wise but in terms of shooting efficiency and turnovers. can you say definitively that either the shooting efficiency would decrease, or the turnovers would increase, because all of their touches (or touches/min) would increase above what they are now?... as a group arenas, jamison, and butler are now shooting a ScFG% of 53.4%, the other five a ScFG% of 52.7%. the career ScFG% of that other five also is 51.2%. the current combined rate of turnovers per touch for the first three players this season is 6% (about 1 turnover for every 16-17 touches on offense), for the other five its 7% (about 1 turnover for every 14-15 touches on offense). so what would happen? I think that. The team that you mention has only scoring role players. That team can't score enough to win games if each player did what he did today (the same role). are you under the impression that if you put these 5 players together today that they would only do what they are doing today, that their touches would not increase? if today you put allen iverson (31.6 pts/g), joe johnson (28.3), carmelo anthony (32.2), zach randolph (25.8 ), and yao ming (25.6) on the same team would that team be scoring 143 points per game? or do you think that maybe each's touches would decrease?...
Back to top

Neil Paine



Joined: 13 Oct 2005
Posts: 774
Location: Atlanta, GA
Posted: Fri Dec 08, 2006 1:18 pm Post subject:

This is why I suggested that there might be some kind of criteria for determining who will take on more possessions/touches, and by the same token who can afford to take more possessions without a decline in efficiency. I said that it was some combination of athleticism, ballhandling, and size that determined a player's "peak usage", but with finals looming I haven't had time to really go any further in the study. But I do think that this matter of usage vs. efficiency is absolutely the fundamental debate of APBRmetrics -- far more so than the Hollinger/Berri tiff over the value of rebounds, etc.
Back to top

Analyze This



Joined: 17 May 2005
Posts: 364
Posted: Fri Dec 08, 2006 1:35 pm Post subject:

bchaikin wrote:
are you under the impression that if you put these 5 players together today that they would only do what they are doing today, that their touches would not increase?
Of course not Bobby. There is a reason why players like antonio daniels, josh childress, quinton ross, shane battier, and ben wallace use a small % of the team scoring possessions and let other players use the biggest % of the team possessions. The reason is that they are not talented enough to be effective in scoring while being the first or second scoring option. These players that you mention are only very effective when they are scoring role players. They profit of the defensive attention that goes to the main scoring weapons and the freedom that they get as a result. (or they have other strengths, for example defense) They don't have the talent to be an effective big time scorer. They can't do what the Bryants of the world can do, namely scoring really a lot and efffective with the defense of the opponent focussed on them. I'm under the impression that if you put those 5 scoring role players on the floor some of them should need to increase their shot attempts enormous (if they want their team to win) and the effect would be that their scoring weaknesses come to the surface and the to ratio goes up, and the ts% goes down. They don't have the talent to be big time scorers, so you would get forced play/shots (to's) and less effective scoring. bchaikin wrote:
if today you put allen iverson (31.6 pts/g), joe johnson (28.3), carmelo anthony (32.2), zach randolph (25.8 ), and yao ming (25.6) on the same team would that team be scoring 143 points per game? or do you think that maybe each's touches would decrease?...
I think that there would be no team actually, but you would get a war of ego's. I don't think that some of those players would be ok in becoming a role player (that would be necessary for team succes) Some of these players are very talented in scoring a lot efficient (for example Ming, ). He can of course stop doing that and become a role player. But a player like for example Ben Wallace does not posses the scoring talent to score as much efficient as Yao Ming. He can't do that because his offensive skills are a lot weaker than those of the great wall of China. If a superstar needs to become a role player, than he needs to swallow his pride and ego and take less attempts. If he can do that it could work. But a player who is not talented enough to score a lot efficient, can simply not do that, ego or no ego. If he has not the talent to score a lot efficient, he has not the talent to score a lot efficient. So if you take a team of 5 scoring role players and you ask some of them (for example Quinten Ross and Ben Wallace) to do what they can't do namely scoring a lot efficient with the defense focussed on them their lack of scoring talent will become visible and you will get more to's and lower ts% from these players then when you ask them to be scoring role players (: the role they are good in)._________________Where There's a WilT There's a Way
Back to top

bchaikin



Joined: 27 Jan 2005
Posts: 689
Location: cleveland, ohio
Posted: Fri Dec 08, 2006 2:15 pm Post subject:

There is a reason why players like antonio daniels, josh childress, quinton ross, shane battier, and ben wallace use a small % of the team scoring possessions and let other players use the biggest % of the team possessions. The reason is that they are not talented enough to be effective in scoring while being the first or second scoring option. These players that you mention are only very effective when they are scoring role players. They profit of the defensive attention that goes to the main scoring weapons and the freedom that they get as a result. (or they have other strengths, for example defense) tell me, what's the difference between these players and say tony campbell and terry catledge prior to 1989-90?... They don't have the talent to be an effective big time scorer. who said they have to be an effective big time scorer? eight teams last season didn't even have a 20 pts/g scorer, the same so far this season... They can't do what the Bryants of the world can do, namely scoring really a lot and effective with the defensive of the opponent focussed on them. so who can do what bryant does? their are very few bryant type players in the league, but there are 30 teams of players... I think that there would be no team actually, but you would get a war of ego's. Some of these players are very talented in scoring a lot efficient (for example Ming, ). He can of course stop doing that and become a role player. It would not be a good idea, and he would need to swallow his ego. But a player like for example Ben Wallace does not posses the scoring talent to score as efficient as Yao Ming. He can't do that because his offensive skills are a lot weaker than those of the great wall of China. If a superstar needs to become a role player, than he needs to swallow his pride and ego. If he can do that it could work. But a player who is not talented enough to score a lot efficient, can simply not do that, ego or not ego. this is a very nice soliloquy, but what about their touches, their shooting efficiencies, their rates of turnovers? how would they be affected?...
Back to top

Analyze This



Joined: 17 May 2005
Posts: 364
Posted: Fri Dec 08, 2006 2:30 pm Post subject:

Bob, I don't like to discuss why the earth is not flat and why if you put 5 starters who are limited in scoring (and are scoring role players) all together in a team without more talented scorers you would see that their ts % would go down and their to ratio go up if they need to do something that they can't do (taking a much bigger bite out of the scoring possessions). Because it's obvious! And responding to your 2 and 3 remarkbchaikin wrote:
who said they have to be an effective big time scorer? eight teams last season didn't even have a 20 pts/g scorer, the same so far this season...so who can do what bryant does? their are very few bryant type players in the league, but there are 30 teams of players...
; let's suppose that you spread the offense and you don't have an offense with one enormous possession user (Bryant type), with a big number one scorer. Than you would need 3 a 4 guys who use a medium/high number of possessions. You don't have so many guys on your team who can become a "second scorer" type in the nba and be as efficient in scoring and to ratio as they were when they were scoring role players. Because they lack the talent. In scoring talent; Daniels in not Billups, Childress is not the masked man, Ross is not Prince, Battier is not Rasheed (you actually have no pf) and well Ben is Ben (I will give you that.) Detroit who in the past has been a spread the offense team,has 4 medium scorers and two of them can step it up. Your team has one such player (Battier) and he can't step it up like the Detroit best two. So forget about spreading the offense, the talent is missing. The discussion was not if there are yes or no nba teams with a starting 5 that is as limited in the scoring department as a team with Ben Wallace, Shane Battier, Josh Childress, Quinten Ross and Antonio Daniels as the starting 5 and how succesfull the team would be. You have good and bad teams and everyhing in between. My point was that if you put 5 scoring role players in a team (namely Ben wallace, Shane Battier, Josh Childress, Quinten Ross, Antonio Daniels as the real starting 5, their ts % and to ratio would become worse because some of them would need to increase their scoring possessions a lot and they don't have the talent to do that and at the same time score a lot very efficient with the defense focussed on them._________________Where There's a WilT There's a WayLast edited by Analyze This on Fri Dec 08, 2006 3:33 pm; edited 5 times in total
Back to top

asimpkins



Joined: 30 Apr 2006
Posts: 245
Location: Pleasanton, CA
Posted: Fri Dec 08, 2006 2:47 pm Post subject:

If we built a team out of those players, I would defintely bet that a few of the players would see drops in their turnover rate and true shooting percentage due to playing a more prominate role in the offense. It's possible, though, that player's like Daniels and Childress might have another level to them and could pick up the slack without much of a dropoff -- allowing the other players to remain in their comfort zones. But I would bet against it. As a side question, bchaikin, what do you think the ceiling for this team would be? Assuming solid coaching and good bench, would you see this starting squad as championship contenders? Or are you only arguing that they'd be competitive?
Back to top

Analyze This



Joined: 17 May 2005
Posts: 364
Posted: Fri Dec 08, 2006 2:59 pm Post subject:

asimpkins wrote:
It's possible, though, that player's like Daniels and Childress might have another level to them and could pick up the slack without much of a dropoff -- allowing the other players to remain in their comfort zones. But I would bet against it.
That is not possible, because if Ross, Wallace and Battier used the same % of possessions as they do now (remain in their comfort zone), to have enough scoring Daniels or Childress must become a number one scorer in terms of possessions and the other one needs to become a second scorer in terms of possessions. They can't do that and stay as effective with ts% and to ratio as they were as limited scoring players. The distance between their current % of possessions and the possessions they would need to get is too big. If the distance was smaller I would perhaps agree with you on that one. And if you don't go that way, and you spread the offense you need 3 or 4 players with a medium/high number of possessions and you don't have them on that team (if you want they keep their ts% and to ratio) asimpkins wrote:
What do you think the ceiling for this team would be? Assuming solid coaching and good bench, would you see this starting squad as championship contenders?
The bottom would be the ceiling_________________Where There's a WilT There's a Way
Back to top

Neil Paine



Joined: 13 Oct 2005
Posts: 774
Location: Atlanta, GA
Posted: Fri Dec 08, 2006 3:35 pm Post subject:

asimpkins wrote:
It's possible, though, that player's like Daniels and Childress might have another level to them and could pick up the slack without much of a dropoff -- allowing the other players to remain in their comfort zones. But I would bet against it.
Before this circular argument -- that we've had a million times before -- goes any further, we really need to find what factors determine a player's peak usage (or, as you put it, "does this guy have another level in him?"). Bob will point out numerous example of guys who increased their usage, but either stayed the same or actually improved their efficiency. But DeanO did a study that found a general trend of an inverse relationship: for every 1% added to %Poss, ORtg drops by 0.6. The inverse relationship has to be the "rule", but it is not linear, and not the same for all players (or even any two players). Therefore, we need to find what factors make guys exceptions to the rule. We also need to find what factors influence the slope of a player's skill curve. That's the next step in this discussion.Last edited by Neil Paine on Fri Dec 08, 2006 4:21 pm; edited 2 times in total
Back to top

kjb



Joined: 03 Jan 2005
Posts: 865
Location: Washington, DC
Posted: Fri Dec 08, 2006 4:01 pm Post subject:

bchaikin wrote:
if tomorrow gilbert arenas, antawn jamison, and caron butler all slip on the same bar of soap in the shower, and are all out for the next month, and your starting five for that time is antonio daniels, deshawn stevenson, jarvis hayes, etan thomas, and brendan haywood - what would happen? not record wise but in terms of shooting efficiency and turnovers. can you say definitively that either the shooting efficiency would decrease, or the turnovers would increase, because all of their touches (or touches/min) would increase above what they are now?... as a group arenas, jamison, and butler are now shooting a ScFG% of 53.4%, the other five a ScFG% of 52.7%. the career ScFG% of that other five also is 51.2%. the current combined rate of turnovers per touch for the first three players this season is 6% (about 1 turnover for every 16-17 touches on offense), for the other five its 7% (about 1 turnover for every 14-15 touches on offense). so what would happen?
Shhhhh -- don't give Eddie ideas. He's already nutty enough with his lineups. That lineup would have problems scoring the ball. Daniels is a solid PG, but I don't think he could be a lead scorer. With good finishers, I think he's a terrific creator. He's good running screen/roll, he's good at getting to the FT line. But, he's not much of a shooter. And, as a passer, he'd need finishers. Neither Stevenson nor Hayes shoot the ball well. Nor do they finish well in traffic. Stevenson would probably become less efficient -- higher usage has coincided with lower offensive efficiency during his career. Last season was the lowest usage of his career and his best shooting. This season, his usage is a touch lower, and his efg is the best of his career. Ask him to create offense, and his efg will drop -- seriously, it's borderline embarrassing to watch him try. Hayes would likely try to pick up the scoring slack -- he's the one guy from that lineup who'd be willing to put up the shot. Unfortunately, despite a nice-looking stroke, his shot doesn't go down enough. He looked to be making strides last season, but 2 kneecap fractures in 2 years haven't helped. The team could go inside to Thomas and Haywood more. That would probably be the best bet, particularly with Thomas in the post. However, Thomas has been poor at passing out of the post when doubled, as has Haywood. And Haywood has iffy footwork on his post moves and gets called for traveling with some frequency. Of course, that lineup doesn't really address the question being posed here. Daniels is efficient, although he doesn't shoot well. Stevenson and Hayes are not efficient players. Haywood isn't really much of a post scoring threat -- he does better getting his points from the offensive glass. Etan is a decent post scorer, but with weak offensive teammates and a weakness passing the ball back out of the post when doubled, there are question marks with him too.
Back to top

asimpkins



Joined: 30 Apr 2006
Posts: 245
Location: Pleasanton, CA
Posted: Fri Dec 08, 2006 4:13 pm Post subject:

Analyze This wrote:
That is not possible, because if Ross, Wallace and Battier used the same % of possessions as they do now (remain in their comfort zone), to have enough scoring Daniels or Childress must become a number one scorer in terms of possessions and the other one needs to become a second scorer in terms of possessions.
I don't think it's probable, but it's certainly possible that some of these players could increase their usage without a loss in efficiency. I'm pretty much in agreement with you. But I do believe that there are players out there that aren't being used to their fullest potential. I'm not sure these are the guys, but I see no reason to push it out of the realm of possibility.
Back to top

Kevin Pelton
Site Admin


Joined: 30 Dec 2004
Posts: 979
Location: Seattle
Posted: Fri Dec 08, 2006 4:21 pm Post subject:

davis21wylie2121 wrote:
Therefore, we need to find what factors make guys exceptions to the rule. We also need to find what factors influence the slope of a player's skill curve. That's the next step in this discussion.
I disagree with that. I think that is still a step away. While there is a general sense of "the rule," it's not very well derived mathematically or studied. Let's get the rule down before we go looking for exceptions.
Back to top

Analyze This



Joined: 17 May 2005
Posts: 364
Posted: Fri Dec 08, 2006 4:33 pm Post subject:

asimpkins wrote:
Analyze This wrote:
That is not possible, because if Ross, Wallace and Battier used the same % of possessions as they do now (remain in their comfort zone), to have enough scoring Daniels or Childress must become a number one scorer in terms of possessions and the other one needs to become a second scorer in terms of possessions.
I don't think it's probable, but it's certainly possible that some of these players could increase their usage without a loss in efficiency. I'm pretty much in agreement with you. But I do believe that there are players out there that aren't being used to their fullest potential. I'm not sure these are the guys, but I see no reason to push it out of the realm of possibility.
I agree that some players are out there that aren't used to their full potential, and I agree that the possibility could exist that some of the players we discuss could increase their usage without a loss of efficiency. What I'm saying is that when the 3 players you mention (Big Ben, Ross, Battier) keep in their comfort zone (which means that one of them uses a rather low number of possesions and the other 2 a marginal number of scoring possessions) the other 2 that you mention (namely Childress and Daniels) should take such a big jump in scoring possessions that it is not possible that they keep the same ts % and to ratio (because are not talented enough to do that). The jump that they need to take in that case (from role player scorers) to an nba first and second scoring option is so big that it's not possible._________________Where There's a WilT There's a Way
Back to top

Neil Paine



Joined: 13 Oct 2005
Posts: 774
Location: Atlanta, GA
Posted: Fri Dec 08, 2006 4:35 pm Post subject:

Well, Dean's inverse relationship was significant at more than 99% -- in his words, "There is roughly a 1 in 10 to the power of 38 that it's mere luck. According to my old quantum phys professor, that would be roughly the odds of banging your head against the wall and none of the molecules in your head actually hitting any of the molecules in the wall, thus avoiding pain." I'd say, then, that "For every increase of 1% in usage, offensive rating drops by about 0.6" probably qualifies as a general rule. But the slope for any player is rarely exactly -0.6, meaning that there are some factors out there that influence what the slope actually is -- factors that make players exceptions to the general 1:-0.6 rule. And I'd like to know what those factors are. But since anybody who's ever played basketball intuitively knows that at some point there's an inverse relationship in their own efficiency, the conventional wisdom (the null hypothesis) espouses an inverse relationship, and Dean's study espouses an inverse relationship, can't we just agree that for most players there is some kind of inverse relationship after a certain degree of usage?

Page 17

Author Message Analyze This



Joined: 17 May 2005
Posts: 364
Posted: Fri Dec 08, 2006 4:47 pm Post subject:

davis21wylie2121 wrote:
Before this circular argument -- that we've had a million times before -- goes any further, we really need to find what factors determine a player's peak usage (or, as you put it, "does this guy have another level in him?"). Bob will point out numerous example of guys who increased their usage, but either stayed the same or actually improved their efficiency. But DeanO did a study that found a general trend of an inverse relationship: for every 1% added to %Poss, ORtg drops by 0.6. The inverse relationship has to be the "rule", but it is not linear, and not the same for all players (or even any two players). Therefore, we need to find what factors make guys exceptions to the rule. We also need to find what factors influence the slope of a player's skill curve. That's the next step in this discussion.
Research showed and I'm seeing for the Belgium competition the same, namely that the general trend (rule) is that how more possessions a player uses (above his level) the lower his of rating becomes. And I also see exceptions to that rule. But these players are part of a team with talented scoring players. In this specific case we discuss we have an unrealistic situation of 5 scoring role players of different teams who would become the starting 5 of a team. 2 of those players are very limited in scoring (Ross who I saw play in Belgium/and Big Ben). I can't believe that if you would make Big Ben a medium scorer or a second scoring option that his ts % and to ratio would stay the same as what it is now as a scoring role player, because this specific player has not the talent to succesfull do that. I don't need further research in that area to realize that. I know the scoring abilities of Ben wallace. That is just common sense. Just like it would be incredible that the other 3 players of that team (Childress, Battier, Daniels) all three are exceptions to the general rule that the of rating drops when the % of possessions goes up (what are the odds) and nobody of their teams realized in their career (Battier and Daniels are already a long time in the nba) that they could carry an offense with the same efficiency as they are scoring role players._________________Where There's a WilT There's a Way
Back to top

schtevie



Joined: 18 Apr 2005
Posts: 411
Posted: Fri Dec 08, 2006 5:40 pm Post subject:

Could someone who has not misplaced their copy of Dean's book read the fine print and relate what exactly the experiment/calculation was that yielded the fact of a 1% increase in possessions usage being associated with a decrease in offensive rating of 0.6? (And for that matter can someone also remind me of what is Dean's definition of Offensive Rating?)
Back to top

94by50



Joined: 01 Jan 2006
Posts: 499
Location: Phoenix
Posted: Fri Dec 08, 2006 6:37 pm Post subject:

I don't think it's in BoP. The rule of thumb he gives in the book is a 1/1 ratio instead of a 0.6/1 ratio.
Back to top

Neil Paine



Joined: 13 Oct 2005
Posts: 774
Location: Atlanta, GA
Posted: Fri Dec 08, 2006 6:47 pm Post subject:

Dean talks about his experiment in this thread. I once estimated the slope based on the graph in BoP, and found it to be -13/14 decrease in ORtg for every 1% increase in %Poss, which is basically -1. The definition of Offensive Rating is Points Produced divided by Possessions Used.
Back to top

bchaikin



Joined: 27 Jan 2005
Posts: 687
Location: cleveland, ohio
Posted: Sun Dec 10, 2006 3:13 pm Post subject:

Do you think that if Ben Wallace doubled his shot attempts he would double his points? How many low-volume shooters do you think there are whose shooting percentages aren't helped by limiting their attempts? do you have statistical evidence of low volume shooters who increase their scoring attempts showing that their shooting percentages decrease (or increase)?... Bob, I don't like to discuss... why if you put 5 starters who are limited in scoring (and are scoring role players) all together in a team without more talented scorers you would see that their ts % would go down and their to ratio go up if they need to do something that they can't do (taking a much bigger bite out of the scoring possessions). Because it's obvious! might you have some statistical evidence to back up this obvious statement? what's so obvious to you might not be so obvious to others... In scoring talent; Daniels in not Billups, Childress is not the masked man, Ross is not Prince, Battier is not Rasheed (you actually have no pf) and well Ben is Ben (I will give you that.) Detroit who in the past has been a spread the offense team,has 4 medium scorers and two of them can step it up. Your team has one such player (Battier) and he can't step it up like the Detroit best two. So forget about spreading the offense, the talent is missing. the pistons last season (64-18 ) won more games than not only any other team in the league last year, but more games than any other team has in the past six seasons. is there some pertinence to this statement to the current discussion? why not compare this supposed team of five players to the 04-05 hawks that went just 13-69? would that be relevant?... My point was that if you put 5 scoring role players in a team (namely Ben wallace, Shane Battier, Josh Childress, Quinten Ross, Antonio Daniels as the real starting 5, their ts % and to ratio would become worse because some of them would need to increase their scoring possessions a lot and they don't have the talent to do that and at the same time score a lot very efficient with the defense focussed on them. again might you have any statistical evidence corroborating this?... It's possible, though, that player's like Daniels and Childress might have another level to them and could pick up the slack without much of a dropoff -- allowing the other players to remain in their comfort zones. both of these players have gotten more touches/min in previous seasons than in this current season. just 3 seasons ago daniels started over 1/3 of the sonics' games and got 50% more touches/min than he's getting now. as a matter of fact, had he played 3000 minutes that season he was on pace to get almost 600 assists, score almost 14 pts/g, with just 120 turnovers. that's awfully efficient. oh, and btw, he shot a career best ScFG% of over 58% that season. so contrary to what some may think daniels can or cannot do with increased touches or possessions, the data shows he has played his best with more touches... childress got over 25% more touches his rookie season than he's getting right now, about 20% more his 2nd season than now. his 2d season he shot a just as good 60% ScFG% as he's shooting now, with a similar rate of turnovers per touch as he's showing now... in 05-06 wallace had one of his highest touches/min of his career, and shot a ScFG% of 48.7%, which was better than his 46.8% career ScFG% coming into that season, and he did this with a turnover per touch rate significantly less than his career rate of turnovers per touch coming into the season... As a side question, bchaikin, what do you think the ceiling for this team would be? Assuming solid coaching and good bench, would you see this starting squad as championship contenders? Or are you only arguing that they'd be competitive? i'm only pointing out that this idea that their shooting efficiency would go south or their turnover rates skyrocket is nonsense, because the past statistical evidence of what players have done with increased touches does not show this. some increase, some decrease, and most stay about the same... That is not possible, because if Ross, Wallace and Battier used the same % of possessions as they do now (remain in their comfort zone),.. why do you insist that these players use the same number of possessions? are you under the assumption that players use the same number of possessions each season? if you are i suggest you check the statistical record... They can't do that and stay as effective with ts% and to ratio as they were as limited scoring players. The distance between their current % of possessions and the possessions they would need to get is too big. again might you have any data supporting this statement? because the data i see suggests otherwise... Of course, that lineup doesn't really address the question being posed here. Daniels is efficient, although he doesn't shoot well. over the past four seasons the average ScFG% for PGs has been 51.1%. daniels has a ScFG% of 55.1% during this same time. of the 39 PGs to have played at least 5000 total minutes those four years, that's the 4th highest ScFG%. his eFG% was at just the average of these 39 PGs, so yes while he may not shoot well, his overall scoring efficiency is very much on the high side. in the past two seasons daniels has gotten fouled more per touch than all PGs other than iverson and arenas (is either one really a PG?), and he shot 83% from the FT line. that's why his ScFG% is so high.... top this off with the fact that his rate of turnovers per touch is absolutely one of the best/lowest of all PGs the past 4 years, and so his points scored per zero point team possession during this time has been the best of all PGs, better than even billups or nash.... doesn't mean he's a better PG (he's not) than either of those two, but it does mean he's very good, and certainly capable of more touches without a decrease in efficiency, either scoring wise or turnover wise... Shhhhh -- don't give Eddie ideas. (shhhh - wonder if eddie realizes this about daniels? i won't tell him if you won't, they're in the same conference as the cavs).... Stevenson and Hayes are not efficient players. Haywood isn't really much of a post scoring threat -- he does better getting his points from the offensive glass. Etan is a decent post scorer, but with weak offensive teammates and a weakness passing the ball back out of the post when doubled, there are question marks with him too. agreed - but the question is would they be even less efficient with more touches, either shooting worse or increasing their rate of turnovers per touch? i just don't see any evidence to suggest either would happen... stevenson shot his career best ScFG% last season, but as of last sunday (16 games played) he was shooting even better than that overall, and with more touches/min, plus less turnovers per touch. hayes was shooting his worst, but with his lowest turnovers per touch and highest touches/min. both haywood and thomas have high rates of turnovers per touch, such that you most likely would not want to increase their touches. but whether their turnovers per touch would increase with an increase in touches/min i see no evidence for this from their statistical histories... when the 3 players you mention (Big Ben, Ross, Battier) keep in their comfort zone (which means that one of them uses a rather low number of possesions and the other 2 a marginal number of scoring possessions) the other 2 that you mention (namely Childress and Daniels) should take such a big jump in scoring possessions that it is not possible that they keep the same ts % and to ratio (because are not talented enough to do that). The jump that they need to take in that case (from role player scorers) to an nba first and second scoring option is so big that it's not possible. you have any data to substantiate these claims?... According to my old quantum phys professor,... remember when eveyone took it for granted that there was an ether for the transmisson of electromagnetic waves? then maxwell came along and nixed that long held belief with evidence to the contrary (now though we have a Higgs field, the 21st century's answer to the long esteemed ether. lucky us...)... and remember when everyone was under the assumption the rate of the expansion of the universe was slowing down due to gravity - everyone knew it was true, right? but saul perlmutter's relatively recent supernova data suggests otherwise, that not only is the expansion of the universe not slowing down, nor even staying constant, but is in fact speeding up (opening up a new or now more envious field of study - dark energy)... just like how now everyone knows that in the nba efficiency falls off with usage, huh?... ...can't we just agree that for most players there is some kind of inverse relationship after a certain degree of usage? it'd be nice if we had the data to show this... I can't believe that if you would make Big Ben a medium scorer or a second scoring option that his ts % and to ratio would stay the same as what it is now as a scoring role player, because this specific player has not the talent to succesful do that. I don't need further research in that area to realize that. you are the fortunate one.... I know the scoring abilities of Ben wallace. That is just common sense. aw, come on... give big ben a chance... unfortunately backing up your claims with comments like "...it's obvious..." or "...that is just common sense..." rather than with statistical evidence does not make them factual.... Just like it would be incredible that the other 3 players of that team (Childress, Battier, Daniels) all three are exceptions to the general rule that the of rating drops when the % of possessions goes up (what are the odds)... the general rule? this pounded into some stone tablet like the ten commandments? now that would be incredible.... ...and nobody of their teams realized in their career (Battier and Daniels are already a long time in the nba) that they could carry an offense with the same efficiency as they are scoring role players. they don't have to be able to carry an offense - just take on more touches without a loss of scoring efficiency and no increase in turnover rate...
Back to top

Eli W



Joined: 01 Feb 2005
Posts: 402
Posted: Sun Dec 10, 2006 3:53 pm Post subject:

bchaikin wrote:
Do you think that if Ben Wallace doubled his shot attempts he would double his points? How many low-volume shooters do you think there are whose shooting percentages aren't helped by limiting their attempts? do you have statistical evidence of low volume shooters who increase their scoring attempts showing that their shooting percentages decrease (or increase)?...
Ben Wallace was given a much larger offensive role under Larry Brown in 03-04 and 04-05 than he had under Rick Carlisle previously or Flip Saunders afterwards. In those two seasons he had the highest touches per minute of his career along with the lowest FG% of his career (ignoring his rookie season). Other players may be different but for Ben Wallace it appears that his percentages dropped as his scoring attempts increased.
Back to top

bchaikin



Joined: 27 Jan 2005
Posts: 687
Location: cleveland, ohio
Posted: Sun Dec 10, 2006 4:33 pm Post subject:

Ben Wallace was given a much larger offensive role under Larry Brown in 03-04 and 04-05 than he had under Rick Carlisle previously or Flip Saunders afterwards. In those two seasons he had the highest touches per minute of his career along with the lowest FG% of his career (ignoring his rookie season). correct - but did you also notice that in those two seasons with his highest touches/min he also had turnovers per touch lower than in all of his other seasons combined? his touches/min those two seasons was at 0.64, but at just 0.49 the rest of his career (that's about a 23% increase). so while his FG% decreased with an increase in touches, his rate of turnovers per touch decreased (rather than increased) with an increase in touches.... Other players may be different but for Ben Wallace it appears that his percentages dropped as his scoring attempts increased. correct - but again i would think that the key is to look at how all players scoring efficiency and turnover rates change with an increase in touches or possessions before proclaiming something an APBR tenet...
Back to top

Ben



Joined: 13 Jan 2005
Posts: 264
Location: Iowa City
Posted: Mon Dec 11, 2006 11:20 am Post subject:

Hi Guys, I maintain that if 5 players were blindfolded with 1 arm tied behind their backs, they would suffer no loss in offensive efficiency. I know this is a minority position - in fact, I think I am the only one who maintains it. Nonetheless, I think the burden is on the rest of you to prove me wrong. And I also think you should do it on this unrelated thread. My other stipulation is that since this is a stats board, you aren't allowed to use any theory, logic or common sense - just give me empirical STATS-based evidence that my claim is wrong. Thanks in advance, Ben
Back to top

Analyze This



Joined: 17 May 2005
Posts: 364
Posted: Mon Dec 11, 2006 3:32 pm Post subject:

Ben, is it the left arm or the right one? My simulation programm needs to know that._________________Where There's a WilT There's a Way
Back to top

kjb



Joined: 03 Jan 2005
Posts: 865
Location: Washington, DC
Posted: Mon Dec 11, 2006 4:56 pm Post subject:

bchaikin wrote:
Of course, that lineup doesn't really address the question being posed here. Daniels is efficient, although he doesn't shoot well. over the past four seasons the average ScFG% for PGs has been 51.1%. daniels has a ScFG% of 55.1% during this same time. of the 39 PGs to have played at least 5000 total minutes those four years, that's the 4th highest ScFG%. his eFG% was at just the average of these 39 PGs, so yes while he may not shoot well, his overall scoring efficiency is very much on the high side. in the past two seasons daniels has gotten fouled more per touch than all PGs other than iverson and arenas (is either one really a PG?), and he shot 83% from the FT line. that's why his ScFG% is so high.... top this off with the fact that his rate of turnovers per touch is absolutely one of the best/lowest of all PGs the past 4 years, and so his points scored per zero point team possession during this time has been the best of all PGs, better than even billups or nash.... doesn't mean he's a better PG (he's not) than either of those two, but it does mean he's very good, and certainly capable of more touches without a decrease in efficiency, either scoring wise or turnover wise...
I was referring to Daniels' shooting from the field, which has been pretty feeble the past couple seasons. His excellent free throw shooting is a good thing. I wonder, though, whether he'd continue to get fouled as often per touch as he does with few touches as he would with what would have to be increased touches as he carried a heavier offensive burden. And, why wouldn't teams start to defend him differently as he carries more of an offensive load? As a #1 option offensively, defenses would focus on him more, and would probably do things to limit his driving opportunities and force him to take more jumpers. And, given his .289 efg on jumpers this season, that wouldn't be a good thing for Daniels or that hypothetical lineup. Quote:
Shhhhh -- don't give Eddie ideas. (shhhh - wonder if eddie realizes this about daniels? i won't tell him if you won't, they're in the same conference as the cavs)....
Excellent use of the "quote out of context to make the other guy's point look silly" technique. I was, of course, talking about using that entire 5-man lineup you'd mentioned, not Daniels as an individual. I've been in favor of playing Daniels more, primarily because it would mean less Stevenson (although DeShawn has been playing pretty well (for him) lately.) Quote:
Stevenson and Hayes are not efficient players. Haywood isn't really much of a post scoring threat -- he does better getting his points from the offensive glass. Etan is a decent post scorer, but with weak offensive teammates and a weakness passing the ball back out of the post when doubled, there are question marks with him too. agreed - but the question is would they be even less efficient with more touches, either shooting worse or increasing their rate of turnovers per touch? i just don't see any evidence to suggest either would happen... stevenson shot his career best ScFG% last season, but as of last sunday (16 games played) he was shooting even better than that overall, and with more touches/min, plus less turnovers per touch. hayes was shooting his worst, but with his lowest turnovers per touch and highest touches/min. both haywood and thomas have high rates of turnovers per touch, such that you most likely would not want to increase their touches. but whether their turnovers per touch would increase with an increase in touches/min i see no evidence for this from their statistical histories...
Stevenson has been inefficient offensively throughout his career. Ditto for Hayes. I don't see any reason to think they'd become more efficient with increased touches, although I do see reasons why they'd become less efficient. Stevenson, for example, gets mostly wide-open mid-range jumpers and layups because defensive attention is focused so heavily on Arenas, Jamison and Butler. He's getting the easiest possible shots, and to his credit, he's making them. The few times when he attempts to create shots is generally pretty ugly. What's the likelihood of him continuing to get wide-open mid-range jumpers and un/lightly contested layups when primary offensive options are on the bench? I think what we'd see with Stevenson would be a drop-off in efficiency -- unless somehow the team could continue creating for him the easy looks he's getting now. By the way, I know you've mentioned it before, but what is the process for estimating touches? When was the last time you updated the formula? When was the tracking done upon which the estimates are constructed?
Back to top

Crow
Posts: 6146
Joined: Thu Apr 14, 2011 11:10 pm

Re: Wins Produced - Wages of Wins (Berri, Schmidt, and Brook

Post by Crow » Fri Apr 15, 2011 8:19 pm

DLew



Joined: 13 Nov 2006
Posts: 224
Posted: Wed Dec 13, 2006 7:41 pm Post subject:

Getting back to the point of this thread, I think there may be a major flaw with the entire premise of the book. The premise is that in all sports payroll does not correlate very strongly with wins. In baseball this is understandable because players are inconsistent. In basketball, players are more consistent in their performance (performance measured by Wins Produced), so if decision makers are making good decisions then payroll should correlate more strongly with wins, which it does not. Therefore decision makers are not making good decisions and they should be using a better metric to evaluate players. This is the central argument of the book. Aside: Berri has also argued that his metric is good because it is consistent across time. This is a stupid argument because it is easy to construct a model that is more predictive of itself in year N+1 then Berri's, PER for example is more predictive. Anyway, Adjusted Plus-Minus ratings tell us a player's effect on the team adjusted for who he played with and against. These numbers are less consistent year to year than Wins Produced or PER (or pretty much any box score stats based metric). This is probably partially due to the cross player effects that are still not entirely accounted for. I would postulate, however, that the larger reason for the lesser year to year correlation of Adjusted Plus-Minus is because it is measuring performance accurately, and actual performance, the way a player effects his team, is more variable than his box score statistics. Essentially, using a better measure of player performance we find that basketball players are not as consistent as Berri thinks, and therefore basketball decision makers are not nearly as bad as he says. I don't believe that NBA decision makers are perfect, but Berri's conclusion that basketball players are consistent is based on his own metric, which may or may not be right, and then based on that he concludes that decision makers need his metric to do better. That is circular logic for you.
Back to top

bchaikin



Joined: 27 Jan 2005
Posts: 687
Location: cleveland, ohio
Posted: Wed Dec 13, 2006 11:59 pm Post subject:

I wonder, though, whether he'd continue to get fouled as often per touch as he does with few touches as he would with what would have to be increased touches as he carried a heavier offensive burden. there are a number of examples of PGs getting fouled more per touch with an increase in touches. considering daniels already gets fouled as much per touch as players/PGs like iverson and arenas, and much more per touch than the vast majority of PGs - and not just this season but each of the last three - i certainly don't see why he wouldn't. just two seasons ago, and with about 27% more touches/min, he was still getting fouled more per touch than about 90% of the league's PGs... And, why wouldn't teams start to defend him differently as he carries more of an offensive load? i don't know - why wouldn't they? three seasons ago he had 37% more touches/min, yet shot even better with a ScFG% of 58%, when he was scoring 8.0 pts/g playing just 21 min/g versus 7.7 pts/g now playing 25 min/g... gilbert arenas is shooting better now than he did three seasons ago with the wizards (but similar to how he shot his rookie season), yet he is getting 20% more touches/min and getting fouled more per touch. are defenses defending him differently now versus how they did three seasons ago? if so how has he been able to increase his shooting and rate of getting fouled per touch with a "..heavier offensive burden..."?... As a #1 option offensively, defenses would focus on him more, and would probably do things to limit his driving opportunities and force him to take more jumpers. are defenses doing this to arenas?... And, given his .289 efg on jumpers this season, that wouldn't be a good thing for Daniels or that hypothetical lineup. i don't know his jumpers' percentage the last few seasons but he still got fouled a ton compared to the majority of other PGs (who probably shot better on jumpers than he) with increased touches... Stevenson has been inefficient offensively throughout his career. agreed - but as of last sunday he was getting his highest touches/min in five seasons yet while shooting a career best ScFG% of 55.2%.... Ditto for Hayes. this is only hayes' fourth season in the league, and each of the first three seasons he increased his ScFG% each year (although his touches/min have been relatively constant)... I don't see any reason to think they'd become more efficient with increased touches... some players do, some don't, there is no hard and fast rule... again hayes is young, and despite being in the league for now his 7th season stevenson is only 25 years of age. most players show an increase in touches when young, after their first few seasons (this is a general trend), and some increase their scoring efficiency while some do not... ...although I do see reasons why they'd become less efficient. Stevenson, for example, gets mostly wide-open mid-range jumpers and layups because defensive attention is focused so heavily on Arenas, Jamison and Butler. He's getting the easiest possible shots, and to his credit, he's making them. The few times when he attempts to create shots is generally pretty ugly. What's the likelihood of him continuing to get wide-open mid-range jumpers and un/lightly contested layups when primary offensive options are on the bench? I think what we'd see with Stevenson would be a drop-off in efficiency... perhaps most would think this too. but there are a number of example of players being as efficient or more efficient with an increase in touches/min, an increase in playing time. a good example is when there are expansion teams, as i mentioned earlier, as when tony campbell and terry catledge went from reserves to starters, increased their touches (by about 20% each), increased their playing time, and maintained their career ScFG%s (while being their team's #1 or #2 option on offense when they never had been before)... when kurt rambis went from the lakers to the expansion charlotte hornets, his touches/min increased 25% over his career touches/min on the lakers to a career high, and although he shot slightly worse overall, his turnovers went down considerably per touch despite the increase in touches/min such that his points scored per zero point team possession was better in charlotte than in seven seasons with the lakers... By the way, I know you've mentioned it before, but what is the process for estimating touches? touches = shots + passes + turnovers + times fouled = FGA + AST/X + TO +FTA/Y where X and Y are calculated yearly (or in season weekly) and by team (pace factors)... When was the last time you updated the formula? every week during a season... When was the tracking done upon which the estimates are constructed? many years ago.... i spot check it pretty much whenever i watch a game from start to finish...
Back to top

Ben



Joined: 13 Jan 2005
Posts: 264
Location: Iowa City
Posted: Thu Dec 14, 2006 12:04 pm Post subject:

For those who are used to different numbers, when the (carefully selected) Kurt Rambis went from LA to Charlotte, his Usage rate went up 17% and his Offensive rating dropped 5% from the year before.
Back to top

Eli W



Joined: 01 Feb 2005
Posts: 402
Posted: Tue Jan 16, 2007 2:39 pm Post subject:

An interesting economics-based critique of Wages of Wins: Quote:
In his book, Berri argues that player productivity can be predicted, with sufficient certainty, on the basis of past performance. Using data from 1994-2004, he shows that Win Scores from the previous season are highly correlated with current performance (R-sq = .70). The problem with this line of reasoning is that “Win Score” as such doesn’t actually exist, but is instead a composite of various inputs. Now, suppose we distinguish between scoring input (points – true shot attempts), and non-scoring input (rebounds + steals - turnovers….). When we reexamine Berri’s 1994-2004 data, we find that non-scoring production is almost perfectly consistent from year to year (R-sq = .85). Yet when we consider scoring production alone, the correlation between past and current performance is much, much weaker (R-sq = .30). Thus, contrary to the assumptions of Berri’s economic model, the ability of GMs to accurately “price” a players’ scoring production is a highly uncertain (i.e. risky) business.
http://freedarko.blogspot.com/2007/01/d ... ience.html
Back to top

Harold Almonte



Joined: 04 Aug 2006
Posts: 616
Posted: Tue Jan 16, 2007 5:48 pm Post subject:

Does it mean that: Scoring Stats (assisted-unassisted-and all the process to convert a possession to points) vs. Non Scoring Stats (isolate defense-team defense-and all the process to obtain a possession, or avoid points) must be adjusted for gambling?, for risky and risk-less way to produce the basketball goal (a positive points differential)?, for skill dificulty? Is it height overpaid?

Page 18

Author Message Harold Almonte



Joined: 04 Aug 2006
Posts: 616
Posted: Wed Jan 17, 2007 5:32 pm Post subject:

Maybe the meaning is that those vintage linear formulas strongly penalize scoring by failed shots (like if scoring is the easiest thing of basketball, and all bad decissions were forced by the scorer), turnovers and offensive fouls; but defense is penalized by defensive fouls only (but not every foul is a shooting foul). But offense is not rightly rewarded by passing vision, by freezing a defender, initiate the offense, create plays (a successfull drive is almost like a player giving an assist to himself). Defense stats are not penalized by opponents offensive rebounds, second chances, points allowed(defensive equivalent to a failed shot), but is not rightly rewarded by taking charges, contested opponent failed shots, intimidation and help. But stars are not rewarded by clutch responsabilities, leadership and coachs' bad decissions. More of the same old stuff. To keep analizing linear ratings is right now a loose of time. There is no enough data, the logic is imcomplete, adjusts are forced. I think when a rating uses "ratio" and "%", they are penalizing every stat by gambling inside the efficiency, but are every stat equally penalized by risk? If we know a 50%FG scorer is an awesome scorer (it depends of the position), then we know that scoring has an implicit risk of about 50% to fail at the best performance (let us not talk about passing risks to contested vs. to non contested shots, long vs. close). If scoring is about 50% of the game, and defense the other 50% (the risk of defense is inside the points allowed, but not as risky as scoring, or is it?), or are the games won only at the defensive end? why don't penalize (maybe in an exponential way) scorers if they perform under an average position-PPP? What if the risky passer share the penalization? Why if an assisted scoring is about 2/3 of the scoring possession value shared with the passer, can't rebounds be 2/3 of the defensive possession value shared with the player who defended the failed shot (if there is someone)? Regressions certainly tell us that defensive stats that imply to take the ball from the opponent contribute strongly on wins, but defense must be penalized with points allowed (defensive attempts, failed to avoid scoring or to take the ball), is not just a lot of positive boxcore stats and only one negative (fouls), and so, some named role players will look as bad as unefficient scorers and viceversa, and players ranking will be rearranged.
Back to top

asimpkins



Joined: 30 Apr 2006
Posts: 245
Location: Pleasanton, CA
Posted: Tue May 01, 2007 12:03 pm Post subject:

bchaikin wrote:
if tomorrow gilbert arenas, antawn jamison, and caron butler all slip on the same bar of soap in the shower, and are all out for the next month, and your starting five for that time is antonio daniels, deshawn stevenson, jarvis hayes, etan thomas, and brendan haywood - what would happen? not record wise but in terms of shooting efficiency and turnovers. can you say definitively that either the shooting efficiency would decrease, or the turnovers would increase, because all of their touches (or touches/min) would increase above what they are now?...
I came across this in this old thread which turned to a usage-efficiency debate -- this is quoted from a post on page 16. This pretty much ended up happening (except for Jamison and the soap). Does anyone have the numbers? Any comments? Are the results what you expected bchaikin?
Back to top

Ben F.



Joined: 07 Mar 2005
Posts: 391
Posted: Tue May 01, 2007 12:30 pm Post subject:

asimpkins wrote:
bchaikin wrote:
if tomorrow gilbert arenas, antawn jamison, and caron butler all slip on the same bar of soap in the shower, and are all out for the next month, and your starting five for that time is antonio daniels, deshawn stevenson, jarvis hayes, etan thomas, and brendan haywood - what would happen? not record wise but in terms of shooting efficiency and turnovers. can you say definitively that either the shooting efficiency would decrease, or the turnovers would increase, because all of their touches (or touches/min) would increase above what they are now?...
I came across this in this old thread which turned to a usage-efficiency debate -- this is quoted from a post on page 16. This pretty much ended up happening (except for Jamison and the soap). Does anyone have the numbers? Any comments? Are the results what you expected bchaikin?
Quoting from what I wrote in the "diversified offense" thread: Ben F. wrote:
As devastating as the injuries to Caron Butler and Gilbert Arenas have been to me as a Wizards fan, they do provide a good test case for what happens when a team's two top usage players go out. Just watching this team you know that they're struggling to get quality shots. In their last 10 games (the ones without both Arenas and Butler) the Wizards went 2-8. They played 4 games against 3 different teams who were out of the playoffs, and went 2-2. They played 6 games against 5 teams who were in the playoffs and fighting for position, and went 0-6. Where was the problem? Intuition would say it's on the offensive end - as much as Butler and Arenas do bring something defensively, their primary contributions are offensive. The numbers say that over those last 10 games, the Wizards gave up 112.8 points per 100 possessions, while scoring only 102.9. Both numbers are quite bad, but the Wizards on the year gave up 112.5 points per 100 poss and scored 111.1. Clearly there was a huge dropoff in offensive production, as expected. According to those who theorize that without star players, role players should play about the same, this shouldn't have been an issue. DeShawn Stevenson was shooting above 50% eFG% - surely he could pick up the offensive slack, with no drop in efficiency? It's not like he was getting better looks because the defense was keying on Arenas and Caron. Besides, player performance is fairly constant regardless of context. [This is tongue in cheek]. DeShawn Stevenson's year long totals were 51% eFG% and 54.2% TS% both with a usage of 16.8. DeShawn Stevenson's totals in that 10 game stretch were 43.4% eFG% and 47.2% TS% both with a usage of 24.3. If you've watched the Wizards you know that a lot of DeShawn's shots come from others creating those opportunities for him. He's been very consistent this year converting on those opportunities, and that has certainly helped the Wizards. But nobody who had seen the Wizards play would have said that DeShawn could start to create for himself, and pick up the offensive slack without a drop in efficiency. He just hasn't been that kind of player. And it's borne out in the numbers. His usage drastically increased, his efficiency drastically decreased.

Back to top

kjb



Joined: 03 Jan 2005
Posts: 864
Location: Washington, DC
Posted: Tue May 01, 2007 12:49 pm Post subject:

kjb wrote:
Shhhhh -- don't give Eddie ideas. He's already nutty enough with his lineups. That lineup would have problems scoring the ball. Daniels is a solid PG, but I don't think he could be a lead scorer. With good finishers, I think he's a terrific creator. He's good running screen/roll, he's good at getting to the FT line. But, he's not much of a shooter. And, as a passer, he'd need finishers. Neither Stevenson nor Hayes shoot the ball well. Nor do they finish well in traffic. Stevenson would probably become less efficient -- higher usage has coincided with lower offensive efficiency during his career. Last season was the lowest usage of his career and his best shooting. This season, his usage is a touch lower, and his efg is the best of his career. Ask him to create offense, and his efg will drop -- seriously, it's borderline embarrassing to watch him try. Hayes would likely try to pick up the scoring slack -- he's the one guy from that lineup who'd be willing to put up the shot. Unfortunately, despite a nice-looking stroke, his shot doesn't go down enough. He looked to be making strides last season, but 2 kneecap fractures in 2 years haven't helped. The team could go inside to Thomas and Haywood more. That would probably be the best bet, particularly with Thomas in the post. However, Thomas has been poor at passing out of the post when doubled, as has Haywood. And Haywood has iffy footwork on his post moves and gets called for traveling with some frequency. Of course, that lineup doesn't really address the question being posed here. Daniels is efficient, although he doesn't shoot well. Stevenson and Hayes are not efficient players. Haywood isn't really much of a post scoring threat -- he does better getting his points from the offensive glass. Etan is a decent post scorer, but with weak offensive teammates and a weakness passing the ball back out of the post when doubled, there are question marks with him too.
That's what I wrote in response. My prediction was accurate about Stevenson. I don't have time right now to look up what happened to Daniels and others. Jarvis (if I recall correctly) shot the ball better, though I don't know if his usage went up significantly. The team assuredly did not go inside to Haywood and Etan, especially since Haywood was on the bench most of the time.Last edited by kjb on Tue May 01, 2007 1:00 pm; edited 1 time in total
Back to top

gabefarkas



Joined: 31 Dec 2004
Posts: 1313
Location: Durham, NC
Posted: Tue May 01, 2007 12:57 pm Post subject:

asimpkins wrote:
bchaikin wrote:
if tomorrow gilbert arenas, antawn jamison, and caron butler all slip on the same bar of soap in the shower, and are all out for the next month, and your starting five for that time is antonio daniels, deshawn stevenson, jarvis hayes, etan thomas, and brendan haywood - what would happen? not record wise but in terms of shooting efficiency and turnovers. can you say definitively that either the shooting efficiency would decrease, or the turnovers would increase, because all of their touches (or touches/min) would increase above what they are now?...
I came across this in this old thread which turned to a usage-efficiency debate -- this is quoted from a post on page 16. This pretty much ended up happening (except for Jamison and the soap). Does anyone have the numbers? Any comments? Are the results what you expected bchaikin?
I missed Bob's comment the first time around, but I wanted to respond to it. In all honesty, I'm not 100% convinced there is a definitive inverse relationship between usage and efficiency. I think it can vary from player to player. So Bob, let me ask you something. Without regard to the specific nature of the relationship (inverse, linear, quadratic, whatever), do you think that an underlying relationship exists, for each player, between touches/min and efficiency? Between touches/min and turnovers? Furthermore, could this relationship be modeled, on a per-player basis? Or, is each game one giant stochastic process? Or, perhaps a relationship does exist, but there is so much intrinsic variability involved that we wouldn't be able to hear anything above the noise?
Back to top

bchaikin



Joined: 27 Jan 2005
Posts: 686
Location: cleveland, ohio
Posted: Tue May 01, 2007 1:02 pm Post subject:

I came across this in this old thread which turned to a usage-efficiency debate -- this is quoted from a post on page 16. This pretty much ended up happening (except for Jamison and the soap). Does anyone have the numbers? Any comments? Are the results what you expected bchaikin? in the 11 games antonio daniels played in since arenas went out, i have his touches/min increasing by 30% from what they were after about 70 games. he played 40 min/g and got 9.5 ast/g with just 2.0 TO/g. you can count on one hand the number of players to have averaged 9 ast/g for a season with just 2 TO/g. granted he did it for just 11 games, but by my numbers his turnovers per touch were at a very low 3% for those 11 games, exactly what they were after 70 games. so with a large increase in touches/min there's no increase in turnovers per touch for daniels in this short 11 game span... daniels had a ScFG% of 53.0% in these 11 games, and was at 56.4% the 70 games prior. but in all of 05-06 he was at a ScFG% of 53%, and in all of 04-05 he was at 54%, so its hard to call that drop due to an increase in touches, or due to the absence of arenas and/or butler... jamison was scoring 19.8 pts/g on a ScFG% of 53.7% after 70 games. in the 12 games since with arenas out, he's scored 25.7 pts/g on a ScFG% of 53.5%, so there's little difference there. i have jamison's touches/min increasing by just 7%-8% in those 12 games versus his first 70 games, and his rate of turnovers per touch at 5% both during those 12 games and the 70 games prior... jarvis hayes played 18 min/g prior to the departure of arenas, 33 min/g after, but i have his touches/min as being the same. his ScFG% increased to 51.3% in those 12 games versus 48.6% in 71 games prior, and his turnovers per touch dropped significantly (he had only 7 turnovers in those 12 games)... I'm not 100% convinced there is a definitive inverse relationship between usage and efficiency. the data certainly does not show that there is, at least in a general sense... do you think that an underlying relationship exists, for each player, between touches/min and efficiency? Between touches/min and turnovers? for each player? not sure how to respond to this - the data for each player is obviously unique for that player. not sure what conclusions can be drawn from one group of players that say do not see a difference and another group that does, unless you can show some similarities between the groups of players...Last edited by bchaikin on Tue May 01, 2007 1:41 pm; edited 3 times in total
Back to top

HoopStudies



Joined: 30 Dec 2004
Posts: 705
Location: Near Philadelphia, PA
Posted: Tue May 01, 2007 1:04 pm Post subject:

kjb wrote:
kjb wrote:
Shhhhh -- don't give Eddie ideas. He's already nutty enough with his lineups. That lineup would have problems scoring the ball. Daniels is a solid PG, but I don't think he could be a lead scorer. With good finishers, I think he's a terrific creator. He's good running screen/roll, he's good at getting to the FT line. But, he's not much of a shooter. And, as a passer, he'd need finishers. Neither Stevenson nor Hayes shoot the ball well. Nor do they finish well in traffic. Stevenson would probably become less efficient -- higher usage has coincided with lower offensive efficiency during his career. Last season was the lowest usage of his career and his best shooting. This season, his usage is a touch lower, and his efg is the best of his career. Ask him to create offense, and his efg will drop -- seriously, it's borderline embarrassing to watch him try. Hayes would likely try to pick up the scoring slack -- he's the one guy from that lineup who'd be willing to put up the shot. Unfortunately, despite a nice-looking stroke, his shot doesn't go down enough. He looked to be making strides last season, but 2 kneecap fractures in 2 years haven't helped. The team could go inside to Thomas and Haywood more. That would probably be the best bet, particularly with Thomas in the post. However, Thomas has been poor at passing out of the post when doubled, as has Haywood. And Haywood has iffy footwork on his post moves and gets called for traveling with some frequency. Of course, that lineup doesn't really address the question being posed here. Daniels is efficient, although he doesn't shoot well. Stevenson and Hayes are not efficient players. Haywood isn't really much of a post scoring threat -- he does better getting his points from the offensive glass. Etan is a decent post scorer, but with weak offensive teammates and a weakness passing the ball back out of the post when doubled, there are question marks with him too.
That's what I wrote in response. My prediction was accurate about Stevenson. I don't have time right now to look up what happened to Daniels and others. Jarvis (if I recall correctly) shot the ball better, though I don't know if his usage went up significantly. The team assuredly did not go inside to Haywood and Etan, especially since Haywood was on the bench most of the time.
Only Hayes looks like he maintained his numbers. Most everyone, including Jamison and Daniels, used more possessions and was less efficient, looking at the metrics in BoP and games from 4/4-4/18. Jamison did get to the foul line more, making up partially for other drop offs. Pretty good natural experiment. Someone should project what the team ORtg should have been if all had maintained their efficiency (maybe just use the minutes played from 4/4 after vs prior to). (For newcomers, this debate has been held before. If you use sophisticated stat methods -- not all that sophisticated, but more than basic -- you see a downward trend to efficiency with increased usage, using efficiency and usage as defined in Basketball on Paper. Because there can be no strict isolation of usage and efficiency without changes in other things -- like age, complementary teammates, etc. -- experiments are difficult to carry out. Some of us are working on ways to better test the phenomenon.)_________________Dean Oliver Author, Basketball on Paper The postings are my own & don't necess represent positions, strategies or opinions of employers.Last edited by HoopStudies on Tue May 01, 2007 6:16 pm; edited 2 times in total
Back to top

Ben F.



Joined: 07 Mar 2005
Posts: 391
Posted: Tue May 01, 2007 1:14 pm Post subject:

Gabe I think there's no question that the relationship has to vary by player. Some players are better at creating shots for themselves than other players. Antonio Daniels is actually MORE comfortable with the ball in his hands, by my observation, as he scores most of his points driving to the basket and getting to the line. He was assisted on only 39% of his made shots last year which doesn't suggest that he depends too heavily on teammates. While I don't think this means he could be a top scorer in the NBA without a drop in efficiency, I think it does mean that he depends far less than someone like Stevenson on his teammates however.
Back to top

gabefarkas



Joined: 31 Dec 2004
Posts: 1313
Location: Durham, NC
Posted: Tue May 01, 2007 2:52 pm Post subject:

bchaikin wrote:
for each player? not sure how to respond to this - the data for each player is obviously unique for that player. not sure what conclusions can be drawn from one group of players that say do not see a difference and another group that does, unless you can show some similarities between the groups of players...
Perhaps you did not understand my question. Allow me to attempt to clarify. Let us assume, for the moment, that this hypothetical relationship is linear in nature, and can be modeled as follows: Yi = Ai + (Bi * Xi) + Ei where "Y" is the scoring efficiency, "X" is the touches per unit of time, "A" is an intercept term, "B" is a coefficient for the for parameter X, and "E" is an error term representing background noise or intrinsic variability. Ei can be defined as a normally distributed random variable with mean 0. Here, let "i" represent the player involved, so that the values of Ai, Bi, and Ei are specific and relevant only to an individual player. My question, in parts, is as follows: First, do you believe that this could (but not necessarily "would") be an adequate model for the relationship between scoring efficiency and touches per unit of time? Secondly, could Bi take on any value from the set of real numbers, either positive or negative, for each "i" player? This implies that either a positive or negative correlation exists, depending on the player in question. Or, do you believe that the range of possible values for Bi would be further constrained in any way (eg, necessarily > 0, or < 0, or between -10 and 10, etc)? Thirdly, assuming a relationship of this sort exists, do you believe the values of Ei, which encapsulate many external influences on Yi (such as teammates on the floor, pace of game, opponents faced), are likely so large as to either mask the true nature of the relationship, or render such a model useless? I look forward to your response.
Back to top

bchaikin



Joined: 27 Jan 2005
Posts: 686
Location: cleveland, ohio
Posted: Tue May 01, 2007 8:12 pm Post subject:

you've presented what looks to me to be a simple equation for a straight line. if the relationship is truly linear (which i do not know if it is), then i guess it can work... the stats database at www.apbr.org or www.bballsports.com has touches/min, ScFG%s, and rates of turnovers per touch for all players, if you are looking for data points to test this...
Back to top

gabefarkas



Joined: 31 Dec 2004
Posts: 1313
Location: Durham, NC
Posted: Tue May 01, 2007 10:35 pm Post subject:

bchaikin wrote:
you've presented what looks to me to be a simple equation for a straight line. if the relationship is truly linear (which i do not know if it is), then i guess it can work...
That is correct! I have presented a simple linear model, yearning for regression. Oftentimes, it is the first step in determining whether or not a functional relationship exists. You have answered the first part of my inquiry, and I look forward to your answers to the rest of it. To reiterate, in your opinion would you envision any constraints on the values of Ai and Bi arising in any way? For instance, would Bi always be positive, negative, centered around zero, centered around some other positive or negative number, or any other similar constraints? Could Bi be a random function? Furthermore, are there any other considerations, in your opinion, that merit specification in the model? Possible candidates that I have thought of could include teammates on the floor, pace of game, or opponent faced. bchaikin wrote:
the stats database at www.apbr.org or www.bballsports.com has touches/min, ScFG%s, and rates of turnovers per touch for all players, if you are looking for data points to test this...
I appreciate the suggestion, but I'm not interested in doing anyone else's heavy lifting, as I've got enough of my own work to do. Perhaps in a few weeks or months. This seems to be a subject of considerable interest to you, as I've noticed you are often involved in discussions involving efficacy and usage. Frankly, it's a subject of interest to me too. So, that is why I'm trying to engage you in a discussion about it here, to attempt to draw out what your actual opinion is on the matter. As a graduate of CWRU, a school that consistently ranks among the best private institutions of higher learning in the US, I believe any insights you could share on this matter would be beneficial, and dare I say promethean, to the advancement of APBRmetrics. You've obviously given it much thought (if you want stats to support that statement just search for posts with "touches" in them and yourself as the author) (screw it, I'll do it for you: 157 of your 485 posts in this forum contain the word in them, that's roughly 1/3 of everything you've ever written on these boards). However, I have to admit, I can't help but feel a little bit of frustration in this regard. At least from what I've seen recently, your responses to others and myself seem elusory, if not puzzling. For instance, I asked you a question in this thread that I felt was fairly straightforward:Quote:
So Bob, let me ask you something. Without regard to the specific nature of the relationship (inverse, linear, quadratic, whatever), do you think that an underlying relationship exists, for each player, between touches/min and efficiency? Between touches/min and turnovers? Furthermore, could this relationship be modeled, on a per-player basis?
Your response only addressed the question tangentially at best. Then, I explained my question again, this time in what I felt was a more illustrative manner. However, once again you cherry-picked my post and responded to only 22 words (by my count). Thus, I'm left grappling with several possible conclusions: 1) Your general reading comprehension skills are not what I had previously assumed 2) You did not understand the other questions, in which case I would hope you would ask for further clarification, since this is a discussion board 3) You already have concluded an answer on your own that is above and beyond the current scope of this discussion, but due to contractual obligations or non-disclosure agreements are prevented from discussing it in this forum 4) You are unsure what the answer could be, and hope that others will take a stab at it, and share their results 5) You find the general level of discourse and my approach to be below yourself 6) You firmly believe that no relationship exists and that touches/min has no effect on scoring efficiency or turnover rate; additionally you feel the onus is on others to disprove or discredit this, or #3 is involved in some way to preclude you from stating this outright I can't come up with anything else to explain what's going on. If the truth is indeed #1, then I'm sorry for you. If it's #2, please don't hesitate to ask for further clarity, and I will be glad to oblige. If it's #3, then good for you, kudos, I can completely respect and understand that. If it's #4, I would say that it is a puerile approach to problem solving. If it's #5, please indicate this, and I will refrain from engaging you in discussions of this nature again. And finally, if it's #6, I think it would only be fair for you to share a comprehensive, rigorous proof, if you believe one exists and are at liberty to disclose it, and I will happily believe you.
Back to top

bchaikin



Joined: 27 Jan 2005
Posts: 686
Location: cleveland, ohio
Posted: Wed May 02, 2007 12:57 am Post subject:

this is what you asked: Without regard to the specific nature of the relationship (inverse, linear, quadratic, whatever), do you think that an underlying relationship exists, for each player, between touches/min and efficiency? for each player - again, i don't know, each is a unique case. if you had asked for all players, i would have asked you what your definition of efficiency was - ScFG%, points per zero point team possession, something else?... Between touches/min and turnovers? again for each player i don't know, you'd have to look at the data for each player specifically. but for all players as a single grouping there is a general trend of less turnovers per touch with increasing touches/min... Furthermore, could this relationship be modeled, on a per-player basis? i'm sure it could... I'm not interested in doing anyone else's heavy lifting, as I've got enough of my own work to do then what it is you are asking? are you supposing a linear relationship, but don't wish to show it? i'm saying i don't see evidence for a general decrease in efficiency with increased touches/min, but i am not saying their is a linear relationship. are you saying there is?... you are often involved in discussions involving efficacy and usage. efficacy?.... As a graduate of CWRU... go Spartans!... a school that consistently ranks among the best private institutions of higher learning in the US with a price tag to match... i think i'm still paying for it decades later... those low interest student loans seem to last forever... I believe any insights you could share on this matter would be beneficial, and dare I say promethean, to the advancement of APBRmetrics. * sniff * You've obviously given it much thought they are of import for the simulation model... However, I have to admit, I can't help but feel a little bit of frustration in this regard because you don't want to do the work yourself showing there is a linear relationship?... At least from what I've seen recently, your responses to others and myself seem elusory, if not puzzling. always keep 'em guessing... For instance, I asked you a question in this thread that I felt was fairly straightforward:..... Your response only addressed the question tangentially at best. see above... once again you cherry-picked... only way i could score... 1) Your general reading comprehension skills are not what I had previously assumed the senior moments are becoming all too frequent.... 2) You did not understand the other questions, in which case I would hope you would ask for further clarification, since this is a discussion board see above... 3) You already have concluded an answer on your own that is above and beyond the current scope of this discussion, but due to contractual obligations or non-disclosure agreements are prevented from discussing it in this forum i should be so lucky... 4) You are unsure what the answer could be, and hope that others will take a stab at it, and share their results i told you i don't know if there is a linear relationship, i simply know the evidence does not show a decrease in efficiency with an increase in touches/min... 5) You find the general level of discourse and my approach to be below yourself no, just confusing... 6) You firmly believe that no relationship exists and that touches/min has no effect on scoring efficiency or turnover rate; additionally you feel the onus is on others to disprove or discredit this, or #3 is involved in some way to preclude you from stating this outright see above... I can't come up with anything else to explain what's going on. If the truth is indeed #1, then I'm sorry for you. thanks - dementia is no laughing matter... If it's #2, please don't hesitate to ask for further clarity, and I will be glad to oblige. see above... If it's #3, then good for you, kudos, I can completely respect and understand that. i don't get your respect and understanding if it's not 3? rats... If it's #4, I would say that it is a puerile approach to problem solving. puerile??.... If it's #5, please indicate this, and I will refrain from engaging you in discussions of this nature again. what? and miss out on all this fun?... And finally, if it's #6, I think it would only be fair for you to share a comprehensive, rigorous proof, if you believe one exists and are at liberty to disclose it, and I will happily believe you. i'll take this under advisement... while you're on a roll, anything else you think would only be fair that i share?...
Back to top

DLew



Joined: 13 Nov 2006
Posts: 224
Posted: Wed May 02, 2007 1:36 am Post subject:

I think tht by this point we all know that it is hopeless to engage Bob in discourse on the usage-efficiency issue, you are not going to get good faith responses. I am a bit curious however, didn't Dean prove this pretty conclusively here: http://sonicscentral.com/apbrmetrics/vi ... y&start=30 It's possible there was some major problem with that study pointed out later that I don't know about, but if not I'm pretty sure we've resolved this. By the way, Bob, I don't think you are unreasonable in general, we have had discussions on other subjects that went fine, but when it comes to usage and efficiency you really go into Socrates mode.
Back to top

gabefarkas



Joined: 31 Dec 2004
Posts: 1313
Location: Durham, NC
Posted: Wed May 02, 2007 7:36 am Post subject:

bchaikin wrote:
for each player - again, i don't know, each is a unique case. if you had asked for all players, i would have asked you what your definition of efficiency was - ScFG%, points per zero point team possession, something else?... Between touches/min and turnovers? again for each player i don't know, you'd have to look at the data for each player specifically. but for all players as a single grouping there is a general trend of less turnovers per touch with increasing touches/min...
Semantics. bchaikin wrote:
3) You already have concluded an answer on your own that is above and beyond the current scope of this discussion, but due to contractual obligations or non-disclosure agreements are prevented from discussing it in this forum i should be so lucky...
Liar. Quote:
Bob Chaikin. NBA consultant and developer of B-Ball, Pro Basketball for the Computer, which is the most statistically accurate basketball simulation on the market today. Chaikin has worked for the New Jersey Nets and is currently employed by the Miami Heat. He compiled the All-Time Sports Statistics Encyclopedias found on the site and also helps with basketball rankings, stats and projections. He is a graduate of Case Western Reserve University.
Or perhaps you need to ask the people at Sandlot Shrink to kindly stop misrepresenting you. We're done here.
Back to top

bchaikin



Joined: 27 Jan 2005
Posts: 686
Location: cleveland, ohio
Posted: Wed May 02, 2007 9:36 am Post subject:

Liar. hmmm.... a puerile response perhaps?... Or perhaps you need to ask the people at Sandlot Shrink to kindly stop misrepresenting you. yep... it was on my recommendation the heat went out and got antoine walker... unfortunately that gig was close to a decade ago, my last season with them being the 97-98 season....
Back to top



Page 19 of 19
Author Message Harold Almonte



Joined: 04 Aug 2006
Posts: 616
Posted: Thu May 03, 2007 10:51 am Post subject:

Everybody wants to link usage and efficiency because economists (WOW among them) think that scoring efficiency and not usage is the measure of the scoring IQ and must be the measure of scoring Power. When they weighted all the basketball skills by regression, they forgot how to weight ballhandling (causes, they did it only with effects: TOs, assists), the main action of the game. The best ballhandlers not only touch the ball more time, they take almost all decissions (create) about which players will have, instead of themselves, the highest usage (who will score). Ballhandling-The capacity to carry the ball to the ring (scoring skill and productivity)-and Efficiency, overall decide the players usage. Efficiency is just a part not the whole, and that's why David Lee is not the NBA scoring leader, and Bibby has a higher usage than Kevin Martin. Lower scoring IQ a player has, a more difficult scheme you need to use for him to score, and overall that's against the economy of the game (at the offensive end).Last edited by Harold Almonte on Sat May 05, 2007 11:58 am; edited 3 times in total
Back to top

mtamada



Joined: 28 Jan 2005
Posts: 377
Posted: Fri May 04, 2007 4:25 am Post subject: Another persepective on team context and teamwork

Beyond the usage vs. efficiency debate is a broader debate about how player's performances will vary in different team contexts or while playing different team roles (PG vs SG, go-to scorer vs role player, etc.) Crew racing, while seemingly dependent only on how well a rower can row, also has a lot of team context issues; some seemingly inferior rowers can help their boat go faster, at least according to Cambridge University's crew team. http://chronicle.com/weekly/v53/i35/35a05601.htm (I think that article is available to non-subscribers.) Although I do believe that Alan Iverson is generally overrated by fans, players, and sportswriters, I also believe that his ability to generate those stats while at high usage levels does, in many cases at least, help make his teammates better, either by letting them concentrate on taking fewer and better shots, or by drawing the defense to him and giving his teammates more open shots period. I.e. I think he's a very good but not great player. The statistics show readily enough that he is not a great player -- but I think he's a bit better than the statistics show, for reasons of team context and teamwork.
Back to top

kjb



Joined: 03 Jan 2005
Posts: 865
Location: Washington, DC
Posted: Fri May 04, 2007 8:43 am Post subject:

Fascinating article. It makes me think of Haywood with the Wizards. Haywood is slow, awkward, uncoordinated, not very strong, gets pushed around, and sometimes just plain looks bad out on the court. Yet every year the Wizards are better when he's on the floor, especially on the defensive end.
Back to top

HoopStudies



Joined: 30 Dec 2004
Posts: 706
Location: Near Philadelphia, PA
Posted: Fri May 04, 2007 9:11 am Post subject: Re: Another perspective on team context and teamwork

mtamada wrote:
Beyond the usage vs. efficiency debate is a broader debate about how player's performances will vary in different team contexts or while playing different team roles (PG vs SG, go-to scorer vs role player, etc.) Crew racing, while seemingly dependent only on how well a rower can row, also has a lot of team context issues; some seemingly inferior rowers can help their boat go faster, at least according to Cambridge University's crew team. http://chronicle.com/weekly/v53/i35/35a05601.htm
The crew article seems to make the point that affability can overwhelm power and strength. There are some indications that this sort of psychological factor does matter and that it can be measured. thebbiq.com is going that way, for instance, in hoops. The usage vs efficiency debate is, as you suggest, underlain by more complex context- and coaching-related things. But it is a simple model of behavior that has been, in my experience, reliable in making predictions. It suggests interaction among players and coaches, something that people can't really debate exists and has some impact. The question of how large the impact is remains, but I'd suggest that the process of looking for how large that impact is and how exactly it manifests itself is one of the biggest contributions we can make. I'd like to think that we've already made headway toward understanding this._________________Dean Oliver Author, Basketball on Paper The postings are my own & don't necess represent positions, strategies or opinions of employers.
Back to top

Harold Almonte



Joined: 04 Aug 2006
Posts: 616
Posted: Sat May 05, 2007 12:47 pm Post subject:

Quote:
Although I do believe that Alan Iverson is generally overrated by fans, players, and sportswriters, I also believe that his ability to generate those stats while at high usage levels does, in many cases at least, help make his teammates better, either by letting them concentrate on taking fewer and better shots, or by drawing the defense to him and giving his teammates more open shots period.
Not only high usage level; the quality of his usage. The fact that he could manage to have average carreer stats (in terms of efficiency) taking a lot of bad, difficult and clutch shots (against or not his will, or coachs's will), just says good things about his talent. Quote:
The crew article seems to make the point that affability can overwhelm power and strength. There are some indications that this sort of psychological factor does matter and that it can be measured. thebbiq.com is going that way, for instance, in hoops. The usage vs efficiency debate is, as you suggest, underlain by more complex context- and coaching-related things. But it is a simple model of behavior that has been, in my experience, reliable in making predictions. It suggests interaction among players and coaches, something that people can't really debate exists and has some impact. The question of how large the impact is remains, but I'd suggest that the process of looking for how large that impact is and how exactly it manifests itself is one of the biggest contributions we can make. I'd like to think that we've already made headway toward understanding this.
Groupal theories and leadership conduct are applied to team sports. Some experts say Prouds/Narcicists Napoleonic leaders (A.I, Jordan, Kobe, Arenas, etc) are the best to reach success. Altruists (Garnett, Nowitzki) are not so good. Selfishs (some Knicks) are the worsts, and of course Humbles can't be leaders. High post offense would be good when you have a highly concentrated (post) power, Triangle when you have non-concentrated several powerful leaders, Motion and Princenton and distributed offense when you don't have a stronger leader personality. Every team would want to have the ideal Detroit, Suns and Spurs cooperativism, but players have their different ranges of egocentrism. It's a matter of fitting them like is done with skills and stats.

Post Reply