Berri Changes Value of Defensive Rebounds in WP

[Italian Stallion]

WinScore without different weights on scoring, rebounding, assists and a different approach to defense...it is another metric that I'd be curious to see results from but it might be better called your own as not that much major is the same as WinScore.

Point well taken.

The reason I quoted WinScore is that I started with that metric. Then over the last couple of years I started making some adjustments as I started understanding the reason some outlier cases were clearly wrong.

I assume Berri has changed DREB to .5 on Win Score also.

I'd be happy to post my top 25 players at some point, but I don't have an easy way to just sort and post them. I have a separate spread sheet for each team and would have to manually cut and paste them and then add comments.

[Crow]

As you like, just curious.

[Guy]

This is an interesting post by Berri: http://www.huffingtonpost.com/david-ber ... 49512.html. He argues that shot creation doesn't matter, because teams basically always take the same number of shots. What's fascinating about this is it demonstrates pretty clearly that Berri, even after 6 years of arguing, doesn't understand the debate he is engaged in. He simply has no idea what the discussion is about. Wherever one comes down on the usage-efficiency issue, I think everyone agrees that the alternative to having Kobe take a lot of shots is not to let the shot clock expire 10 times per game. And yet Berri seems to believe, or chooses to believe, that is what this debate is about. Very revealing, I think.....

[bchaikin]

I think everyone agrees that the alternative to having Kobe take a lot of shots is not to let the shot clock expire 10 times per game.

before claiming this as fact you might want to take a look at high scoring players who have switched teams during their careers, either via trade or free agency, and then look at what that team did after they left. charles barkley, allen iverson, grant hill, dominique wilkins, world b. free, adrian dantley, and a host of others all switched teams in their careers...

did all of their teams suffer from higher turnovers after they left? or did their previous team simply absorb all those touches amongst other players without an increase in turnovers?...

[Crow]

Not counting Free (because I wasn't sure which team departure you wanted to emphasize) 3 of the teams who had the others stars you mentioned saw team turnovers go down the year after the star left, 2 went up. No clear pattern but certainly not a consistent increase in turnovers.

[mystic]

Shot creation becomes not just an issue of turnovers, but also eFG%. Those star players are taking the last second shot in order to avoid a shot clock violation and are usually above league average at those last second shots. And those tough shots at the end of the shotclock are results of broken plays, not seldom caused by players who didn't take the open shot and passed it up. Players like Landry Fields are looking pretty good also, because they are not taking tough shots, but rather passing the ball to someone else who has to take the shot.

The names listed by bchaikin are also a bit random. Adrian Dantley for example was not that great at creating shooting opportunities under pressure. He was the guy melting the shot clock and then passing it up to his teammates who had to take the tough shot. Dantley's efficiency looked great, but he also did not really improve his team overall offensively. Allen Iverson on the other end was never afraid of the tough shot, and in fact from 2003 to 2008 his teams played always better offensively with him on the court than without him (in average we see an increased eFG% and a lower amount of turnovers). The latter is the value of a high usage player, who really has positiv impact on the team's offense.

So, overall bchaikin's question is a strawman, nothing else.

Regarding the article of Berri: Well, Berri is arguing again on an intellectual level which fits his own. Given the fact that Berri isn't the brightest bulb in the box, he argues that ppg is not a great indicator of "greatness"; he likes to argue against supposed "conventional wisdom". I stopped at that point, because obviously Berri likes to argue with people who don't have a clue. Great stuff, but that is as useful as trying to teach a tree to speak. Yeah, no kidding Dave, ppg is not that great in order to determine how good a player is, but neither is WP48, a metric which sucks at predicting lineup performance levels and does not have good predictive power overall as shown in those retrodictive tests.
Well, Ellis and Anthony are also two complete different cases. While Anthony throughout his career improved his team offensively, Ellis didn't do that. Why would I put both into the same category? Only a fool would do that, a fool with no knowledge about basketball.

[Mike G]

The most abrupt departure of a high-usage player would probably be Jordan's first retirement in 1993.
The '92-93 Bulls had a TO% of .120, best in the league and 86% of the league avg.
The Jordanless Bulls of '93-94 were .146, or 102% of league avg.
Of course, he was more than just a scorer. He was a ball handler, and seldom could anyone get the ball away from him without fouling.

[mystic]

The issue with such comparisons is also the differences of the replacement players. The Bulls were able to maintain their defensive strength for example, because they aquired Myers, a player known for his defense rather than his offense. It is also the case that the Bulls during the 92-93 campaign coasted through the regular season and played clearly below the level they showed in 1992 and later in the playoffs 1993. So, the overall offensive impact by Jordan was likely bigger than the difference between the 1993 and the 1994 Bulls.

Iverson for example was replaced by André Miller on the 76ers, a player who had always influenced the offense positively. Thus, we can't expect the numbers seen for the 76er after Iverson leaving giving us a good impression about his offensive impact. The Nuggets replaced Anthony with multiple players who were pretty good offensively anyway. How much of a drop was really expected? Basically none. But that doesn't mean that Anthony did not bring positive value to the Nuggets offense. So, the whole argumentation is a bit silly to me, because it ignores a lot of things.

[Crow]

Looking at Iverson for 2004-5 thru 2008-9, his team was above league average efficiency with him on the court twice and below three times. I couldn't, for some reason, get pages for earlier years at 82games to open for me.

[mystic]

Looking at Iverson for 2004-5 thru 2008-9, his team was above league average efficiency with him on the court twice and below three times. I couldn't, for some reason, get pages for earlier years at 82games to open for me.

Coding issue. You can use an old version of a browser in order to look at those pages. I use Flock 2.5.6 and it is working: http://www.oldapps.com/flock.php?old_flock=2

I also don't see a value in looking at the offensive efficiency in comparison to league average in such a case. If Iverson plays with bad offensive players on the court, you should not expect a crazy offense at all.

[Crow]

I find value in looking at the offensive efficiency in comparison to league average in such a case. Comparing Iverson to his sub is insufficient information to me and adding league average efficiency information to the previous team on / off information helps, especially in an extreme case such as this one where everything was built around Iverson and yet the team with him was below average on team offensive efficiency more often than not, because of the teammates and to some extent the star and the way he played too.


Thanks for offering a software solution to the problem I had with the old data.

[Guy]

I think everyone agrees that the alternative to having Kobe take a lot of shots is not to let the shot clock expire 10 times per game.

before claiming this as fact you might want to take a look at high scoring players who have switched teams during their careers, either via trade or free agency, and then look at what that team did after they left.

I don't understand what it is you are disagreeing with. Let me recap:

Berri: belief in the idea of "shot creation" means that you think the only alternative to having high-usage shooters take their shots is to have no one at all take their shots. So, for example, if a 30% player was limited to 20% usage, an additional 10% of current FGA would become shot clock violations. (Berri then goes on to provide anecdotal evidence that this is not true.)

Me: No basketball analyst who thinks shot creation matters believes what Berri claims. It is a straw man.

So, are you saying that there are in fact analysts who subscribe to the view Berri concept of "shot creation?" If so, who are they? Or are you saying you personally subscribe to it?

[mystic]

I find value in looking at the offensive efficiency in comparison to league average in such a case. Comparing Iverson to his sub is insufficient information to me and adding league average efficiency information to the previous team on / off information helps, especially in an extreme case such as this one where everything was built around Iverson and yet the team with him was below average on team offensive efficiency more often than not, because of the teammates and to some extent the star and the way he played too.

I think we are running into a chicken/egg problem here. The team was build around Iverson, because Iverson provided the stuff, which made it possible to have sucess with limited offensive players. That doesn't mean that we have to expect above league average efficiency with Iverson on the court, just enough improvement offensively with Iverson in order to outscore the opponents overall. The question is: What is cheaper and easier to do? Getting enough above average offensive players with above average defense? Or take a star like Iverson and put defensive minded players with less offense around him? I would argue that with a CBA in place which limits the payment for star players by having a max salary, the latter option is not just easier to achieve, but also cheaper overall.
It seems like that defensive players are cheaper than offensive player. It also seems like the team defense can overall be good with lesser talented players, but the team offense isn't with weak offensive players.

@Guy

Interesting, when I first read your post, I thought you claimed that the alternative to Bryant taking shots is a shot clock violation. Thus, I was a bit confused overall. But now it makes sense. Indeed, the team is not just stupidely let the shot clock expire, someone else has to take that shot. But not everyone is equally good at making those tough shots. That is the value of high usage players like Bryant. On the other end, the teams without their star players on the court have in most cases indeed more turnovers than with that player.

[xkonk]

@Guy

Interesting, when I first read your post, I thought you claimed that the alternative to Bryant taking shots is a shot clock violation. Thus, I was a bit confused overall. But now it makes sense. Indeed, the team is not just stupidely let the shot clock expire, someone else has to take that shot. But not everyone is equally good at making those tough shots. That is the value of high usage players like Bryant. On the other end, the teams without their star players on the court have in most cases indeed more turnovers than with that player.

I think this makes the point very clear. If we want to find that high usage players are helping their teams strictly through their usage, shouldn't we have to demonstrate a) high usage players are better than average at making tough shots (however 'tough shots' might be defined) and b) teams turn the ball over more than would be expected given that a high usage player is off the court? And as part of b, don't we have to account for the fact that the guy coming in to play instead of the high usage player is likely not as good as the high usage player? Kobe Bryant is an interesting example here, because it has been demonstrated a variety of times that in the clutch (which might be an example of tough shot time) Kobe is not actually any better than average.

[mystic]

strictly through their usage

That is the point where I find it difficult to even argue about. Why should we expect that someone is helping strictly through their usage? We should rather expect that a guy is having high usage, because he is better than others.
Well, we can take Monta Ellis as an example for whom we know that this is not the case. I would have given touches rather to Stephen Curry instead of Ellis. So, why exactly did Ellis get more touches? Maybe Evan can put some light on this, as a fan of the Warriors he seems to be predestinated to give a useful answer.

Kobe Bryant is an interesting example here, because it has been demonstrated a variety of times that in the clutch (which might be an example of tough shot time) Kobe is not actually any better than average.

Actually it was shown that he is better than average, just not as much better as some people believe he is. It is also a fact that the Lakers with Bryant played incredible better than without him. So, Bryant seems to be someone who is deserving the touches.