recovered page 1 of 6
Author Message
Ed Küpfer
Joined: 30 Dec 2004
Posts: 785
Location: Toronto
PostPosted: Fri Jan 28, 2011 3:44 pm Post subject: Scorecasting: The Hidden Influences Behind How Sports Are Pl Reply with quote
I just got Scorecasting: The Hidden Influences Behind How Sports Are Played and Games Are Won by Tobias Moskowitz and L. Jon Wertheim. I figured I'd blog the basketball content as I read it. But first, the blurbs:
Quote:
From Booklist
“Defense wins championships!” So declared a triumphant Michael Jordan in 1991, invoking a hallowed sports mantra. But Jordan’s assertion melts into cliché when Moskowitz and Wertheim expose it to statistical calculations revealing that, regardless of the sport, offense proves just as decisive as defense. Indeed, in their wide-ranging iconoclasm, the authors repeatedly poke arithmetic holes in what everyone in sports supposedly knows. Typical is their number-crunching assault on the popular explanation of home-field advantage as a consequence of visiting teams’ road fatigue. Home teams win, the authors demonstrate, chiefly because referees tend to see plays their way—especially when the crowd of spectators grows large. Parsing of data illuminates off-field behavior, too, explaining which athletes use steroids and which ones use marijuana. Even the curse hanging over the Chicago Cubs comes into focus then the analysts ignore the billy-goat myth and statistically assess a management style fostered by fans perversely loyal to “lovable losers”! Sports buffs eager to win their next barroom argument will be lining up for this book. --Bryce Christensen
Review
"The closest thing to Freakonomics I've seen since the original. A rare combination of terrific storytelling and unconventional thinking. I love this book..."
—Steven D. Levitt, Alvin H. Baum Professor of Economics, University of Chicago, and co-author of Freakonomics and SuperFreakonomics
"I love this book. If I told you why, the NBA would fine me again."
—Mark Cuban, owner of the Dallas Mavericks
“Scorecasting is both scholarly and entertaining, a rare double. It gets beyond the cliched narratives and tried-but-not-necessarily-true assumptions to reveal significant and fascinating truths about sports.”
—Bob Costas
"A counterintuitive, innovative, unexpected handbook for sports fans interested in the truths that underpin our favorite games. With their lively minds and prose, Moskowitz and Wertheim will change the way you think about and watch sports. Not just for stats nerds, Scorecasting enlightens and entertains. I wish I had thought of it!"
—Jeremy Schaap, ESPN reporter, Author of Cinderella Man.
"(Sports + numbers) x great writing = winning formula. A must read for all couch analysts."
—Richard Thaler, Professor of Behavioral Science and Economics, best-selling author of Nudge.
“Scorecasting will change the way you watch sports, but don’t start reading it during a game; you’re liable to get lost in it and miss the action. I’m not giving anything away because you’ll want to read exactly how they arrived at their conclusions."--Allen Barra, NJ Star Ledger
Chapter one, "Whistle swallowing."
Talks about the human tendency to view acts of omission less harshly than acts of commission that result in the same thing. Example: Drowning someone is viewed as worse as watching someone drown and doing nothing to help.
What does this have to do with basketball? The authors (M&W) take on "star calls":
That's kind of a neat experiment, focusing on loose balls. DeanO loves to remind us that it's hard to compare statistics between different circumstances, since not only has the environment changed but also the style of play may also have changed and we don't really account for that (eg home vs away stats). The loose ball idea I think takes most of the "style of play" element out of the equation.
There are no footnotes, but I think the research being discussed is in Daniel Cervone and Tobias Moskowitz's working paper "Whistle Swallowing: Officiating and Omission Bias", which I cannot find online.
_________________
ed
Last edited by Ed Küpfer on Wed Feb 02, 2011 11:46 am; edited 1 time in total
Back to top
View user's profile Send private message Send e-mail
Guy
Joined: 02 May 2007
Posts: 128
PostPosted: Fri Jan 28, 2011 3:46 pm Post subject: Reply with quote
FYI, Phil Birnbaum has a post up on the book's look at home field advantage in sports here: http://sabermetricresearch.blogspot.com ... ntage.html
Back to top
View user's profile Send private message
Ed Küpfer
Joined: 30 Dec 2004
Posts: 785
Location: Toronto
PostPosted: Fri Jan 28, 2011 3:48 pm Post subject: Reply with quote
Page 24
This is the kind of thing DeanO worries about: maybe traveling is called less often at the end of tight games because players play differently, knowing their impact on the game outcome is much higher?
_________________
ed
Last edited by Ed Küpfer on Wed Feb 02, 2011 11:46 am; edited 1 time in total
Back to top
View user's profile Send private message Send e-mail
EvanZ
Joined: 22 Nov 2010
Posts: 279
PostPosted: Fri Jan 28, 2011 3:51 pm Post subject: Reply with quote
I suppose one could play devil's advocate and argue that the "star" gets less fouls called because he has more experience and wisdom about how to get loose balls.
I know, I know. Just had to put it out there. Laughing
_________________
http://www.thecity2.com
http://www.ibb.gatech.edu/evan-zamir
Back to top
View user's profile Send private message
Ed Küpfer
Joined: 30 Dec 2004
Posts: 785
Location: Toronto
PostPosted: Fri Jan 28, 2011 4:51 pm Post subject: Reply with quote
Pages 28-30 Not basketball, but tennis: M&W discuss the final game of an important match between two players, where one player was called for a foot fault, which pretty much ruined her chances of a comeback. The point M&W are making is that when officials "insert" themselves into a match by making a call that significantly determines the outcome, they risk ostracism. The official in question was indeed ostracized, even though the call was correct.
There is another way to interpret these events: the official was not ostracized for "inserting" herself and determining the outcome, but for calling a rarely-called foul, which seemed entirely arbitrary to the player and fans. My interpretation is that fans and players may not want correct calls so much as consistent calls.
M&W previously discussed the famous "tuck rule" NFL game, aslo interpreting it as suggesting that making an accurate call may be dangerous to the career health of an official. They quote NFL referee Mike Carey:
Quote:
Making the hard call, or the unpopular call, that's where guts are tested, that's the mark of a true official. You might have a longer career as an official if you back off. But you won't have a more accurate career.
I think my alternate interpretation of correct vs consistent calls is equally plausible.
_________________
ed
Last edited by Ed Küpfer on Wed Feb 02, 2011 11:46 am; edited 1 time in total
Back to top
View user's profile Send private message Send e-mail
EvanZ
Joined: 22 Nov 2010
Posts: 279
PostPosted: Fri Jan 28, 2011 4:56 pm Post subject: Reply with quote
Jim Joyce is a hero for calling it as he saw it! Laughing
_________________
http://www.thecity2.com
http://www.ibb.gatech.edu/evan-zamir
Back to top
View user's profile Send private message
Ed Küpfer
Joined: 30 Dec 2004
Posts: 785
Location: Toronto
PostPosted: Sat Jan 29, 2011 10:02 am Post subject: Reply with quote
Chapter two: Go For It
Long discussion of The Pulaski Academy Bruins football coach, who has an unorthodox offensive philosophy of never punting and always onside kicking. This leads into a discussion of the overly-conservative tactics of most football coaches. Nerd love risky strategies because we think they make sense. Traditionalists don't because they think they make no sense. It's hard for me to make sense of this: the numbers speak for themselves, but the accumulated wisdom of thousands of coaches can't be totally wrong, right? Right? Cynics out there will say "yes of course these coaches can be totally wrong". But I can't make myself believe that. If it's true that conventional wisdom is correct here, then what could we be missing to resolve this?
Pg 43
Does it make sense to take out a player with five fouls with lots of time remaining to save him for the end of the game ? M&W answer "no".
This is not right. Everyone's clutch efficiency is lower than their nonclutch efficiency. They need to compare the star player's efficiency to his replacement's clutch efficiency, not to the star player's nonclutch efficiency.
M&W say
Quote:
Leave a player in with five fouls and what happens? The average player with five fouls will pick up his sixth foul and foul out of the game only 21 percent of the time.
But this merely calls up the DeanO Conundrum: what it the player with five fouls is playing more timidly to avoid that sixth foul? What if this timidity then reduces his efficiency to a level where his replacement then exceeds it?
M&W pin this five-foul strategy on loss aversion, the same thing that makes football coaches punt on 4th and short. I'll accept the probability of an overly conservative 4th down strategy, but I don't think the five-foul strategy is necessarily a mistake.
_________________
ed
Last edited by Ed Küpfer on Wed Feb 02, 2011 11:47 am; edited 1 time in total
Back to top
View user's profile Send private message Send e-mail
Mike G
Joined: 14 Jan 2005
Posts: 3596
Location: Hendersonville, NC
PostPosted: Sat Jan 29, 2011 10:20 am Post subject: Reply with quote
Ed Küpfer wrote:
Quote:
Leave a player in with five fouls and what happens? The average player with five fouls will pick up his sixth foul and foul out of the game only 21 percent of the time.
But this merely calls up the DeanO Conundrum: what it the player with five fouls is playing more timidly to avoid that sixth foul? .
Is this 21% of the time when a player with 5 fouls is left in the game that he eventually fouls out?
Or is it 21% of all 5th fouls lead to a 6th?
If the latter -- and if such players generally don't stay in the game -- then it hardly makes sense to call this 'what happens' if you leave the player in the game.
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
Ed Küpfer
Joined: 30 Dec 2004
Posts: 785
Location: Toronto
PostPosted: Sat Jan 29, 2011 10:47 am Post subject: Reply with quote
Mike G wrote:
Is this 21% of the time when a player with 5 fouls is left in the game that he eventually fouls out?
Or is it 21% of all 5th fouls lead to a 6th?
If the latter -- and if such players generally don't stay in the game -- then it hardly makes sense to call this 'what happens' if you leave the player in the game.
I believe they mean the former, but I don't think it's spelled out explicitly. Or perhaps I am misreading the text.
Some data here would be helpful, if any of you PBP wizards are reading this.
_________________
ed
Back to top
View user's profile Send private message Send e-mail
Ed Küpfer
Joined: 30 Dec 2004
Posts: 785
Location: Toronto
PostPosted: Sat Jan 29, 2011 4:05 pm Post subject: Reply with quote
pg 47 The case of Bill Belichick. M&W note that the coach who is well known for risky strategies today did not begin his career this way (they contrast the Patriot's 4th down go-for-its versus the Browns from his previous employment). They chalk up this change to Belichick's increasing job security over the years, pointing out that he didn't go for it on 4th downs as often in his early Patriots career as he does today.
It's a neat story. I've heard this explanation before. I wonder if it's true. Are there other coaches who employ more risky strategies as they become more successful? Are there coaches who employ risky strategies who don't have as much job security as a multi Super Bowl winner? Seems that this explanation, if true, should be generalizable.
p55 Belichick is contrasted with Paul Westhead, who tried to install a different kind of risky strategy in Denver with a much less success-filled pedigree. The failure of his experiment is tied by the authors to his players. But that's not the important point the authors want to make: in contrast with Belichick, whose risky strategies were praised,
_________________
ed
Last edited by Ed Küpfer on Wed Feb 02, 2011 11:47 am; edited 1 time in total
Back to top
View user's profile Send private message Send e-mail
Jeff Fogle
Joined: 11 Jan 2011
Posts: 68
PostPosted: Sat Jan 29, 2011 6:21 pm Post subject: Reply with quote
From the wikipedia entry on Westhead:
"Laker owner Jerry Buss wanted to fire Westhead several days prior to the actual occurrence. The entire Laker team did not get along with Westhead."
Magic Johnson was seen as running Westhead out...just before the Lakers blossomed under Pat Riley. The players apparently backed Magic. Subsequent Westhead NBA records:
28-54 with Chicago in '82-83
20-62 with Denver in '90-91
24-58 with Denver in '91-92
That's 72-174 over three years, AFTER a good group of players ran him off their team. That's not the kind of portfolio that's going to get you hired somewhere else in the NBA. A guy that's hard to get along with who looks to be in over his head at the pro level.
In '93, Westhead went to George Mason, where he was 38-70 over a few seasons. An NBA team is going to hire him away from George Mason while he's compiling a 38-70 college record...after going 44-120 with Denver?
There are other possible descriptions besides calling him an eccentric mad professor without tenure. And, the guy was a head coach at either the NBA or Division I college level through the 80's and most of the 90's.
Anyway, isn't it generally accepted that his style exhausts his own defenses over time because they're the only team playing that pace over the course of a season? It's not one that befuddles or challenges opposing NBA coaches.
Back to top
View user's profile Send private message
Ed Küpfer
Joined: 30 Dec 2004
Posts: 785
Location: Toronto
PostPosted: Sun Jan 30, 2011 11:49 am Post subject: Reply with quote
p73 A fun example of loss aversion from football:
Scenario 1 - offense starts a series on the opponents one yard line, and now faces fourth and goal after running three straight times and not gaining anything.
Scenario 2 - offense starts on the opponent's ten yard line, gains nine yards on three downs and now faces fourth and goal on the opponents one yard line.
Both teams are 4th and goal from the one yard line. Which scenario is more likely to pass up the FG to go for the TD? The scenario where the team has been stuffed at the goal line three times in a row? Or the scenario where the team has gained three yards per play on the previous three downs? I wouldn't be writing this if it was Scenario 2. The team in scenario 1 has already mentally accounted for the TD because they started the series 1st and goal at the one-yard line, and are scared of losing the TD they thought was "in the bank". The team in scenario 2 had less expectation of a TD, and are more willing to settle for the FG.
Like I said, a fun example. But the difference in go-for-itness doesn't seem that large to me (67% versus 59%). There is no research in the bibliography that seems relevant to this. I'm wondering to what extent this difference is significant, and were other factors accounted for?
_________________
ed
Last edited by Ed Küpfer on Wed Feb 02, 2011 11:47 am; edited 1 time in total
Back to top
View user's profile Send private message Send e-mail
Ed Küpfer
Joined: 30 Dec 2004
Posts: 785
Location: Toronto
PostPosted: Sun Jan 30, 2011 2:30 pm Post subject: Reply with quote
p75-6
Is shooting more three pointers really an example of playing more aggressively? Haven't we all heard the commentator trope that teams should stop "settling" for jumpers and go strong to the rim? But that's just a minor point. The rest of this looks plausible to me, although again no actual numbers are presented an it doesn't look like any relevant research is listed in the bibliography -- what is the n here? Out of 5000 games, how many meet the conditions of A) <= 5 point score differential at the start of the 4th Q, and B) one team had a >= 15 point lead in the 3rd Q? To what extent is the difference in aggressiveness significant, both in the practical and significant senses?
_________________
ed
Last edited by Ed Küpfer on Wed Feb 02, 2011 11:48 am; edited 1 time in total
Back to top
View user's profile Send private message Send e-mail
Jeff Fogle
Joined: 11 Jan 2011
Posts: 68
PostPosted: Sun Jan 30, 2011 2:57 pm Post subject: Reply with quote
Agree with you EK about three-point shooting not representing aggression. I think typically flying at the basket and trying to score or draw a foul is considered being aggressive, while teams who are being passive settle for jumpers (though, I know we've all seen games where teams were aggressively moving the ball around trying to free up a shooter, so the shades of gray are pretty significant).
My issues would be with:
*Conventional wisdom being that teams get passive AFTER blowing a lead. Never heard that. They get passive with a big lead. They blow the big lead. The coach complains after the game that they got passive with the big lead. Not familiar with the concept of teams getting passive AFTER they've blown a lead. Authors said, "Conventional widsom suggests that Team A will play passively" after their lead has already been blown. Disagree with that assessment.
*If you throw out any wild fluctuations that happened earlier in the game, and just start with one team having a lead entering the fourth quarter...isn't it natural for the trailing team to shoot faster and try more threes? They're trailing. Time is running out. Isn't it natural for the team with a lead to treasure their possessions more, hoping to protect their lead as time runs out?
That would seem "normal" for fourth quarters, regardless of what happened earlier. So, the evidence they uncovered in their study might just be a "fourth quarter lead" phenomenon rather than a "what happens after a big lead is erased" phenomenon. I would have preferred they compare their type of fourth quarter with the standard fourth quarter.
Back to top
View user's profile Send private message
EvanZ
Joined: 22 Nov 2010
Posts: 279
PostPosted: Sun Jan 30, 2011 3:51 pm Post subject: Reply with quote
As far as the three-pointers go, I would go with DeanO's terminology of "risky" rather than aggressive. I agree that "aggressive" would be driving to the basket, largely with intention of getting fouled.
_________________
http://www.thecity2.com
http://www.ibb.gatech.edu/evan-zamir
Scorecasting:Hidden Influences Behind How Sports Are Played
Re: Scorecasting:Hidden Influences Behind How Sports Are Pla
additional pages recovered from the files of WiLQ
Guy
Joined: 02 May 2007
Posts: 104
PostPosted: Sun Jan 30, 2011 5:19 pm Post subject: Reply with quote
Ed Küpfer wrote:
Both teams are 4th and goal from the one yard line. Which scenario is more likely to pass up the FG to go for the TD? The scenario where the team has been stuffed at the goal line three times in a row? Or the scenario where the team has gained three yards per play on the previous three downs?
Seems like this could be explained by game situation. The team that got stuffed three times is more likely to be losing, I would think. In which case, going for it is explained by the score not by loss aversion.
Also, going for it greatly increases the risk of getting zero points on the possession, which is arguably a greater "loss" than just getting a field goal. It seems to me that these authors could -- and likely would -- cite this as evidence of loss aversion no matter that the two go-for-it rates were, unless both teams were exactly 63%.
Back to top
View user's profile Send private message
EvanZ
Joined: 22 Nov 2010
Posts: 177
PostPosted: Sun Jan 30, 2011 5:36 pm Post subject: Reply with quote
Is "loss aversion" a new euphism for "winning"? Laughing
_________________
http://www.thecity2.com
http://www.ibb.gatech.edu/evan-zamir
Back to top
View user's profile Send private message
Ed Küpfer
Joined: 30 Dec 2004
Posts: 783
Location: Toronto
PostPosted: Mon Jan 31, 2011 1:39 pm Post subject: Reply with quote
EvanZ wrote:
Is "loss aversion" a new euphism for "winning"? Laughing
I know you're joking, but it strikes me that maybe I didn't explain what they mean by loss aversion: it is the finding that people will have stronger feelings about losing a dollar they had in their pocket than not getting a dollar they find in the street when it blows away before they get their hands on it. Am I explaining this right? It would be the difference between a team heading into the last minute of a game tied with their opponents, who they dominated for 47 minutes but blew the lead, and a team heading into the last minute of a game tied, having been down the entire game. In both cases the present situation is the same, but the first team is more desperate to win, having already "won" that game in their heads.
_________________
ed
Back to top
View user's profile Send private message Send e-mail
Ed Küpfer
Joined: 30 Dec 2004
Posts: 783
Location: Toronto
PostPosted: Mon Jan 31, 2011 1:41 pm Post subject: Reply with quote
Chapter Five: Offense Wins Championships, Too
After showing that defensive prowess is no more likely to predict championships than offensive prowess in a bunch of sports, M&W reflect on what the phrase "defense wins championships" could possibly mean. They settle on the following explanation: there are plenty of incentives for individual offensive effort in the form of statistics, but there are relatively few for defensive effort. The saying "defense wins championships" is a way to incentivize defensive effort.
M&W manage to tie this back to loss aversion too.
[img]file:///C:/My%20Documents/recovery/viewtopic.php2_pliki/iypWu.jpg[/img]
I like this.
_________________
ed
Last edited by Ed Küpfer on Wed Feb 02, 2011 11:48 am; edited 1 time in total
Back to top
View user's profile Send private message Send e-mail
EvanZ
Joined: 22 Nov 2010
Posts: 177
PostPosted: Mon Jan 31, 2011 2:19 pm Post subject: Reply with quote
Ed Küpfer wrote:
EvanZ wrote:
Is "loss aversion" a new euphism for "winning"? Laughing
I know you're joking, but it strikes me that maybe I didn't explain what they mean by loss aversion: it is the finding that people will have stronger feelings about losing a dollar they had in their pocket than not getting a dollar they find in the street when it blows away before they get their hands on it. Am I explaining this right? It would be the difference between a team heading into the last minute of a game tied with their opponents, who they dominated for 47 minutes but blew the lead, and a team heading into the last minute of a game tied, having been down the entire game. In both cases the present situation is the same, but the first team is more desperate to win, having already "won" that game in their heads.
Yeah, I got it. I just think it's a funny term.
Maybe "win expectation" would be more useful (i.e. Team A had a lower win expectation going into the last minute.)
_________________
http://www.thecity2.com
http://www.ibb.gatech.edu/evan-zamir
Back to top
View user's profile Send private message
EvanZ
Joined: 22 Nov 2010
Posts: 177
PostPosted: Mon Jan 31, 2011 2:25 pm Post subject: Reply with quote
I think people say "defense wins championships" with the implicit assumption that most people think offense is somehow more important, and by saying "defense wins championships" it makes you somehow seem more knowledgeable. They are equally important. More appropriate would be to say "defense also wins championships" or "defense is equally important to offense for winning championship." Doesn't sound as wise, though. Wink
_________________
http://www.thecity2.com
http://www.ibb.gatech.edu/evan-zamir
Back to top
View user's profile Send private message
Mike G
Joined: 14 Jan 2005
Posts: 3458
Location: Hendersonville, NC
PostPosted: Mon Jan 31, 2011 2:45 pm Post subject: Reply with quote
In playoffs, don't teams tend to maintain their defensive prowess more than they do their offensive prowess?
Another way to look at it is that offense is more volatile, and defense is more reliable. If your shots aren't falling, your offense sucks, and you hope to snap out of it. Defense is more about staying on the job and remaining intense.
An individual, highly competitive player will sometimes have a bad game on offense, but he can partly make up for that on defense. You can't just will yourself to shoot better, but you can always hustle on D.
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
Jeff Fogle
Joined: 11 Jan 2011
Posts: 24
PostPosted: Mon Jan 31, 2011 4:12 pm Post subject: Reply with quote
EK, could you post any specific thing the authors said about basketball when discussing the "bunch of sports" you referenced? Thanks.
In terms of incentives. Defensive players in college or pro football are self-contained in a way that wouldn't need additional motivation to perform their best. They're only on the field for defense. They're not accumulating stats in a way that's meaningful beyond sacks or interceptions. They're "stats" are keeping the other team off the scoreboard. Their motivation to play at their best is that they are excluded from the lineup if they don't play their best.
In baseball, the biggest influence against run scoring is the pitcher, who's trying to get people out, and may also be motivated by his statistics. There don't need to be any extra incentives there in a "defense wins championships" way.
In basketball, one could see the line of thinking that you have to motivate "two-way players" to also defend. But, I've heard the phrase as far back as the Celtics championship era when Bill Russell was a dominant defender who was seen as leading the team to titles. Maybe that line of thinking started in basketball because Russell was seen as a dominating force at the time pro hoops made inroads in the national media...
Interested in seeing data on the offense/defense dynanimc in pro and college data. Anyone know of any studies on that?
Back to top
View user's profile Send private message
BobboFitos
Joined: 21 Feb 2009
Posts: 180
Location: Cambridge, MA
PostPosted: Mon Jan 31, 2011 11:17 pm Post subject: Reply with quote
EvanZ wrote:
I think people say "defense wins championships" with the implicit assumption that most people think offense is somehow more important, and by saying "defense wins championships" it makes you somehow seem more knowledgeable. They are equally important. More appropriate would be to say "defense also wins championships" or "defense is equally important to offense for winning championship." Doesn't sound as wise, though. Wink
Technically preventing points is slightly more valuable than scoring points, but by and large both are pretty important!
_________________
-Rob
Back to top
View user's profile Send private message AIM Address
gabefarkas
Joined: 31 Dec 2004
Posts: 1306
Location: Durham, NC
PostPosted: Tue Feb 01, 2011 2:28 pm Post subject: Reply with quote
Ed Küpfer wrote:
Pages 28-30 Not basketball, but tennis: M&W discuss the final game of an important match between two players, where one player was called for a foot fault, which pretty much ruined her chances of a comeback. The point M&W are making is that when officials "insert" themselves into a match by making a call that significantly determines the outcome, they risk ostracism. The official in question was indeed ostracized, even though the call was correct.
There is another way to interpret these events: the official was not ostracized for "inserting" herself and determining the outcome, but for calling a rarely-called foul, which seemed entirely arbitrary to the player and fans. My interpretation is that fans and players may not want correct calls so much as consistent calls.
M&W previously discussed the famous "tuck rule" NFL game, aslo interpreting it as suggesting that making an accurate call may be dangerous to the career health of an official. They quote NFL referee Mike Carey:
Quote:
Making the hard call, or the unpopular call, that's where guts are tested, that's the mark of a true official. You might have a longer career as an official if you back off. But you won't have a more accurate career.
I think my alternate interpretation of correct vs consistent calls is equally plausible.
Maybe the player actually committed a foot fault?
Back to top
View user's profile Send private message Send e-mail AIM Address
schtevie
Joined: 18 Apr 2005
Posts: 392
PostPosted: Tue Feb 01, 2011 4:41 pm Post subject: Reply with quote
Ed Küpfer wrote:
....Nerd love risky strategies because we think they make sense. Traditionalists don't because they think they make no sense. It's hard for me to make sense of this: the numbers speak for themselves, but the accumulated wisdom of thousands of coaches can't be totally wrong, right? Right? Cynics out there will say "yes of course these coaches can be totally wrong". But I can't make myself believe that. If it's true that conventional wisdom is correct here, then what could we be missing to resolve this?
It is because of my utmost respect for your perspicacity that I sadly conclude that you have been corrupted by wealth, power, and fame. Coaches (acting as a herd) can be totally wrong (with large, counterfactual, competitive benefits foregone for the innovators who weren't). And this collective cud chewing can last decades (and has). You can make yourself believe that! (That said, I think you are basically correct about the fifth foul thing.)
I dutifully repeat the most (?) glaring evidence of MOOOOOOOOOO!
(1) The peculiar pace and productivity paradox, circa the beginning of the shot clock era (probably) and ending in the mid-80s (more or less). Slowing game pace cannot be reconciled with increasing offensive productivity (efficiency) without concluding that there were secular, net offensive improvements, in particular, teams culled hasty (i.e. bad) shots from the offense and they turned the ball over less. But to my understanding there were absolutely no conceptual revolutions or incremental innovations undergirding this dramatic improvement. Said another way, there was nothing stopping a team in the 60s from doing some the most simple math (really just having an intuitive sense about the sloppiness of run and gun offenses vs. well run half court plays) and reaping the huge competitive advantages of modernity. Moo.
(2) The slooooooow adoption of the 3 point shot. Is it plausible to believe that no coaches in the early 80s had an intuitive awareness of the crappiness of contested mid-range jump shots compared to the potential of 3 pointers? It cannot be that there was a general inability to multiply by 1.5. The concept of the benefits of spacing existed. And furthermore coaches surely knew the potential of the rule change given the very recent history of the ABA. There were big private gains to be had by aggressively innovating in this area, but these were not seen, just incremental, tentative small steps, a slow transformation with the herd moving essentially as one.
Anyone else with other examples of coaches being "totally wrong"?
Regarding the (near) "optimality" of sitting a starter for long minutes on getting a fifth foul, some fake, plausible numbers to generate an upper bound estimate of the cost of doing this. Suppose a starter is 2 points per game better than his replacement. Furthermore suppose that as a worst case he sits a whole quarter, to ensure that he doesn't foul out. What is the cost to the team of this strategy. If he were to stay in the game, there is about a 0.25 chance that he fouls out, say, on average, at the midpoint of the quarter. In this case the expected cost is 0.44 points foregone. Is this big or small? I am thinking that in a coach's consideration, this is small compared to the downside abuse that would be had if the player fouls out sooner than expected, especially if the team were to lose a close game. And then there is the relatively small likelihood that the 0.44 would be determinative in the outcome, and that the game outcome would be important for the season. And so the $10 bills aren't picked up from the sidewalk.
Back to top
View user's profile Send private message
Ed Küpfer
Joined: 30 Dec 2004
Posts: 783
Location: Toronto
PostPosted: Tue Feb 01, 2011 5:29 pm Post subject: Reply with quote
Jeff Fogle wrote:
EK, could you post any specific thing the authors said about basketball when discussing the "bunch of sports" you referenced? Thanks.
I skimmed over it earlier because I felt that this particular result wasn't very surprising to the more quant-oriented folks here.
* Comparison of super bowl winners versus losers, better defensive team vs better offensive team shows similar win%. I think pts scored and allowed was the offensive/defensive metric.
* Same comparison, all NFL playoff games.
* Best NFL offense vs best NFL defense.
* Same comparison in NBA for playoffs and regular season. Not sure how they measure offense and defense.
* Same comparison in NHL for playoffs and regular season. Not sure how they measure offense and defense.
* World Series winners and losers.
* "Okay, but does defense give an underdog more of a chance? Are upsets more likely to be sprung by defensive-minded teams? Sifting through the statistics, we found that the answer is no. We calculated that in the regular season, playoffs, and championships, underdog teams are no more likely to win if they are good defenders than if they are good scorers." Intriguing. but that's all they have to say about that aspect.
_________________
ed
Back to top
View user's profile Send private message Send e-mail
Ed Küpfer
Joined: 30 Dec 2004
Posts: 783
Location: Toronto
PostPosted: Tue Feb 01, 2011 5:31 pm Post subject: Reply with quote
gabefarkas wrote:
Maybe the player actually committed a foot fault?
No maybe, she definitely did. I don't know anything about tennis, but M&W made it sound like this particular foot fault was something that was virtually never called. It would be like getting a ticket for feeding pigeons in a park -- it may actually be an infraction, but since it is almost never enforced this particular ticket seems arbitrary and unjust.
_________________
ed
Back to top
View user's profile Send private message Send e-mail
Ed Küpfer
Joined: 30 Dec 2004
Posts: 783
Location: Toronto
PostPosted: Tue Feb 01, 2011 5:34 pm Post subject: Reply with quote
schtevie wrote:
It is because of my utmost respect for your perspicacity that I sadly conclude that you have been corrupted by wealth, power, and fame. Coaches (acting as a herd) can be totally wrong (with large, counterfactual, competitive benefits foregone for the innovators who weren't). And this collective cud chewing can last decades (and has).
This may be true, but I have been much more wrong far too often to assume it. In any case, the question isn't whether they are wrong, but whether they are wronger than anyone else in that position (sort of a Strategy Wins Above Replacement).
_________________
ed
Back to top
View user's profile Send private message Send e-mail
BobboFitos
Joined: 21 Feb 2009
Posts: 180
Location: Cambridge, MA
PostPosted: Tue Feb 01, 2011 5:37 pm Post subject: Reply with quote
Ed Küpfer wrote:
gabefarkas wrote:
Maybe the player actually committed a foot fault?
No maybe, she definitely did. I don't know anything about tennis, but M&W made it sound like this particular foot fault was something that was virtually never called. It would be like getting a ticket for feeding pigeons in a park -- it may actually be an infraction, but since it is almost never enforced this particular ticket seems arbitrary and unjust.
I don't know if any of you are Tennis fans, but I remember watching this match last year:
http://larrybrownsports.com/tennis/andy ... ideo/28723
Foot faults are somewhat rare but it's not like they're never called. 10 seconds to shoot a FT, for example, is a far rarer call!
_________________
-Rob
Back to top
View user's profile Send private message AIM Address
Ed Küpfer
Joined: 30 Dec 2004
Posts: 783
Location: Toronto
PostPosted: Tue Feb 01, 2011 5:39 pm Post subject: Reply with quote
Chapter Six: The Value of a Blocked Shot
Discusses Weil and Huizinga's paper, already threaded here. I don't really have much to add -- see Albert Lyu's series of blogs for another look.
_________________
ed
Last edited by Ed Küpfer on Wed Feb 02, 2011 11:45 am; edited 1 time in total
Back to top
View user's profile Send private message Send e-mail
mtamada
Joined: 28 Jan 2005
Posts: 375
PostPosted: Tue Feb 01, 2011 5:50 pm Post subject: Reply with quote
An observation from another sport, which relates to both the "does defense win championships" question as well as the "do coaches make optimal decisions" question. Soccer scoring at the international level (probably domestically too but I haven seen stats) has been or at least was slowly decreasing for decades. In soccer, I think it is/was clear that defense did win championships in the sense that teams which put more emphasis on defense did better, i.e. a more conservative, defend-your-half-of-the-field strategy led to more wins and more championships. I remember reading old soccer books which listed the positions, and there'd be FIVE forwards. I don't think any team uses more than two these days, and often they go with just one striker (albeit supported with some attacking midfielders).
But, it took awhile for this to sink in, and soccer gradually became more and more defensively oriented. So much so that FIFA had to start tweaking the rules, much like the NBA with its constant tweaking of the handchecking rules. Offsides only if you're ahead of the 2nd-to-last defender, not if you're even with him. Goalies can't handle the ball if a teammate kicks it back to him. I think they also raised the penalty, or widened the definition, of tackling from behind? Etc. etc. (Which has helped some; the 1990 World Cup was practically unwatchable, it was like NCAA basketball during the pre-shot clock, 4-corners offense era; the 1994 and subsequent World Cups have been better.)
So it's not out of the question that similar movements happened in the NBA. I think defenses did become more sophisticated comparing say the 1990s to the 1970s (although offense probably became more sophisticated too -- and also slower-paced, which is a different thing). But it took years.
But pretty much any revolution takes years (okay maybe not Tunisia). Dick Fosbury totally changed high-jumping at the 1968 Olympics but competing jumpers couldn't instantly switch to the Fosbury Flop. According to Wikipedia, 4 years later 28 out of 40 Olympians were using it. I'd have to think that in high stakes team sports such as basketball, most large innovations start at low levels -- high school or small colleges, and slowly percolate upwards. An NBA coach doesn't have a lot of time much less opportunity to experiment with a big innovation which might flop and cost the team an entire season (and the coach his job). Simply telling your players "shoot more 3 pointers" without an offensive strategy surrounding that dictum is more likely to simply result in bad shots rather than good offense.
OTOH, it wouldn't be a huge deal for the NBA coach to gradually add a few plays to the playbook designed to free the best shooter for an open 3-pointer. That would've been an easy, low-risk innovation, and as Schtevie says, NBA coaches seem to have been slow to do so.
Back to top
View user's profile Send private message
Jeff Fogle
Joined: 11 Jan 2011
Posts: 24
PostPosted: Tue Feb 01, 2011 6:03 pm Post subject: Reply with quote
Thanks EK. Is it fair to say that the general studies that have been done by quants in this regard would entail:
*Look at regular season offensive ranking
*Look at regular season defensive ranking
*See who won the championship
As opposed to something like...looking at the championship game or series itself, as it was played out...to see whether it was defense or offense that was "imposing it's will" in a way that determined who won the championship that day?
Or, the prior studies turned the phrase "defense wins championships" into a premise like "regular season defense wins post-season championships" which they then disproved?
I'm afraid of moving into what the definition of "is" is type stuff...but, I'm trying to reconcile the various ways of looking at what the phrase means, and how defense has been defined over the years (and whether regular season defense has been adjusted for schedule quirks, injuries, ballpark effects in baseball, etc...).
For example: Boston and Bill Russell won the '58-59 championship despite ranking 6th in an 8 team league in points allowed. Then again in 59-60 with the 5th ranking of 8 in points allowed. It looks from shot totals that Boston played a faster pace than everyone else in those years. But, Russell was still Russell...and as a kid I can remember hearing the phrase "defense wins championships" linked back to Russell's perceived edges over Chamberlain in the stuff that mattered. Defense can still "win" the championship if it was defense being the difference-maker in the games that determined who won.
I guess I want to make sure that all-time data being used in studies isn't polluted. And, I've seen too many people saying the equivalent of "we sifted through the stats, trust us," when they hadn't done nearly the due diligence one would have hoped for.
Thanks...
Back to top
View user's profile Send private message
schtevie
Joined: 18 Apr 2005
Posts: 392
PostPosted: Tue Feb 01, 2011 6:05 pm Post subject: Reply with quote
Ed Küpfer wrote:
schtevie wrote:
It is because of my utmost respect for your perspicacity that I sadly conclude that you have been corrupted by wealth, power, and fame. Coaches (acting as a herd) can be totally wrong (with large, counterfactual, competitive benefits foregone for the innovators who weren't). And this collective cud chewing can last decades (and has).
This may be true, but I have been much more wrong far too often to assume it. In any case, the question isn't whether they are wrong, but whether they are wronger than anyone else in that position (sort of a Strategy Wins Above Replacement).
FOOT FAULT!
How is the question not whether "they" are wrong? Who else is in that position but they (including those above and below replacement value)? The point is that the accumulated wisdom isn't always. If there is an apparently, competitively significant sub-optimal strategy that persists over decades, the phenomenon needs to be explained (or explained away).
Back to top
View user's profile Send private message
Ed Küpfer
Joined: 30 Dec 2004
Posts: 783
Location: Toronto
PostPosted: Wed Feb 02, 2011 11:43 am Post subject: Reply with quote
Jeff Fogle wrote:
As opposed to something like...looking at the championship game or series itself, as it was played out...to see whether it was defense or offense that was "imposing it's will" in a way that determined who won the championship that day?
This was also done, if not in this book than elsewhere. I'm pretty sure it's not controversial at all to say that defense has as much to do with winning, however you want to define any of those terms, as offense. I mean, there's probably a way to frame this so that you'll find defense contributes more, but that would just be cherry picking your result.
M&W weren't exhaustive in their attempts to compare offensive and defensive contributions to winning because it wasn't their main objective: what they were trying to do was make a point about incentives.
_________________
ed
Back to top
View user's profile Send private message Send e-mail
Ed Küpfer
Joined: 30 Dec 2004
Posts: 783
Location: Toronto
PostPosted: Wed Feb 02, 2011 11:45 am Post subject: Reply with quote
Chapter Seven: Rounding First
Discusses the findings that players perform differently when they are on the cusp of a statistical threshold than when they are slightly passed it (eg players hitting .299 perform differently on their next at bat than when they are hitting .300). There's no basketball content in this chapter, probably because basketball isn't as obsessed with this kind of irrelevant numerology to the same extent baseball is, but some ambitious nerd out there may be able to extend this finding to triple- or double-doubles.
_________________
ed
Back to top
View user's profile Send private message Send e-mail
EvanZ
Joined: 22 Nov 2010
Posts: 177
PostPosted: Wed Feb 02, 2011 12:08 pm Post subject: Reply with quote
I would think in basketball, there are many more 50 or 51 point performances than, say, 48 or 49. Just a hunch.
_________________
http://www.thecity2.com
http://www.ibb.gatech.edu/evan-zamir
Back to top
View user's profile Send private message
EvanZ
Joined: 22 Nov 2010
Posts: 177
PostPosted: Wed Feb 02, 2011 12:15 pm Post subject: Reply with quote
I thought wrong. Shocked Very Happy
Since 1986 there were 114 performances of 47-49 points and 99 of 50-52 points. Pretty random. I love that Basketball-Reference game finder.
Did you know...since 1986 there were three games where a player had >=10 AST and >= 10 blocks? Hakeem had two (both in 1990). Robinson had the other ('94).
_________________
http://www.thecity2.com
http://www.ibb.gatech.edu/evan-zamir
Last edited by EvanZ on Wed Feb 02, 2011 12:18 pm; edited 1 time in total
Back to top
View user's profile Send private message
Guy
Joined: 02 May 2007
Posts: 104
PostPosted: Wed Feb 02, 2011 12:16 pm Post subject: Reply with quote
Ed Küpfer wrote:
Chapter Seven: Rounding First
Discusses the findings that players perform differently when they are on the cusp of a statistical threshold than when they are slightly passed it (eg players hitting .299 perform differently on their next at bat than when they are hitting .300).
Ed, this is a totally discredited study they are relying upon. It purports to show that hitters perform better when a hit in their final at-bat will get them to the .300 level. What the study failed to account for was selection bias: when a hitter gets a hit and passes .300, he often leaves the lineup. If he doesn't get a hit, he keeps playing in hopes of passing .300. So his "last" at bat is a function of his success. It is true that hitters care about .300 (or they wouldn't leave the lineup), but it is not true that professional athletes can dramatically improve their performance just because they "really want to." Which I think should have been obvious to the researchers. I think Phil Birnbaum has a post or two on this a while back...
Back to top
View user's profile Send private message
Guy
Joined: 02 May 2007
Posts: 104
PostPosted: Wed Feb 02, 2011 12:28 pm Post subject: Reply with quote
schtevie wrote:
How is the question not whether "they" are wrong? Who else is in that position but they (including those above and below replacement value)? The point is that the accumulated wisdom isn't always. If there is an apparently, competitively significant sub-optimal strategy that persists over decades, the phenomenon needs to be explained (or explained away).
This is a really great and important question: to what extent should we assume sports decision-makers are doing things optimally? Or at least close? I think the answer is really "it depends." On the one hand, Schtevie presents a couple of good examples of non-optimal behavior on a significant scale. Mistakes are made. On the other hand, I think the behaviorial economists now digging into sports (I would include Berri here, and it sounds like the author of this book as well) tend to greatly overestimate the extent of non-rational decisions in professional sports. Most of the studies I see claiming to show irrational behavior by athletes or coaches ends up revealing only the ignorance of the researcher.
I would offer a couple of caveats to Schtevie's examples:
1) his examples are rather old now. I think the longer a sport has been played at a high professional level, and with serious money involved, the less likely it is we will see large-scale suboptimal behavior. Our presumption of rationality should be greater for the game played in 2011 than in 1991, or 1961.
2) We need to distinguish between failure to discover a successful innovation and sub-optimal behavior. Baseball players could have hit HRs before Babe Ruth showed them how in 1919, but I don't think you can say teams were acting irrationally before that because they failed to figure out a new possibility. To some extent the same could be said of 3-point shots. On the other hand, if NFL teams really attempt 4th-down conversions less than they should, I would say that is clearly sub-optimal behavior -- teams have had all the information they need to make that decision for a long time.
Back to top
View user's profile Send private message
greyberger
Joined: 27 Sep 2010
Posts: 33
PostPosted: Wed Feb 02, 2011 2:48 pm Post subject: Reply with quote
Quote:
On the other hand, I think the behaviorial economists now digging into sports (I would include Berri here, and it sounds like the author of this book as well) tend to greatly overestimate the extent of non-rational decisions in professional sports.
I hate to sound cynical but I always assumed this was just marketing. It's true and worth pointing out that decision-makers will walk past tens on the sidewalk to avoid the risk of looking dumb. Too often this just gets pushed farther into the emperor has no clothes argument, though.
Back to top
View user's profile Send private message Send e-mail
acollard
Joined: 22 Sep 2010
Posts: 49
Location: MA
PostPosted: Wed Feb 02, 2011 3:33 pm Post subject: Reply with quote
Greybereger, Guy, I agree. I think for the most part decision makers get it right. It may seem like they don't get it right very often because those instances are sought out and highlighted by writers/economists because they are interesting, and because hindsight is often 20/20.
However, there are letgit examples of times when the decision makers in sports (or anywhere) get it wrong. I think almost all of those fall into on or more of four categories:
1) Averse to looking dumb with a risky (but technically right) move
2) Favoring short-term vs. long-term success
3) Following the conventional wisdom blindly
4) Conflicting objectives (profiting vs. winning in sports)
I also think there are some aspects that numbers fail to grasp. For example, the fourth down call in football. Just looking at the odds, I agree it makes sense to go for it on fourth down reasonably frequently. But, momentum is such a huge factor in football, and a big failure like a gamble and loss in that situation can change the game a lot more than just giving a team good field position. Momentum is something I think would actually be pretty interesting to look at, and I'm not sure how to measure it, or if it could be measured.
Back to top
View user's profile Send private message
gabefarkas
Joined: 31 Dec 2004
Posts: 1306
Location: Durham, NC
PostPosted: Wed Feb 02, 2011 9:41 pm Post subject: Reply with quote
Ed Küpfer wrote:
gabefarkas wrote:
Maybe the player actually committed a foot fault?
No maybe, she definitely did. I don't know anything about tennis, but M&W made it sound like this particular foot fault was something that was virtually never called. It would be like getting a ticket for feeding pigeons in a park -- it may actually be an infraction, but since it is almost never enforced this particular ticket seems arbitrary and unjust.
We're talking about Serena Williams, right? If so, I've seen the replay a few times (tennis is my second love, after basketball), and it was a clear foot fault.
It's not called often because it doesn't happen often. But when it does, it should be called, regardless of the circumstances. Isn't that what we always talk about in basketball with refereeing -- consistency in the calls?
It's a matter of focus. She lost focus at that time, and committed an error.
Back to top
View user's profile Send private message Send e-mail AIM Address
schtevie
Joined: 18 Apr 2005
Posts: 392
PostPosted: Thu Feb 03, 2011 10:35 am Post subject: Reply with quote
Apologies for the thread drift, but a couple of points of clarification and reply.
First, I am not of the view that there are as many dollar bills left on the sidewalk today as there were decades ago - too many people nowadays are productively employed picking them up - nor that those that remain are of similar denominations. But they surely exist and are very significant.
Second, I think I stepped on my lede a bit. Ed was referring to a reluctance of coaches to adopt risky strategies for apparent, potential gain. I was making a much stronger point: that coaches throughout NBA history have neglected to adopt non-risky strategies that had immediate and ginormous gain. The first example of decreasing the sloppiness of offenses is obvious in this regard, but so too is the slow implementation of three point offenses.
Consider the curious case of Larry Bird (and in considering this case, let me disagree with Mike T. My strong impression is that strategic progress doesn't percolate to the NBA from the high school ranks; rather, it is all top down, awaiting the implementation by perceived NBA winners, champions really, who by having achieved that status, aren't much interested in tampering with anything. That is the essence of the slow-adoption/stagnation phenomenon.)
Larry Bird was a really good shot, from 3 point range and everywhere else. His overall history with the shot is a bit strange but telling, a 0.400 shot when it was a larger (but not large) part of his game, but only a mid-0.200s when neglected early in his career. If we assume (as his potential exhibited) that he was indeed a 0.400 3 point shooter, and we make some ordinary assumptions about the fraction of shots he took about the rim that had a very high success rate, the picture that emerges is that there was a very high marginal trade-off between his 3 pointers and the average of all other non-close shots (and what must have been higher still for the inevitably crappy, contested, long 2s). Probably between 0.33 and 0.5 points per shot. Given how many shots Bird took, a little enlightened self-interest would have had huge effect.
Imagine how the history of the league would have been different were these non-risky swaps of 0.400 shots to have been made. My guess is a couple extra banners for the Celtics and a rapid transformation of NBA offenses to adopt the demonstrated benefits of 3 point shooting.
But alas, Eleanor Roosevelt cannot fly....
Back to top
View user's profile Send private message
Guy
Joined: 02 May 2007
Posts: 104
PostPosted: Thu Feb 03, 2011 10:57 am Post subject: Reply with quote
Ed: If you are interested in the .300 hitter issue, see Phil Birnbaum here: http://sabermetricresearch.blogspot.com ... nning.html, and here: http://sabermetricresearch.blogspot.com ... ng_06.html, and here: http://sabermetricresearch.blogspot.com ... ng_07.html.
Bottom line: hitters can NOT improve their performance simply because they want to reach a statistical benchmark.
Guy
Joined: 02 May 2007
Posts: 104
PostPosted: Sun Jan 30, 2011 5:19 pm Post subject: Reply with quote
Ed Küpfer wrote:
Both teams are 4th and goal from the one yard line. Which scenario is more likely to pass up the FG to go for the TD? The scenario where the team has been stuffed at the goal line three times in a row? Or the scenario where the team has gained three yards per play on the previous three downs?
Seems like this could be explained by game situation. The team that got stuffed three times is more likely to be losing, I would think. In which case, going for it is explained by the score not by loss aversion.
Also, going for it greatly increases the risk of getting zero points on the possession, which is arguably a greater "loss" than just getting a field goal. It seems to me that these authors could -- and likely would -- cite this as evidence of loss aversion no matter that the two go-for-it rates were, unless both teams were exactly 63%.
Back to top
View user's profile Send private message
EvanZ
Joined: 22 Nov 2010
Posts: 177
PostPosted: Sun Jan 30, 2011 5:36 pm Post subject: Reply with quote
Is "loss aversion" a new euphism for "winning"? Laughing
_________________
http://www.thecity2.com
http://www.ibb.gatech.edu/evan-zamir
Back to top
View user's profile Send private message
Ed Küpfer
Joined: 30 Dec 2004
Posts: 783
Location: Toronto
PostPosted: Mon Jan 31, 2011 1:39 pm Post subject: Reply with quote
EvanZ wrote:
Is "loss aversion" a new euphism for "winning"? Laughing
I know you're joking, but it strikes me that maybe I didn't explain what they mean by loss aversion: it is the finding that people will have stronger feelings about losing a dollar they had in their pocket than not getting a dollar they find in the street when it blows away before they get their hands on it. Am I explaining this right? It would be the difference between a team heading into the last minute of a game tied with their opponents, who they dominated for 47 minutes but blew the lead, and a team heading into the last minute of a game tied, having been down the entire game. In both cases the present situation is the same, but the first team is more desperate to win, having already "won" that game in their heads.
_________________
ed
Back to top
View user's profile Send private message Send e-mail
Ed Küpfer
Joined: 30 Dec 2004
Posts: 783
Location: Toronto
PostPosted: Mon Jan 31, 2011 1:41 pm Post subject: Reply with quote
Chapter Five: Offense Wins Championships, Too
After showing that defensive prowess is no more likely to predict championships than offensive prowess in a bunch of sports, M&W reflect on what the phrase "defense wins championships" could possibly mean. They settle on the following explanation: there are plenty of incentives for individual offensive effort in the form of statistics, but there are relatively few for defensive effort. The saying "defense wins championships" is a way to incentivize defensive effort.
M&W manage to tie this back to loss aversion too.
[img]file:///C:/My%20Documents/recovery/viewtopic.php2_pliki/iypWu.jpg[/img]
I like this.
_________________
ed
Last edited by Ed Küpfer on Wed Feb 02, 2011 11:48 am; edited 1 time in total
Back to top
View user's profile Send private message Send e-mail
EvanZ
Joined: 22 Nov 2010
Posts: 177
PostPosted: Mon Jan 31, 2011 2:19 pm Post subject: Reply with quote
Ed Küpfer wrote:
EvanZ wrote:
Is "loss aversion" a new euphism for "winning"? Laughing
I know you're joking, but it strikes me that maybe I didn't explain what they mean by loss aversion: it is the finding that people will have stronger feelings about losing a dollar they had in their pocket than not getting a dollar they find in the street when it blows away before they get their hands on it. Am I explaining this right? It would be the difference between a team heading into the last minute of a game tied with their opponents, who they dominated for 47 minutes but blew the lead, and a team heading into the last minute of a game tied, having been down the entire game. In both cases the present situation is the same, but the first team is more desperate to win, having already "won" that game in their heads.
Yeah, I got it. I just think it's a funny term.
Maybe "win expectation" would be more useful (i.e. Team A had a lower win expectation going into the last minute.)
_________________
http://www.thecity2.com
http://www.ibb.gatech.edu/evan-zamir
Back to top
View user's profile Send private message
EvanZ
Joined: 22 Nov 2010
Posts: 177
PostPosted: Mon Jan 31, 2011 2:25 pm Post subject: Reply with quote
I think people say "defense wins championships" with the implicit assumption that most people think offense is somehow more important, and by saying "defense wins championships" it makes you somehow seem more knowledgeable. They are equally important. More appropriate would be to say "defense also wins championships" or "defense is equally important to offense for winning championship." Doesn't sound as wise, though. Wink
_________________
http://www.thecity2.com
http://www.ibb.gatech.edu/evan-zamir
Back to top
View user's profile Send private message
Mike G
Joined: 14 Jan 2005
Posts: 3458
Location: Hendersonville, NC
PostPosted: Mon Jan 31, 2011 2:45 pm Post subject: Reply with quote
In playoffs, don't teams tend to maintain their defensive prowess more than they do their offensive prowess?
Another way to look at it is that offense is more volatile, and defense is more reliable. If your shots aren't falling, your offense sucks, and you hope to snap out of it. Defense is more about staying on the job and remaining intense.
An individual, highly competitive player will sometimes have a bad game on offense, but he can partly make up for that on defense. You can't just will yourself to shoot better, but you can always hustle on D.
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
Jeff Fogle
Joined: 11 Jan 2011
Posts: 24
PostPosted: Mon Jan 31, 2011 4:12 pm Post subject: Reply with quote
EK, could you post any specific thing the authors said about basketball when discussing the "bunch of sports" you referenced? Thanks.
In terms of incentives. Defensive players in college or pro football are self-contained in a way that wouldn't need additional motivation to perform their best. They're only on the field for defense. They're not accumulating stats in a way that's meaningful beyond sacks or interceptions. They're "stats" are keeping the other team off the scoreboard. Their motivation to play at their best is that they are excluded from the lineup if they don't play their best.
In baseball, the biggest influence against run scoring is the pitcher, who's trying to get people out, and may also be motivated by his statistics. There don't need to be any extra incentives there in a "defense wins championships" way.
In basketball, one could see the line of thinking that you have to motivate "two-way players" to also defend. But, I've heard the phrase as far back as the Celtics championship era when Bill Russell was a dominant defender who was seen as leading the team to titles. Maybe that line of thinking started in basketball because Russell was seen as a dominating force at the time pro hoops made inroads in the national media...
Interested in seeing data on the offense/defense dynanimc in pro and college data. Anyone know of any studies on that?
Back to top
View user's profile Send private message
BobboFitos
Joined: 21 Feb 2009
Posts: 180
Location: Cambridge, MA
PostPosted: Mon Jan 31, 2011 11:17 pm Post subject: Reply with quote
EvanZ wrote:
I think people say "defense wins championships" with the implicit assumption that most people think offense is somehow more important, and by saying "defense wins championships" it makes you somehow seem more knowledgeable. They are equally important. More appropriate would be to say "defense also wins championships" or "defense is equally important to offense for winning championship." Doesn't sound as wise, though. Wink
Technically preventing points is slightly more valuable than scoring points, but by and large both are pretty important!
_________________
-Rob
Back to top
View user's profile Send private message AIM Address
gabefarkas
Joined: 31 Dec 2004
Posts: 1306
Location: Durham, NC
PostPosted: Tue Feb 01, 2011 2:28 pm Post subject: Reply with quote
Ed Küpfer wrote:
Pages 28-30 Not basketball, but tennis: M&W discuss the final game of an important match between two players, where one player was called for a foot fault, which pretty much ruined her chances of a comeback. The point M&W are making is that when officials "insert" themselves into a match by making a call that significantly determines the outcome, they risk ostracism. The official in question was indeed ostracized, even though the call was correct.
There is another way to interpret these events: the official was not ostracized for "inserting" herself and determining the outcome, but for calling a rarely-called foul, which seemed entirely arbitrary to the player and fans. My interpretation is that fans and players may not want correct calls so much as consistent calls.
M&W previously discussed the famous "tuck rule" NFL game, aslo interpreting it as suggesting that making an accurate call may be dangerous to the career health of an official. They quote NFL referee Mike Carey:
Quote:
Making the hard call, or the unpopular call, that's where guts are tested, that's the mark of a true official. You might have a longer career as an official if you back off. But you won't have a more accurate career.
I think my alternate interpretation of correct vs consistent calls is equally plausible.
Maybe the player actually committed a foot fault?
Back to top
View user's profile Send private message Send e-mail AIM Address
schtevie
Joined: 18 Apr 2005
Posts: 392
PostPosted: Tue Feb 01, 2011 4:41 pm Post subject: Reply with quote
Ed Küpfer wrote:
....Nerd love risky strategies because we think they make sense. Traditionalists don't because they think they make no sense. It's hard for me to make sense of this: the numbers speak for themselves, but the accumulated wisdom of thousands of coaches can't be totally wrong, right? Right? Cynics out there will say "yes of course these coaches can be totally wrong". But I can't make myself believe that. If it's true that conventional wisdom is correct here, then what could we be missing to resolve this?
It is because of my utmost respect for your perspicacity that I sadly conclude that you have been corrupted by wealth, power, and fame. Coaches (acting as a herd) can be totally wrong (with large, counterfactual, competitive benefits foregone for the innovators who weren't). And this collective cud chewing can last decades (and has). You can make yourself believe that! (That said, I think you are basically correct about the fifth foul thing.)
I dutifully repeat the most (?) glaring evidence of MOOOOOOOOOO!
(1) The peculiar pace and productivity paradox, circa the beginning of the shot clock era (probably) and ending in the mid-80s (more or less). Slowing game pace cannot be reconciled with increasing offensive productivity (efficiency) without concluding that there were secular, net offensive improvements, in particular, teams culled hasty (i.e. bad) shots from the offense and they turned the ball over less. But to my understanding there were absolutely no conceptual revolutions or incremental innovations undergirding this dramatic improvement. Said another way, there was nothing stopping a team in the 60s from doing some the most simple math (really just having an intuitive sense about the sloppiness of run and gun offenses vs. well run half court plays) and reaping the huge competitive advantages of modernity. Moo.
(2) The slooooooow adoption of the 3 point shot. Is it plausible to believe that no coaches in the early 80s had an intuitive awareness of the crappiness of contested mid-range jump shots compared to the potential of 3 pointers? It cannot be that there was a general inability to multiply by 1.5. The concept of the benefits of spacing existed. And furthermore coaches surely knew the potential of the rule change given the very recent history of the ABA. There were big private gains to be had by aggressively innovating in this area, but these were not seen, just incremental, tentative small steps, a slow transformation with the herd moving essentially as one.
Anyone else with other examples of coaches being "totally wrong"?
Regarding the (near) "optimality" of sitting a starter for long minutes on getting a fifth foul, some fake, plausible numbers to generate an upper bound estimate of the cost of doing this. Suppose a starter is 2 points per game better than his replacement. Furthermore suppose that as a worst case he sits a whole quarter, to ensure that he doesn't foul out. What is the cost to the team of this strategy. If he were to stay in the game, there is about a 0.25 chance that he fouls out, say, on average, at the midpoint of the quarter. In this case the expected cost is 0.44 points foregone. Is this big or small? I am thinking that in a coach's consideration, this is small compared to the downside abuse that would be had if the player fouls out sooner than expected, especially if the team were to lose a close game. And then there is the relatively small likelihood that the 0.44 would be determinative in the outcome, and that the game outcome would be important for the season. And so the $10 bills aren't picked up from the sidewalk.
Back to top
View user's profile Send private message
Ed Küpfer
Joined: 30 Dec 2004
Posts: 783
Location: Toronto
PostPosted: Tue Feb 01, 2011 5:29 pm Post subject: Reply with quote
Jeff Fogle wrote:
EK, could you post any specific thing the authors said about basketball when discussing the "bunch of sports" you referenced? Thanks.
I skimmed over it earlier because I felt that this particular result wasn't very surprising to the more quant-oriented folks here.
* Comparison of super bowl winners versus losers, better defensive team vs better offensive team shows similar win%. I think pts scored and allowed was the offensive/defensive metric.
* Same comparison, all NFL playoff games.
* Best NFL offense vs best NFL defense.
* Same comparison in NBA for playoffs and regular season. Not sure how they measure offense and defense.
* Same comparison in NHL for playoffs and regular season. Not sure how they measure offense and defense.
* World Series winners and losers.
* "Okay, but does defense give an underdog more of a chance? Are upsets more likely to be sprung by defensive-minded teams? Sifting through the statistics, we found that the answer is no. We calculated that in the regular season, playoffs, and championships, underdog teams are no more likely to win if they are good defenders than if they are good scorers." Intriguing. but that's all they have to say about that aspect.
_________________
ed
Back to top
View user's profile Send private message Send e-mail
Ed Küpfer
Joined: 30 Dec 2004
Posts: 783
Location: Toronto
PostPosted: Tue Feb 01, 2011 5:31 pm Post subject: Reply with quote
gabefarkas wrote:
Maybe the player actually committed a foot fault?
No maybe, she definitely did. I don't know anything about tennis, but M&W made it sound like this particular foot fault was something that was virtually never called. It would be like getting a ticket for feeding pigeons in a park -- it may actually be an infraction, but since it is almost never enforced this particular ticket seems arbitrary and unjust.
_________________
ed
Back to top
View user's profile Send private message Send e-mail
Ed Küpfer
Joined: 30 Dec 2004
Posts: 783
Location: Toronto
PostPosted: Tue Feb 01, 2011 5:34 pm Post subject: Reply with quote
schtevie wrote:
It is because of my utmost respect for your perspicacity that I sadly conclude that you have been corrupted by wealth, power, and fame. Coaches (acting as a herd) can be totally wrong (with large, counterfactual, competitive benefits foregone for the innovators who weren't). And this collective cud chewing can last decades (and has).
This may be true, but I have been much more wrong far too often to assume it. In any case, the question isn't whether they are wrong, but whether they are wronger than anyone else in that position (sort of a Strategy Wins Above Replacement).
_________________
ed
Back to top
View user's profile Send private message Send e-mail
BobboFitos
Joined: 21 Feb 2009
Posts: 180
Location: Cambridge, MA
PostPosted: Tue Feb 01, 2011 5:37 pm Post subject: Reply with quote
Ed Küpfer wrote:
gabefarkas wrote:
Maybe the player actually committed a foot fault?
No maybe, she definitely did. I don't know anything about tennis, but M&W made it sound like this particular foot fault was something that was virtually never called. It would be like getting a ticket for feeding pigeons in a park -- it may actually be an infraction, but since it is almost never enforced this particular ticket seems arbitrary and unjust.
I don't know if any of you are Tennis fans, but I remember watching this match last year:
http://larrybrownsports.com/tennis/andy ... ideo/28723
Foot faults are somewhat rare but it's not like they're never called. 10 seconds to shoot a FT, for example, is a far rarer call!
_________________
-Rob
Back to top
View user's profile Send private message AIM Address
Ed Küpfer
Joined: 30 Dec 2004
Posts: 783
Location: Toronto
PostPosted: Tue Feb 01, 2011 5:39 pm Post subject: Reply with quote
Chapter Six: The Value of a Blocked Shot
Discusses Weil and Huizinga's paper, already threaded here. I don't really have much to add -- see Albert Lyu's series of blogs for another look.
_________________
ed
Last edited by Ed Küpfer on Wed Feb 02, 2011 11:45 am; edited 1 time in total
Back to top
View user's profile Send private message Send e-mail
mtamada
Joined: 28 Jan 2005
Posts: 375
PostPosted: Tue Feb 01, 2011 5:50 pm Post subject: Reply with quote
An observation from another sport, which relates to both the "does defense win championships" question as well as the "do coaches make optimal decisions" question. Soccer scoring at the international level (probably domestically too but I haven seen stats) has been or at least was slowly decreasing for decades. In soccer, I think it is/was clear that defense did win championships in the sense that teams which put more emphasis on defense did better, i.e. a more conservative, defend-your-half-of-the-field strategy led to more wins and more championships. I remember reading old soccer books which listed the positions, and there'd be FIVE forwards. I don't think any team uses more than two these days, and often they go with just one striker (albeit supported with some attacking midfielders).
But, it took awhile for this to sink in, and soccer gradually became more and more defensively oriented. So much so that FIFA had to start tweaking the rules, much like the NBA with its constant tweaking of the handchecking rules. Offsides only if you're ahead of the 2nd-to-last defender, not if you're even with him. Goalies can't handle the ball if a teammate kicks it back to him. I think they also raised the penalty, or widened the definition, of tackling from behind? Etc. etc. (Which has helped some; the 1990 World Cup was practically unwatchable, it was like NCAA basketball during the pre-shot clock, 4-corners offense era; the 1994 and subsequent World Cups have been better.)
So it's not out of the question that similar movements happened in the NBA. I think defenses did become more sophisticated comparing say the 1990s to the 1970s (although offense probably became more sophisticated too -- and also slower-paced, which is a different thing). But it took years.
But pretty much any revolution takes years (okay maybe not Tunisia). Dick Fosbury totally changed high-jumping at the 1968 Olympics but competing jumpers couldn't instantly switch to the Fosbury Flop. According to Wikipedia, 4 years later 28 out of 40 Olympians were using it. I'd have to think that in high stakes team sports such as basketball, most large innovations start at low levels -- high school or small colleges, and slowly percolate upwards. An NBA coach doesn't have a lot of time much less opportunity to experiment with a big innovation which might flop and cost the team an entire season (and the coach his job). Simply telling your players "shoot more 3 pointers" without an offensive strategy surrounding that dictum is more likely to simply result in bad shots rather than good offense.
OTOH, it wouldn't be a huge deal for the NBA coach to gradually add a few plays to the playbook designed to free the best shooter for an open 3-pointer. That would've been an easy, low-risk innovation, and as Schtevie says, NBA coaches seem to have been slow to do so.
Back to top
View user's profile Send private message
Jeff Fogle
Joined: 11 Jan 2011
Posts: 24
PostPosted: Tue Feb 01, 2011 6:03 pm Post subject: Reply with quote
Thanks EK. Is it fair to say that the general studies that have been done by quants in this regard would entail:
*Look at regular season offensive ranking
*Look at regular season defensive ranking
*See who won the championship
As opposed to something like...looking at the championship game or series itself, as it was played out...to see whether it was defense or offense that was "imposing it's will" in a way that determined who won the championship that day?
Or, the prior studies turned the phrase "defense wins championships" into a premise like "regular season defense wins post-season championships" which they then disproved?
I'm afraid of moving into what the definition of "is" is type stuff...but, I'm trying to reconcile the various ways of looking at what the phrase means, and how defense has been defined over the years (and whether regular season defense has been adjusted for schedule quirks, injuries, ballpark effects in baseball, etc...).
For example: Boston and Bill Russell won the '58-59 championship despite ranking 6th in an 8 team league in points allowed. Then again in 59-60 with the 5th ranking of 8 in points allowed. It looks from shot totals that Boston played a faster pace than everyone else in those years. But, Russell was still Russell...and as a kid I can remember hearing the phrase "defense wins championships" linked back to Russell's perceived edges over Chamberlain in the stuff that mattered. Defense can still "win" the championship if it was defense being the difference-maker in the games that determined who won.
I guess I want to make sure that all-time data being used in studies isn't polluted. And, I've seen too many people saying the equivalent of "we sifted through the stats, trust us," when they hadn't done nearly the due diligence one would have hoped for.
Thanks...
Back to top
View user's profile Send private message
schtevie
Joined: 18 Apr 2005
Posts: 392
PostPosted: Tue Feb 01, 2011 6:05 pm Post subject: Reply with quote
Ed Küpfer wrote:
schtevie wrote:
It is because of my utmost respect for your perspicacity that I sadly conclude that you have been corrupted by wealth, power, and fame. Coaches (acting as a herd) can be totally wrong (with large, counterfactual, competitive benefits foregone for the innovators who weren't). And this collective cud chewing can last decades (and has).
This may be true, but I have been much more wrong far too often to assume it. In any case, the question isn't whether they are wrong, but whether they are wronger than anyone else in that position (sort of a Strategy Wins Above Replacement).
FOOT FAULT!
How is the question not whether "they" are wrong? Who else is in that position but they (including those above and below replacement value)? The point is that the accumulated wisdom isn't always. If there is an apparently, competitively significant sub-optimal strategy that persists over decades, the phenomenon needs to be explained (or explained away).
Back to top
View user's profile Send private message
Ed Küpfer
Joined: 30 Dec 2004
Posts: 783
Location: Toronto
PostPosted: Wed Feb 02, 2011 11:43 am Post subject: Reply with quote
Jeff Fogle wrote:
As opposed to something like...looking at the championship game or series itself, as it was played out...to see whether it was defense or offense that was "imposing it's will" in a way that determined who won the championship that day?
This was also done, if not in this book than elsewhere. I'm pretty sure it's not controversial at all to say that defense has as much to do with winning, however you want to define any of those terms, as offense. I mean, there's probably a way to frame this so that you'll find defense contributes more, but that would just be cherry picking your result.
M&W weren't exhaustive in their attempts to compare offensive and defensive contributions to winning because it wasn't their main objective: what they were trying to do was make a point about incentives.
_________________
ed
Back to top
View user's profile Send private message Send e-mail
Ed Küpfer
Joined: 30 Dec 2004
Posts: 783
Location: Toronto
PostPosted: Wed Feb 02, 2011 11:45 am Post subject: Reply with quote
Chapter Seven: Rounding First
Discusses the findings that players perform differently when they are on the cusp of a statistical threshold than when they are slightly passed it (eg players hitting .299 perform differently on their next at bat than when they are hitting .300). There's no basketball content in this chapter, probably because basketball isn't as obsessed with this kind of irrelevant numerology to the same extent baseball is, but some ambitious nerd out there may be able to extend this finding to triple- or double-doubles.
_________________
ed
Back to top
View user's profile Send private message Send e-mail
EvanZ
Joined: 22 Nov 2010
Posts: 177
PostPosted: Wed Feb 02, 2011 12:08 pm Post subject: Reply with quote
I would think in basketball, there are many more 50 or 51 point performances than, say, 48 or 49. Just a hunch.
_________________
http://www.thecity2.com
http://www.ibb.gatech.edu/evan-zamir
Back to top
View user's profile Send private message
EvanZ
Joined: 22 Nov 2010
Posts: 177
PostPosted: Wed Feb 02, 2011 12:15 pm Post subject: Reply with quote
I thought wrong. Shocked Very Happy
Since 1986 there were 114 performances of 47-49 points and 99 of 50-52 points. Pretty random. I love that Basketball-Reference game finder.
Did you know...since 1986 there were three games where a player had >=10 AST and >= 10 blocks? Hakeem had two (both in 1990). Robinson had the other ('94).
_________________
http://www.thecity2.com
http://www.ibb.gatech.edu/evan-zamir
Last edited by EvanZ on Wed Feb 02, 2011 12:18 pm; edited 1 time in total
Back to top
View user's profile Send private message
Guy
Joined: 02 May 2007
Posts: 104
PostPosted: Wed Feb 02, 2011 12:16 pm Post subject: Reply with quote
Ed Küpfer wrote:
Chapter Seven: Rounding First
Discusses the findings that players perform differently when they are on the cusp of a statistical threshold than when they are slightly passed it (eg players hitting .299 perform differently on their next at bat than when they are hitting .300).
Ed, this is a totally discredited study they are relying upon. It purports to show that hitters perform better when a hit in their final at-bat will get them to the .300 level. What the study failed to account for was selection bias: when a hitter gets a hit and passes .300, he often leaves the lineup. If he doesn't get a hit, he keeps playing in hopes of passing .300. So his "last" at bat is a function of his success. It is true that hitters care about .300 (or they wouldn't leave the lineup), but it is not true that professional athletes can dramatically improve their performance just because they "really want to." Which I think should have been obvious to the researchers. I think Phil Birnbaum has a post or two on this a while back...
Back to top
View user's profile Send private message
Guy
Joined: 02 May 2007
Posts: 104
PostPosted: Wed Feb 02, 2011 12:28 pm Post subject: Reply with quote
schtevie wrote:
How is the question not whether "they" are wrong? Who else is in that position but they (including those above and below replacement value)? The point is that the accumulated wisdom isn't always. If there is an apparently, competitively significant sub-optimal strategy that persists over decades, the phenomenon needs to be explained (or explained away).
This is a really great and important question: to what extent should we assume sports decision-makers are doing things optimally? Or at least close? I think the answer is really "it depends." On the one hand, Schtevie presents a couple of good examples of non-optimal behavior on a significant scale. Mistakes are made. On the other hand, I think the behaviorial economists now digging into sports (I would include Berri here, and it sounds like the author of this book as well) tend to greatly overestimate the extent of non-rational decisions in professional sports. Most of the studies I see claiming to show irrational behavior by athletes or coaches ends up revealing only the ignorance of the researcher.
I would offer a couple of caveats to Schtevie's examples:
1) his examples are rather old now. I think the longer a sport has been played at a high professional level, and with serious money involved, the less likely it is we will see large-scale suboptimal behavior. Our presumption of rationality should be greater for the game played in 2011 than in 1991, or 1961.
2) We need to distinguish between failure to discover a successful innovation and sub-optimal behavior. Baseball players could have hit HRs before Babe Ruth showed them how in 1919, but I don't think you can say teams were acting irrationally before that because they failed to figure out a new possibility. To some extent the same could be said of 3-point shots. On the other hand, if NFL teams really attempt 4th-down conversions less than they should, I would say that is clearly sub-optimal behavior -- teams have had all the information they need to make that decision for a long time.
Back to top
View user's profile Send private message
greyberger
Joined: 27 Sep 2010
Posts: 33
PostPosted: Wed Feb 02, 2011 2:48 pm Post subject: Reply with quote
Quote:
On the other hand, I think the behaviorial economists now digging into sports (I would include Berri here, and it sounds like the author of this book as well) tend to greatly overestimate the extent of non-rational decisions in professional sports.
I hate to sound cynical but I always assumed this was just marketing. It's true and worth pointing out that decision-makers will walk past tens on the sidewalk to avoid the risk of looking dumb. Too often this just gets pushed farther into the emperor has no clothes argument, though.
Back to top
View user's profile Send private message Send e-mail
acollard
Joined: 22 Sep 2010
Posts: 49
Location: MA
PostPosted: Wed Feb 02, 2011 3:33 pm Post subject: Reply with quote
Greybereger, Guy, I agree. I think for the most part decision makers get it right. It may seem like they don't get it right very often because those instances are sought out and highlighted by writers/economists because they are interesting, and because hindsight is often 20/20.
However, there are letgit examples of times when the decision makers in sports (or anywhere) get it wrong. I think almost all of those fall into on or more of four categories:
1) Averse to looking dumb with a risky (but technically right) move
2) Favoring short-term vs. long-term success
3) Following the conventional wisdom blindly
4) Conflicting objectives (profiting vs. winning in sports)
I also think there are some aspects that numbers fail to grasp. For example, the fourth down call in football. Just looking at the odds, I agree it makes sense to go for it on fourth down reasonably frequently. But, momentum is such a huge factor in football, and a big failure like a gamble and loss in that situation can change the game a lot more than just giving a team good field position. Momentum is something I think would actually be pretty interesting to look at, and I'm not sure how to measure it, or if it could be measured.
Back to top
View user's profile Send private message
gabefarkas
Joined: 31 Dec 2004
Posts: 1306
Location: Durham, NC
PostPosted: Wed Feb 02, 2011 9:41 pm Post subject: Reply with quote
Ed Küpfer wrote:
gabefarkas wrote:
Maybe the player actually committed a foot fault?
No maybe, she definitely did. I don't know anything about tennis, but M&W made it sound like this particular foot fault was something that was virtually never called. It would be like getting a ticket for feeding pigeons in a park -- it may actually be an infraction, but since it is almost never enforced this particular ticket seems arbitrary and unjust.
We're talking about Serena Williams, right? If so, I've seen the replay a few times (tennis is my second love, after basketball), and it was a clear foot fault.
It's not called often because it doesn't happen often. But when it does, it should be called, regardless of the circumstances. Isn't that what we always talk about in basketball with refereeing -- consistency in the calls?
It's a matter of focus. She lost focus at that time, and committed an error.
Back to top
View user's profile Send private message Send e-mail AIM Address
schtevie
Joined: 18 Apr 2005
Posts: 392
PostPosted: Thu Feb 03, 2011 10:35 am Post subject: Reply with quote
Apologies for the thread drift, but a couple of points of clarification and reply.
First, I am not of the view that there are as many dollar bills left on the sidewalk today as there were decades ago - too many people nowadays are productively employed picking them up - nor that those that remain are of similar denominations. But they surely exist and are very significant.
Second, I think I stepped on my lede a bit. Ed was referring to a reluctance of coaches to adopt risky strategies for apparent, potential gain. I was making a much stronger point: that coaches throughout NBA history have neglected to adopt non-risky strategies that had immediate and ginormous gain. The first example of decreasing the sloppiness of offenses is obvious in this regard, but so too is the slow implementation of three point offenses.
Consider the curious case of Larry Bird (and in considering this case, let me disagree with Mike T. My strong impression is that strategic progress doesn't percolate to the NBA from the high school ranks; rather, it is all top down, awaiting the implementation by perceived NBA winners, champions really, who by having achieved that status, aren't much interested in tampering with anything. That is the essence of the slow-adoption/stagnation phenomenon.)
Larry Bird was a really good shot, from 3 point range and everywhere else. His overall history with the shot is a bit strange but telling, a 0.400 shot when it was a larger (but not large) part of his game, but only a mid-0.200s when neglected early in his career. If we assume (as his potential exhibited) that he was indeed a 0.400 3 point shooter, and we make some ordinary assumptions about the fraction of shots he took about the rim that had a very high success rate, the picture that emerges is that there was a very high marginal trade-off between his 3 pointers and the average of all other non-close shots (and what must have been higher still for the inevitably crappy, contested, long 2s). Probably between 0.33 and 0.5 points per shot. Given how many shots Bird took, a little enlightened self-interest would have had huge effect.
Imagine how the history of the league would have been different were these non-risky swaps of 0.400 shots to have been made. My guess is a couple extra banners for the Celtics and a rapid transformation of NBA offenses to adopt the demonstrated benefits of 3 point shooting.
But alas, Eleanor Roosevelt cannot fly....
Back to top
View user's profile Send private message
Guy
Joined: 02 May 2007
Posts: 104
PostPosted: Thu Feb 03, 2011 10:57 am Post subject: Reply with quote
Ed: If you are interested in the .300 hitter issue, see Phil Birnbaum here: http://sabermetricresearch.blogspot.com ... nning.html, and here: http://sabermetricresearch.blogspot.com ... ng_06.html, and here: http://sabermetricresearch.blogspot.com ... ng_07.html.
Bottom line: hitters can NOT improve their performance simply because they want to reach a statistical benchmark.
Re: Scorecasting:Hidden Influences Behind How Sports Are Pla
Guy
Joined: 02 May 2007
Posts: 104
PostPosted: Thu Feb 03, 2011 11:58 am Post subject: Reply with quote
schtevie wrote:
Ed was referring to a reluctance of coaches to adopt risky strategies for apparent, potential gain. I was making a much stronger point: that coaches throughout NBA history have neglected to adopt non-risky strategies that had immediate and ginormous gain.
It seems to me that taking more 3-point shots is indeed a riskier strategy, in that it increases the probability of a bad short-term outcome (zero points on the possession). That is similar to the 4th-down play in football: punting decreases the chance of a very bad immediate outcome, i.e. turning the ball over at current field position.
Was it really obvious in the 1980s that players should take more 3s? Did teams have good data on the success rate on long-2s, to allow that comparison? I'm not disagreeing, I just don't know the relevant history here.
Back to top
View user's profile Send private message
Jeff Fogle
Joined: 11 Jan 2011
Posts: 24
PostPosted: Thu Feb 03, 2011 1:28 pm Post subject: Reply with quote
If you accept the premise that it takes years of practice to truly master long range shooting, then it's tough to judge 3-point strategies in the NBA in the 1980's based on how we see the world now.
The NBA adopted the line for the '79-80 season.
It wasn't universal in the colleges until the 86-87 season.
High schools match the colleges
So, you didn't have a set of players who had been shooting the shot their whole lives until the 90's. That may have influenced the confidence coaches had in using that as an important part of the offense.
Plus, Larry Bird was impossible to guard in his 20's! He was an amazing shooter, earned 400+FT attempts most years where he hit 88% or so, and he had a ton of assists. You want a guy like that driving to the basket because odds are good you'll get something...often getting something that attaches a personal foul to one of the opposing bigs too. Can see why the emphasis for Bird would be attacking the basket until his body couldn't take it any more. High trey attempts over a period of years didn't begin until he was 29 or so.
http://www.basketball-reference.com/pla ... dla01.html
Feel fortunate to have seen a lot of him on TV. Indiana State ended up getting a lot of "Game of the Week" telecasts across the middle section of the country when he was there (even down in Texas). Was that the Hughes Sports Network? Something like that. No three's in the college game at the time, so it would have been odd for him to emphasize three's as a pro until he had to.
Quick note on the .299/.300 baseball discussion. Worth considering I think that pitching quality may not be all that intense on the last day of the regular season. Usually a meaningless game...possibly a September call-up or late rotation guy throwing...possibly in a "pitch-to-contact" way rather than bearing down. Even if the hitter can't make himself a better hitter on command, he might be facing a less challenging task then during a typical regular season at bat. He's a .299 hitter for the season, but a .340 hitter vs. a September call-up or a back-of-the-rotation guy throwing to contact as both teams are just finishing out the season in front of a three-quarters empty stadium...
On fourth down stuff in football. I'm pretty confident that defenses would come up with some very creative attacks if they knew in advance that a team was going to go for it on all fourth downs. Maybe that doesn't matter at lower levels. But, among elite college programs and the NFL,you have great athletes disrupting plays. A chance to disrupt a play deep in opposing territory doesn't seem like a big mystery to solve. The studies I've seen in favor of going for it on 4th down all the time tend to use averages from past fourth down plays. Past 4th down plays weren't deep in a team's own territory generally, and often came agaisnt a prevent when the game was out of reach. Game in reach? Different set of circumstances, and successful defensive strategies would likely evolve even if they don't exist already. Defenses generally control the flow of football games if you look at success on a drive-by-drive basis.
Back to top
View user's profile Send private message
Guy
Joined: 02 May 2007
Posts: 104
PostPosted: Thu Feb 03, 2011 1:43 pm Post subject: Reply with quote
Quote:
Quick note on the .299/.300 baseball discussion. Worth considering I think that pitching quality may not be all that intense on the last day of the regular season....He's a .299 hitter for the season, but a .340 hitter vs. a September call-up or a back-of-the-rotation guy.
No, that's not plausible -- .340 vs. .299 is huge. It's the difference between an average hitter and a guy who can't make the majors, or an average hitter and a Hall of Famer. Pitching quality does not change remotely that much in September. And in any case, Phil Birnbaum actually looked at the data, and hitters don't improve at all (without any control for pitcher quality). There's just nothing there.....
Back to top
View user's profile Send private message
parinella
Joined: 16 Dec 2005
Posts: 10
PostPosted: Thu Feb 03, 2011 2:38 pm Post subject: Reply with quote
EvanZ wrote:
I thought wrong. Shocked Very Happy
Since 1986 there were 114 performances of 47-49 points and 99 of 50-52 points. Pretty random. I love that Basketball-Reference game finder.
Did you know...since 1986 there were three games where a player had >=10 AST and >= 10 blocks? Hakeem had two (both in 1990). Robinson had the other ('94).
Yes, but no matter how many points you're looking at, there are fewer games at N+1 than there are at N. This was part of a Bill James study that looked at how often pitchers won 20, hitters got 30 HR, etc.
Looking from 2000-present, I found:
37 287
38 224
39 174
40 157
41 122
42 92
43 87
44 61
45 46
46 30
47 26
48 23
49 6
50 22
51 19
52 10
53 8
54 5
So we'd expect maybe only half as many 50-52 as 47-49 point performances if it were a smooth distribution, and we're seeing just a small decrease, so yes, your original thought was right.
Back to top
View user's profile Send private message
Mike G
Joined: 14 Jan 2005
Posts: 3458
Location: Hendersonville, NC
PostPosted: Thu Feb 03, 2011 2:42 pm Post subject: Reply with quote
Jeff Fogle wrote:
If you accept the premise that it takes years of practice to truly master long range shooting, then it's tough to judge 3-point strategies in the NBA in the 1980's based on how we see the world now.
The NBA adopted the line for the '79-80 season.
It wasn't universal in the colleges until the 86-87 season.
Yes, and in the ABA, toward the end of the 9-year experiment, 3-pt shooting had been pretty much declared dead as a viable strategy. About as useful as the 2-point conversion in football.
Early NBA experimenters on a largish scale (through 1986) --
Griffith .330, Bird .359, Free .345, Joey Hassett .336 (early 3-pt "specialist"), MR Richardson .220 (not a typo), Macy .340, Buse .333, Theus .220, Evans .288, Bratz .305, Cooper .314 ...
http://bkref.com/tiny/gJQfN
Yes, that Coooper, who at the time was 2nd to Bird in career playoff treys.
In 1988, Danny Ainge connected on 148 -- almost 2 per game -- and the race was on.
Pre-short-arc 3pt specialists (more than 1/3 of their FGA) , chronologically: Hassett .336, Leon Wood .326, M Adams .331, Les .405, V Maxwell .320, Barros .398, Henson .432, DScott .386, Bullard .362, Iuzzolino .404, KJennings .381, Van Exel .338
Most of these guys didn't have enough game to stick around long. Some were just mad bombers without a conscience. Dennis Scott parlayed his skill into major contribution.
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
EvanZ
Joined: 22 Nov 2010
Posts: 177
PostPosted: Thu Feb 03, 2011 2:47 pm Post subject: Reply with quote
parinella wrote:
EvanZ wrote:
I thought wrong. Shocked Very Happy
Since 1986 there were 114 performances of 47-49 points and 99 of 50-52 points. Pretty random. I love that Basketball-Reference game finder.
Did you know...since 1986 there were three games where a player had >=10 AST and >= 10 blocks? Hakeem had two (both in 1990). Robinson had the other ('94).
Yes, but no matter how many points you're looking at, there are fewer games at N+1 than there are at N. This was part of a Bill James study that looked at how often pitchers won 20, hitters got 30 HR, etc.
Looking from 2000-present, I found:
37 287
38 224
39 174
40 157
41 122
42 92
43 87
44 61
45 46
46 30
47 26
48 23
49 6
50 22
51 19
52 10
53 8
54 5
So we'd expect maybe only half as many 50-52 as 47-49 point performances if it were a smooth distribution, and we're seeing just a small decrease, so yes, your original thought was right.
Of course I was right! Laughing
Thanks, that makes sense.
_________________
http://www.thecity2.com
http://www.ibb.gatech.edu/evan-zamir
Back to top
View user's profile Send private message
Jeff Fogle
Joined: 11 Jan 2011
Posts: 24
PostPosted: Thu Feb 03, 2011 3:04 pm Post subject: Reply with quote
Shin-Soo Choo is a career .297 hitter, but he has splits that suggest he'll better than that in favorable situations and even high focus situations.
.310 at home for career
.285 on road for career
.280 in first half of season
.312 in second half of season
.373 ahead in the count
.217 behind in the count
.323 in high leverage situations (as defined by baseball reference)
.283 in low leverage situations
(loosely defined as "importance" of the situation by br)
.324 vs. groundball pitchers
.266 vs. flyball pitchers
A guy hitting near .300 for a season represents a mix of a variety of situations that may or may not favor his production. I don't think it's implausible to suggest he could be the equivalent of a .340 hitter if he's facing a minor league call-up, or back of the rotation guy, or a team splitting innings amongst the bullpen...or even a name starter who's pitching to contact rather than bearing down.
A great .330 hitter doesn't go .330 vs. everyone, he obliterates bad pitching but settles in below .300 vs. elite pitchers. That's a general rule, I'm sure there are some exceptions (I think Ichiro beats out grounders successfully vs. everyone, lol).
Agree that pitching doesn't drop off as much in September as some had thought. But, this isn't a generic September day, it's the last day of the season when motivation is at its lowest level for almost everyone on the field. If there's one guy who's motivated...and he's facing vulnerable pitching it could influence his personal results, and the conclusions we draw from the results. He doesn't increase his talent level. He's the same guy. He's the same guy when others have decreased their effectiveness because they're just going through the motions.
Disagree that it's implausible. Not saying it's an absolute certainty.
Back to top
View user's profile Send private message
Guy
Joined: 02 May 2007
Posts: 104
PostPosted: Thu Feb 03, 2011 3:32 pm Post subject: Reply with quote
Quote:
Disagree that it's implausible. Not saying it's an absolute certainty.
Yes, it's totally implausible. First, .299 hitter on last day of season are probably something like .285 hitters in real talent. So for them to hit .340 would require opposing pitchers to be about 55 points worse than average, say about .325. I doubt that a single pitcher allowed to pitch for a major league team, even in game 162, is that bad. In any case, there is no possibility whatsoever that the pitchers in game 162 are anything close to being that bad.
Trust me on this -- there isn't a chance in the world this could happen. And, as I indicated, we actually know that it doesn't happen, which should count for something....
Back to top
View user's profile Send private message
iamawesomer
Joined: 01 Sep 2008
Posts: 14
PostPosted: Thu Feb 03, 2011 4:28 pm Post subject: Reply with quote
Jeff Fogle wrote:
Shin-Soo Choo is a career .297 hitter, but he has splits that suggest he'll better than that in favorable situations and even high focus situations.
.310 at home for career
.285 on road for career
.280 in first half of season
.312 in second half of season
.373 ahead in the count
.217 behind in the count
.323 in high leverage situations (as defined by baseball reference)
.283 in low leverage situations
(loosely defined as "importance" of the situation by br)
.324 vs. groundball pitchers
.266 vs. flyball pitchers
Pretty much everyone hits better at home, much better ahead in the count, against groundball pitchers (higher babip) and he has all of a little more than 300 AB in high leverage situations.
Back to top
View user's profile Send private message
John Hollinger
Joined: 14 Feb 2005
Posts: 172
PostPosted: Thu Feb 03, 2011 5:19 pm Post subject: Reply with quote
The thing about the NBA 3-point line is you needed a generation of players to come along who had developed that skill. The original NBA 3-point shot was intended to be rare and beyond the range of what a typical jumper was in that day. Players just practiced the shot, whereas prior to that time it made no sense.
I played HS basketball when they put in the 3-point rule, and what was interesting was that the players understood its value almost immediately (the college line, which we used, is close enough that good HS shooters were already shooting from roughly that range), but the coaches, having never been in that environment before, took a long time to get it.
Back to top
View user's profile Send private message AIM Address
Jeff Fogle
Joined: 11 Jan 2011
Posts: 24
PostPosted: Thu Feb 03, 2011 5:29 pm Post subject: Reply with quote
Well...I'm not awesome...I'll grant that...
There are researchers in the markets who have established to their satisfaction that the world is different enough in "get away" games to profit from. The last day of the season is the ultimate get-away day. When the mainstream sabermetric world does some studies on that...we can talk more about what's plausible for a hitter to do in a game that barely matters to anyone but him, just by showing up and being himself, when everyone else is setting up their Tuesday tee times. He'll be in a favorable hitting situation, and hitters do better than their norms in favorable hitting sitations...
PS to JH's note...I played HS ball in 79 and 80...before the line...to my utter chagrin. Agree with everyone pointing out timeline issues...
Back to top
View user's profile Send private message
Guy
Joined: 02 May 2007
Posts: 104
PostPosted: Thu Feb 03, 2011 7:26 pm Post subject: Reply with quote
Jeff Fogle wrote:
When the mainstream sabermetric world does some studies on that...we can talk more about what's plausible for a hitter to do in a game that barely matters to anyone but him, just by showing up and being himself, when everyone else is setting up their Tuesday tee times. He'll be in a favorable hitting situation, and hitters do better than their norms in favorable hitting sitations....
You don't need much of a study. Retrosheet assembles the box scores for each day of the season. For example, here is the last day of the 2010 season: http://www.retrosheet.org/boxesetc/2010/10032010.htm. On this date, the 30 teams produced 230 hits in 277 half-innings, or an average of 7.5 hits per 9 innings. But for the season, teams average 8.8 per 9. So hitters hit significantly worse, not better, in the final game. It's just one date, but it represents over 1,000 plate appearances, so it's a decent sample. If you want to check out a few more games, be my guest -- it only takes a few minutes. I guarantee you that you will find that the average final day batting average does not even approach .290, much less .340. And if you care only about the .299 guys, well Phil Birnbaum has already addressed that.
There's just nothing here......
Back to top
View user's profile Send private message
Jeff Fogle
Joined: 11 Jan 2011
Posts: 24
PostPosted: Fri Feb 04, 2011 1:07 am Post subject: Reply with quote
Yes, you'll find hitters are generally worse on all getaway days. During the regular season, pitchers are typically in their normal form on rotation in terms of effort whether it's a getaway game or not. They're taking their turn and giving it their all. Hitters swing more freely (which is important for strikeout props). Sequential offense takes a back seat to getting things over with and moving on (not for every single team but it's a good general rule).
On the last day of the season though, nobody cares in a way that's consistent with regular season norms. Or, hardly anyone. This still favors pitchers (because hitters are taking their cuts and everyone wants to go home)...but, there's less must-win intensity from everyone. Should there be a hitter who has individual motivation to achieve a goal or a contract bonus threshold or something...he maintains his normal intensity in world where hardly anyone cares (which is worth knowing for market players looking at individual player props)...then he's theoretically in a more favorable hitting situation than normal.
Teams race through a 3-1 game, but he's there for his moments. Or, maybe he's lucky enough to draw Phil Coke or Justin Germano or Dustin Moseley (or middle relievers after those guys go out) in a way that makes his life even easier.
I'm not suggesting there are a zillion hitters who will be in this position. But, it could influence a study of the specific few over time who had been in that position.
I hope we can agree to disagree at this point. Your position is very clear. I appreciate you taking time to present the case.
Back to top
View user's profile Send private message
Mogilny
Joined: 05 Aug 2010
Posts: 12
PostPosted: Fri Feb 04, 2011 6:48 am Post subject: Reply with quote
Regarding the defense wins championships in soccer mentioned earlier - I think the popular belief nowadays is that offense wins championships because of the 3-1-0 point system. Low scoring teams plays more draws which is costly when a win gives 3p and a draw only 1p compared to 2p and 1p earlier. I often read lines like "a great defensive team might win a knock out-cup but in order to win the league you need a great offense".
Note, I'm not saying this is the correct theory, only the one I hear most commonly from "experts" and the general fan. And there is a chance my selective memory is playing a trick on me too. Smile
Back to top
View user's profile Send private message
EvanZ
Joined: 22 Nov 2010
Posts: 177
PostPosted: Fri Feb 04, 2011 7:36 am Post subject: Reply with quote
LeBron scored 51 last night. He must be a lurker here. Laughing
_________________
http://www.thecity2.com
http://www.ibb.gatech.edu/evan-zamir
kjb
Joined: 03 Jan 2005
Posts: 857
Location: Washington, DC
PostPosted: Fri Feb 04, 2011 10:48 am Post subject: Reply with quote
John Hollinger wrote:
The thing about the NBA 3-point line is you needed a generation of players to come along who had developed that skill. The original NBA 3-point shot was intended to be rare and beyond the range of what a typical jumper was in that day. Players just practiced the shot, whereas prior to that time it made no sense.
I played HS basketball when they put in the 3-point rule, and what was interesting was that the players understood its value almost immediately (the college line, which we used, is close enough that good HS shooters were already shooting from roughly that range), but the coaches, having never been in that environment before, took a long time to get it.
Good points. The 3pt shot was introduced to high school for my senior season. I went to several basketball camps that preceding summer, and it was interesting to see the coaches trying to figure out what to do with it.
At 5-Star, my coach (Bill Bayno) at first had us running a fairly traditional offense, and he didn't want us shooting the three. About halfway through the week (and after several big losses) he had a change of heart. We started pushing the pace, shooting spot-up 3s in transition, playing drive-and-kick. And we started winning. Part of that was because the strategy fit the players pretty well; part of that was that it was a novel tactic at the time. Of course, we got bounced from the playoffs in the first game, so maybe it didn't work that great after all.
I think the transition to the three may have been tougher for established coaches who had success. Denny Crum comes immediately to mind as a coach who eschewed the 3pt shot while teams surpassed him by shooting it.
Back to top
View user's profile Send private message AIM Address Yahoo Messenger
schtevie
Joined: 18 Apr 2005
Posts: 392
PostPosted: Fri Feb 04, 2011 11:01 am Post subject: Reply with quote
Picking up the arguments...
Jeff Fogle wrote:
If you accept the premise that it takes years of practice to truly master long range shooting, then it's tough to judge 3-point strategies in the NBA in the 1980's based on how we see the world now.
This premise is irrelevant to the issue at hand. The issue is why was the NBA so damn slow in integrating 3 point shooing into its offenses. The fact of the matter is that there were players in 1980 who were excellent long range shooters (more on Larry et al in a minute) and their demonstrated abilities were such that the rewards of 3 point shooting should have been/were clear (unless one is prepared to believe some rather unkind things about the acuity of coaching staffs of the day).
Jeff Fogle wrote:
The NBA adopted the line for the '79-80 season.
It wasn't universal in the colleges until the 86-87 season.
High schools match the colleges
So, you didn't have a set of players who had been shooting the shot their whole lives until the 90's. That may have influenced the confidence coaches had in using that as an important part of the offense.
Again, the ABA experience belies your claims. The history of that league - what was surely generally familiar, no? - is of a pool of talented scorers, the best of whom were converting 3 pointers such that the eFG% was greater than the relevant averages and many of whom were proficient enough that the returns were higher than many, many of the shots taken in a typical offense. If you build the 3 point line, they will come. This is the essential point.
In this light, the lack of "confidence" of coaches makes no sense at all, the shooters were already there.
Jeff Fogle wrote:
Plus, Larry Bird was impossible to guard in his 20's! He was an amazing shooter, earned 400+FT attempts most years where he hit 88% or so, and he had a ton of assists. You want a guy like that driving to the basket because odds are good you'll get something...often getting something that attaches a personal foul to one of the opposing bigs too. Can see why the emphasis for Bird would be attacking the basket until his body couldn't take it any more. High trey attempts over a period of years didn't begin until he was 29 or so.
http://www.basketball-reference.com/pla ... dla01.html
I would urge you to return to this page and revisit the facts, because I think they tell a rather different story than the one you relate. And more generally, as went Larry, so went the league. (Actually Larry lagged the league, which I add for alliterative pleasure, and because it aids the narrative.)
Larry Bird, as we all know, entered the NBA with the 3 point rule. And he, like the league, apparently enjoyed the novelty. Larry shot many more 3 pointers that year (as a share of FGA) than in any of the four following years. An inexplicable consequence, really, given that he hit 40% of them. But the issue raised is whether he cut down because he found better things to do with his possessions. And I think the clear answer is "no". Against a baseline expectation of general improvement with experience, the distinct improvement awaits the 1984-85 and 1985-86 campaigns, where his long-range game returns to the importance of his rookie year and non-coincidentally, his FTA/2FGA improves. (Nothing like the threat of being able to step back for a three to open up the floor and draw fouls.)
But this aside, there is the more interesting issue of Larry as representing, as well as lagging, the league. That his 3 point talents weren't utilized in his early career, to the detriment of the Celtics, is one thing. But this technological retrogression was a league-wide phenomenon. After the novelty (and success) of the first year, 3 point shooting largely disappeared from the league, exceeding the initial year's 3.1% of FGAs only five years later (3.5%). Same with Larry, but worse! He started off with 9.8% of his FGAs, and only six years later did he regain (and surpass) that level (12.1%).
The point is that Larry is illustrative in terms of being the revolutionary who wasn't, but the ostrich-like tendency to look away from the potential of the 3 point shot was a league-wide phenomenon.
Mike G wrote:
Yes, and in the ABA, toward the end of the 9-year experiment, 3-pt shooting had been pretty much declared dead as a viable strategy. About as useful as the 2-point conversion in football.
What can I say? If the point is that this is what was the conventional wisdom, I guess I agree. But it was wrong. Full stop. A fact as plain as day for anyone back then who could multiply by 1.5 and had a basic intuitive sense of the relative worth of the range of observed scoring opportunities. Reality is a bitch sometimes.
John Hollinger wrote:
The thing about the NBA 3-point line is you needed a generation of players to come along who had developed that skill. The original NBA 3-point shot was intended to be rare and beyond the range of what a typical jumper was in that day. Players just practiced the shot, whereas prior to that time it made no sense.
Why you hate Larry Bird?! Don't hate Larry Bird! The story of Chris Ford is also interesting from that same first year of the three point shot. He killed the shot to the tune of shooting 0.427, and then in the last two years of his career that followed, he suffered the consequences of the Dark Ages alongside his far more storied teammate.
But clearly he should have had a sense of the potential value, no? No evidence of that in his stretch as coach of the Cs in the early 90s however, not that he was blessed with a lot of raw material to work with (but did he make it a priority?) And then Larry was coach of the Pacers which had a lot, a lot of long range talent. It is a superficial reading, but it isn't clear to me that he pushed the curve with it, in terms of maximizing attempts.
Conventional wisdom is a tyrant.
John Hollinger wrote:
I played HS basketball when they put in the 3-point rule, and what was interesting was that the players understood its value almost immediately (the college line, which we used, is close enough that good HS shooters were already shooting from roughly that range), but the coaches, having never been in that environment before, took a long time to get it.
Doesn't make much sense, does it?
Back to top
View user's profile Send private message
Mike G
Joined: 14 Jan 2005
Posts: 3458
Location: Hendersonville, NC
PostPosted: Fri Feb 04, 2011 1:06 pm Post subject: Reply with quote
schtevie wrote:
Mike G wrote:
...in the ABA, toward the end of the 9-year experiment, 3-pt shooting had been pretty much declared dead as a viable strategy. About as useful as the 2-point conversion in football.
What can I say? If the point is that this is what was the conventional wisdom, I guess I agree. But it was wrong. Full stop. A fact as plain as day for anyone back then who could multiply by 1.5 and had a basic intuitive sense of the relative worth of the range of observed scoring opportunities. Reality is a bitch sometimes.
The last ABA player to attempt as many as 50 threes in a season and make over 35% was George McGinnis in 1975.
http://bkref.com/tiny/dB0j1
Sort by season.
Of a total 49 ABA player-seasons with >50 3FGA, the median 3fg% was .314 .
Multiplying by 1.5 gives us .457 as an eFG% on threes, when taken by a league's premier shooters.
This was a 9-year experiment undertaken by several hundred players and a few dozens of coaches.
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
schtevie
Joined: 18 Apr 2005
Posts: 392
PostPosted: Fri Feb 04, 2011 1:54 pm Post subject: Reply with quote
Mike, I believe that confusion reigns on multiple levels, and hopefully it is not mine.
First of all, the threshold you consider relevant, is it? Let's take the last year of the ABA, just as the closest bridge year. No cherry-picking. The 2 point fg% that year was 47.3%. So, the effective 3 point equivalent would be 31.5%, right? And that is only if you consider the threshold for three point shooting to be the average. It clearly isn't. The threshold is the opportunity cost, the worst alternative, which as we know (we do, right?) is much lower than that. The same class of shots that are still the worst today, mid to long range jumpers, contested more than not.
But let's take 31.5% as the threshold anyway, setting the bar high to see if we can clear it. And let's take the 50 attempts reference point as well, though I am by no means arguing that the ABA was enlightened in integrating 3 point shooting into its offenses. My only point in invoking the defunct league is that there was a history to ponder than should have been familiar to the NBA.
So, ABA, 3fg% > or = 31.5% and 50 attempts min., basketball reference here we come:
1975-76: 6 players
1974-75: 9 players
1973-74: 11 players
1972-73: 11 players
1971-72: 12 players
1970-71: 12 players
1969-70: 8 players
1968-69: 11 players
1967-68: 8 players
So, you are looking at about one guy per team on average, but many of these players shot a lot higher percentage than 31.5% and had a lot more than 50 attempts. I don't know why your B-R query left you short, but there are the data.
And for comparison, in the NBA's first 3 point year, there were 12 players taking 50 attempts (where there analogous threshold was 32.5%), roughly one player for every two teams, compared to about one per team, on average, in the ABA.
Back to top
View user's profile Send private message
Mike G
Joined: 14 Jan 2005
Posts: 3458
Location: Hendersonville, NC
PostPosted: Sat Feb 05, 2011 9:04 am Post subject: Reply with quote
OUCH! My bad. Player Season Finder set to Forwards only.
Try again -- top 100 ABA 3fga seasons lists those with >118 3fga.
Sorting by 3P%, the median is .320
Sorting by Season, just one player (Billy Shepherd, '75) hit over .354 after 1974.
Just 4 of the 100 'prolific' 3-shooters are in the list from the final ABA season.
Quote:
I am by no means arguing that the ABA was enlightened in integrating 3 point shooting into its offenses.
Louie Dampier (Ky), from '69 onward always managed to hit at least .348 from the arc (.358 career). He had 550 attempts/yr before Issel came, then cut that in half, and when Gilmore arrived dropped further every year.
Even as his 3fg% rose to near .400, his 3 attempts dropped below 10% of his FGA. Coach Hubie Brown apparently didn't think much of it, except perhaps as a distraction.
Of course, not everyone has a Gilmore in the painted area, nor also an Issel in the wings. But everyone gets more trips to the FT line from shots inside the arc. And these shots skew quite a bit the effectiveness of the average 2 vs the avg 3.
If a team's TS% is .540, then shouldn't a 3fg% of 2/3 that rate be the threshold? This would be .360, and it would be a long time before teams stocked players with that ability -- and/or with adequate additional basketball skills.
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
schtevie
Joined: 18 Apr 2005
Posts: 392
PostPosted: Sat Feb 05, 2011 3:38 pm Post subject: Reply with quote
Mike G wrote:
.....But everyone gets more trips to the FT line from shots inside the arc. And these shots skew quite a bit the effectiveness of the average 2 vs the avg 3.
If a team's TS% is .540, then shouldn't a 3fg% of 2/3 that rate be the threshold? This would be .360, and it would be a long time before teams stocked players with that ability -- and/or with adequate additional basketball skills.
Um. No. Definitely not.
This is not how the concept of a threshold works, given the realities of offensive basketball.
On average, ex post, one expects the average productivity/efficiency to be realized. But that has essentially nothing to do with actual choices faced, ex ante, on a given possession, where the threshold concept is operative.
You start a possession (and the shot clock) hoping for luck and good fortune, that a good shot attempt around the basket can be realized, one that yields, in expectation, a way above average shot. This happens relatively infrequently however (otherwise all shots would be above average, and....) and offenses are compelled to take a large fraction of below average shots.
Given that the basic physical dimensions of the basketball court haven't changed over the relevant history, nor has the reality that defense extends out from the basket, contemporary data about the relative returns of shots from different locations can inform our views of the ABA and 1980s NBA past (http://hoopdata.com/teamshotlocs.aspx and http://www.82games.com/locations.htm).
Accordingly, the relevant threshold for a 3 point attempt was and is approximately that of a mid-range to long 2, and these aren't (and likely weren't) fouled all that much (so no large FT bonus should be added). The bottom line being: the relevant 3 point threshold was and is well below the average TS%. And as such, there has always been an "excess supply" of 3 point shooting ability in the league.
THE 3 POINT SHOT HAS EVER AND ALWAYS, AT A MINIMUM, BEEN AN ABOVE AVERAGE, BELOW AVERAGE SHOT IN ITS POTENTIAL.
As this discussion continues, I guess I am obliged to revise my expectations slightly. Maybe coaches back then really couldn't multiply by 1.5 and really didn't have a sense how unproductive jump shooting was. But a lack of the ability to keep score in a game where score is everything seems such an odd assumption.
Back to top
View user's profile Send private message
Mike G
Joined: 14 Jan 2005
Posts: 3458
Location: Hendersonville, NC
PostPosted: Mon Feb 07, 2011 6:15 am Post subject: Reply with quote
The dimensions of the court are the same as ever, but the effective reach of the defense has seemingly expanded.
In the ABA, and in the post-merger NBA, the long 2-pointer was always available. As was the 3-pointer. But it also wasn't hard to get a shot from 16-18 feet. Without the 3 as an incentive, the vast majority of players' shooting ranges drop off dramatically somewhere before 20'.
In the mid-80s, it was rare to see perimeter passing as far as 25' out, and now it's standard; even bigs who are never going to shoot from there are out there, because that's where the offense is happening.
If milling around the 3pt line is taking you out of the offense, you had better be a damn good 3pt shooter. Lots better than 32%. You aren't going to set many picks, get many offensive rebounds, get many assists out there.
At some point -- the Bad Boy Pistons era, perhaps -- the open 16-18 foot shot had become so rare that offenses had 2 choices: drive it inside, or set up at the arc. Thus the 3-pt era was created. Not so much from great insight, as from necessity.
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
schtevie
Joined: 18 Apr 2005
Posts: 392
PostPosted: Mon Feb 07, 2011 3:34 pm Post subject: Reply with quote
I very much look forward to Ed K reprising his chapter reviews, but until then, let me try to better establish the fact that opinions on the shape of the earth should not be allowed to differ when it comes to the history of the adoption of 3 point shooting in the NBA. There can no fundamental ambiguity. Taking a broad brush (and apologies to the unappreciated best practitioners of their day) NBA coaches, players, all relevant decision makers were, collectively, horrible innovators. Many, many points, and dramatic competitive advantage for the innovators who weren't, were foregone.
Invoking changing (improving) defenses over the relevant history doesn't impact this fundamental determination. Arguing that there were supply side failures is a non-starter. The relevant shortcomings were all on the demand side. For ignorance or strategic risk-aversion (aka keeping one's head down) big mistakes were made, persisting over decades.
I pursue this pedantically, not because NBA empirical history should necessarily be of interest to those interested in stats (though one would be a better person accordingly) but because the belief that we have ever and always been in a quasi-optimal state of best practice is false and influences current thinking. It was screamingly false regarding the 3 point shot way back when, but all that has changed to the present day is a question of degree. The 3 point story doesn't end in the 80s. And if this, what else? This should be the informed prior.
OK.
The threshold. It is probably very fair to say that the reasonable opportunity cost of a 3PA throughout NBA history has been about 0.400. Why? As noted, the fundamental reality of defense extending out from the basket hasn't changed since the game began. This is a very useful fact. Roland posted some extreme charting summaries (previously linked) for the 2004-05 season (I believe). These tell a very clear, and ultimately intuitive story. Comparing a straight average of the three categories that comprise his zones labeled 6 through 10, we have a FG% of .426. And we also note that these long (and some not-long 2s) had a straight average "play" efficiency of 0.394 (Roland labels this column "Pts per 100 poss" but I am pretty sure he means per 100 plays, according to the Oliverian nomenclature.) This latter concept is very close to the actual decision threshold. If I am to take a longish 2, I expect a certain augment for being fouled (in the instance, including being merely in the zone, not just attempting a shot...I believe that this is a correct interpretation of the data) as well as a certain decrement for the possibility of turning the ball over and getting no points. So, in 2004-05, 0.400 was a useful opportunity cost.
And this is essentially corroborated by the more contemporary shot location data at hoopdata, that shows that from 10 feet out to the 3 point line the average FG% is about 0.400.
Now, is there any reason to believe that this benchmark has changed over the years? I don't think so, but I would be interested in hearing alternative views. Going back to 1979-80, when the 3 point shot came to be in the NBA, the average offensive efficiency was basically the same. The reality of defenses extending out was the same. The basic rationale that high quality shots are the ones worthy of fouling was surely the same. Two differences that might be noted is that FG% was higher and FT/FGA was lower back in the day, but that is consistent with the story of later defensive tightening on high quality shots (i.e. "no layups" allowed) and suggests nothing about long 2s. These weren't fouled much in 2004-05, and it is highly implausible to believe that they were fouled more (when fouling was generally less) back in the early NBA shot clock era.
So until persuasively argued otherwise, a threshold of 0.400 it is. Bump it up a bit if you choose, no matter.
Next issue: was slow adoption due to demand or supply?
I repeat the NBA's initial year's success rate on 3s: 0.280 or a eFG% of 0.420. "Ah", the skeptic might say, "Right at the threshold. See. Quasi-optimality prevailed." And in a sense, yes. The average use was just at (above) the threshold. Call it a baseline. But it is important to realize that in the initial, half-assed (usage rate compared to the ABA) "implementation", there were inevitably buzzer beating heaves, not consciously integrated as part of an offense. But call it a baseline, and we move on to year two, 1980-81, the beginning of the mini-Dark Age of the 3 point shot.
Attempts dropped from 2.8 to 2 per game, and the success rate dropped to 0.245, making the realized eFG% 0.37, now discernibly below the alleged threshold. But what of the marginal change? Tracking this, you see that what was culled from the previous year's offense were the 3 point shots having an eFG% of 0.566! Technological retrogression indeed.
And this story, of "marginal" yearly changes being well above the 0.400 threshold, indeed often well above overall offensive efficiency, is the crucial fact in understanding the history of the 3 point shot. Despite realized changes being incredibly effective, nothing but the most tentative steps in implementation.
Year "Marginal" eFG% (* denotes shorted 3 point line)
1981, 0.566
1982, 0.571
1983, 3.750
1984, 0.750
1985, 0.581
1986, 0.353
1987, 0.526
1988, 0.886
1989, 0.508
1990, 2.250
1991, 0.267
1992, 0.750
1993, 0.556
1994, 0.448
1995, 0.608*
1996, 0.811*
1997, 0.320*
1998, 0.609
1999, 0.535
2000, 0.559
2001, 0.000
2002, 0.547
2003, 2.100
2004, 0.300
2005, 0.750
2006, 0.917
2007, 0.532
2008, 0.609
2009, 12.000
2010, -27.000
Pretty much no matter the year (the few odd numbers mostly reflecting small usage changes and rules changes) and encompassing what is generally acclaimed to be the era of improved team defense, there was adequate supply of "3 point shooting ability", defined as the combination of raw shooting talent and the ability to find its effective use in the offenses of the day. (And completely unmentioned so far in the argument is that demand could have easily called forth additional supply. There was never any impediment, and all incentives would have been perfectly aligned, for coaches to have instructed their players in the off-season to work intensively on their long-range shooting.)
Eyeballing this data, we are talking 0.3 points per shot transferred from a long 2 to a considered 3 pointer (and likely more in the sense that one first culls the contested long 2s by mediocre shooters from the offense, not the wide open ones by the more skilled). A huge opportunity, always available, and decisively taken advantage of by none.
And still, 30 years on, and its not yet clear that the optimal usage rate of the 3 pointer has been achieved in the league.
Back to top
View user's profile Send private message
Mike G
Joined: 14 Jan 2005
Posts: 3458
Location: Hendersonville, NC
PostPosted: Tue Feb 08, 2011 9:31 am Post subject: Reply with quote
schtevie wrote:
...
I pursue this pedantically... because the belief that we have ever and always been in a quasi-optimal state of best practice is false and influences current thinking. It was screamingly false regarding the 3 point shot way back when, ...
Perhaps nobody really believes in such a thing?
Nevertheless:
In 9 years, the ABA had 92 team-seasons, not counting teams that folded early in some year.
The median 3PA is 350.
The 46 teams shooting more than 350 3's in a year had a median Win% of .470; 9 teams over .600 and 13 teams under .400 .
Those attempting fewer than 350 had a median W% of .540, with 15 over .600 and 9 under .400 .
http://bkref.com/tiny/sFZ5j
In the years 1980-85, there are 137 NBA team-seasons, with a median team 3PA of 187-188.
Of 68 teams shooting 188 or more threes, the median W% is .476
Of 69 trying fewer than 188, the median is .537.
Where in this broad brush view is evidence that shooting more 3's is a good strategy?
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
Joined: 02 May 2007
Posts: 104
PostPosted: Thu Feb 03, 2011 11:58 am Post subject: Reply with quote
schtevie wrote:
Ed was referring to a reluctance of coaches to adopt risky strategies for apparent, potential gain. I was making a much stronger point: that coaches throughout NBA history have neglected to adopt non-risky strategies that had immediate and ginormous gain.
It seems to me that taking more 3-point shots is indeed a riskier strategy, in that it increases the probability of a bad short-term outcome (zero points on the possession). That is similar to the 4th-down play in football: punting decreases the chance of a very bad immediate outcome, i.e. turning the ball over at current field position.
Was it really obvious in the 1980s that players should take more 3s? Did teams have good data on the success rate on long-2s, to allow that comparison? I'm not disagreeing, I just don't know the relevant history here.
Back to top
View user's profile Send private message
Jeff Fogle
Joined: 11 Jan 2011
Posts: 24
PostPosted: Thu Feb 03, 2011 1:28 pm Post subject: Reply with quote
If you accept the premise that it takes years of practice to truly master long range shooting, then it's tough to judge 3-point strategies in the NBA in the 1980's based on how we see the world now.
The NBA adopted the line for the '79-80 season.
It wasn't universal in the colleges until the 86-87 season.
High schools match the colleges
So, you didn't have a set of players who had been shooting the shot their whole lives until the 90's. That may have influenced the confidence coaches had in using that as an important part of the offense.
Plus, Larry Bird was impossible to guard in his 20's! He was an amazing shooter, earned 400+FT attempts most years where he hit 88% or so, and he had a ton of assists. You want a guy like that driving to the basket because odds are good you'll get something...often getting something that attaches a personal foul to one of the opposing bigs too. Can see why the emphasis for Bird would be attacking the basket until his body couldn't take it any more. High trey attempts over a period of years didn't begin until he was 29 or so.
http://www.basketball-reference.com/pla ... dla01.html
Feel fortunate to have seen a lot of him on TV. Indiana State ended up getting a lot of "Game of the Week" telecasts across the middle section of the country when he was there (even down in Texas). Was that the Hughes Sports Network? Something like that. No three's in the college game at the time, so it would have been odd for him to emphasize three's as a pro until he had to.
Quick note on the .299/.300 baseball discussion. Worth considering I think that pitching quality may not be all that intense on the last day of the regular season. Usually a meaningless game...possibly a September call-up or late rotation guy throwing...possibly in a "pitch-to-contact" way rather than bearing down. Even if the hitter can't make himself a better hitter on command, he might be facing a less challenging task then during a typical regular season at bat. He's a .299 hitter for the season, but a .340 hitter vs. a September call-up or a back-of-the-rotation guy throwing to contact as both teams are just finishing out the season in front of a three-quarters empty stadium...
On fourth down stuff in football. I'm pretty confident that defenses would come up with some very creative attacks if they knew in advance that a team was going to go for it on all fourth downs. Maybe that doesn't matter at lower levels. But, among elite college programs and the NFL,you have great athletes disrupting plays. A chance to disrupt a play deep in opposing territory doesn't seem like a big mystery to solve. The studies I've seen in favor of going for it on 4th down all the time tend to use averages from past fourth down plays. Past 4th down plays weren't deep in a team's own territory generally, and often came agaisnt a prevent when the game was out of reach. Game in reach? Different set of circumstances, and successful defensive strategies would likely evolve even if they don't exist already. Defenses generally control the flow of football games if you look at success on a drive-by-drive basis.
Back to top
View user's profile Send private message
Guy
Joined: 02 May 2007
Posts: 104
PostPosted: Thu Feb 03, 2011 1:43 pm Post subject: Reply with quote
Quote:
Quick note on the .299/.300 baseball discussion. Worth considering I think that pitching quality may not be all that intense on the last day of the regular season....He's a .299 hitter for the season, but a .340 hitter vs. a September call-up or a back-of-the-rotation guy.
No, that's not plausible -- .340 vs. .299 is huge. It's the difference between an average hitter and a guy who can't make the majors, or an average hitter and a Hall of Famer. Pitching quality does not change remotely that much in September. And in any case, Phil Birnbaum actually looked at the data, and hitters don't improve at all (without any control for pitcher quality). There's just nothing there.....
Back to top
View user's profile Send private message
parinella
Joined: 16 Dec 2005
Posts: 10
PostPosted: Thu Feb 03, 2011 2:38 pm Post subject: Reply with quote
EvanZ wrote:
I thought wrong. Shocked Very Happy
Since 1986 there were 114 performances of 47-49 points and 99 of 50-52 points. Pretty random. I love that Basketball-Reference game finder.
Did you know...since 1986 there were three games where a player had >=10 AST and >= 10 blocks? Hakeem had two (both in 1990). Robinson had the other ('94).
Yes, but no matter how many points you're looking at, there are fewer games at N+1 than there are at N. This was part of a Bill James study that looked at how often pitchers won 20, hitters got 30 HR, etc.
Looking from 2000-present, I found:
37 287
38 224
39 174
40 157
41 122
42 92
43 87
44 61
45 46
46 30
47 26
48 23
49 6
50 22
51 19
52 10
53 8
54 5
So we'd expect maybe only half as many 50-52 as 47-49 point performances if it were a smooth distribution, and we're seeing just a small decrease, so yes, your original thought was right.
Back to top
View user's profile Send private message
Mike G
Joined: 14 Jan 2005
Posts: 3458
Location: Hendersonville, NC
PostPosted: Thu Feb 03, 2011 2:42 pm Post subject: Reply with quote
Jeff Fogle wrote:
If you accept the premise that it takes years of practice to truly master long range shooting, then it's tough to judge 3-point strategies in the NBA in the 1980's based on how we see the world now.
The NBA adopted the line for the '79-80 season.
It wasn't universal in the colleges until the 86-87 season.
Yes, and in the ABA, toward the end of the 9-year experiment, 3-pt shooting had been pretty much declared dead as a viable strategy. About as useful as the 2-point conversion in football.
Early NBA experimenters on a largish scale (through 1986) --
Griffith .330, Bird .359, Free .345, Joey Hassett .336 (early 3-pt "specialist"), MR Richardson .220 (not a typo), Macy .340, Buse .333, Theus .220, Evans .288, Bratz .305, Cooper .314 ...
http://bkref.com/tiny/gJQfN
Yes, that Coooper, who at the time was 2nd to Bird in career playoff treys.
In 1988, Danny Ainge connected on 148 -- almost 2 per game -- and the race was on.
Pre-short-arc 3pt specialists (more than 1/3 of their FGA) , chronologically: Hassett .336, Leon Wood .326, M Adams .331, Les .405, V Maxwell .320, Barros .398, Henson .432, DScott .386, Bullard .362, Iuzzolino .404, KJennings .381, Van Exel .338
Most of these guys didn't have enough game to stick around long. Some were just mad bombers without a conscience. Dennis Scott parlayed his skill into major contribution.
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
EvanZ
Joined: 22 Nov 2010
Posts: 177
PostPosted: Thu Feb 03, 2011 2:47 pm Post subject: Reply with quote
parinella wrote:
EvanZ wrote:
I thought wrong. Shocked Very Happy
Since 1986 there were 114 performances of 47-49 points and 99 of 50-52 points. Pretty random. I love that Basketball-Reference game finder.
Did you know...since 1986 there were three games where a player had >=10 AST and >= 10 blocks? Hakeem had two (both in 1990). Robinson had the other ('94).
Yes, but no matter how many points you're looking at, there are fewer games at N+1 than there are at N. This was part of a Bill James study that looked at how often pitchers won 20, hitters got 30 HR, etc.
Looking from 2000-present, I found:
37 287
38 224
39 174
40 157
41 122
42 92
43 87
44 61
45 46
46 30
47 26
48 23
49 6
50 22
51 19
52 10
53 8
54 5
So we'd expect maybe only half as many 50-52 as 47-49 point performances if it were a smooth distribution, and we're seeing just a small decrease, so yes, your original thought was right.
Of course I was right! Laughing
Thanks, that makes sense.
_________________
http://www.thecity2.com
http://www.ibb.gatech.edu/evan-zamir
Back to top
View user's profile Send private message
Jeff Fogle
Joined: 11 Jan 2011
Posts: 24
PostPosted: Thu Feb 03, 2011 3:04 pm Post subject: Reply with quote
Shin-Soo Choo is a career .297 hitter, but he has splits that suggest he'll better than that in favorable situations and even high focus situations.
.310 at home for career
.285 on road for career
.280 in first half of season
.312 in second half of season
.373 ahead in the count
.217 behind in the count
.323 in high leverage situations (as defined by baseball reference)
.283 in low leverage situations
(loosely defined as "importance" of the situation by br)
.324 vs. groundball pitchers
.266 vs. flyball pitchers
A guy hitting near .300 for a season represents a mix of a variety of situations that may or may not favor his production. I don't think it's implausible to suggest he could be the equivalent of a .340 hitter if he's facing a minor league call-up, or back of the rotation guy, or a team splitting innings amongst the bullpen...or even a name starter who's pitching to contact rather than bearing down.
A great .330 hitter doesn't go .330 vs. everyone, he obliterates bad pitching but settles in below .300 vs. elite pitchers. That's a general rule, I'm sure there are some exceptions (I think Ichiro beats out grounders successfully vs. everyone, lol).
Agree that pitching doesn't drop off as much in September as some had thought. But, this isn't a generic September day, it's the last day of the season when motivation is at its lowest level for almost everyone on the field. If there's one guy who's motivated...and he's facing vulnerable pitching it could influence his personal results, and the conclusions we draw from the results. He doesn't increase his talent level. He's the same guy. He's the same guy when others have decreased their effectiveness because they're just going through the motions.
Disagree that it's implausible. Not saying it's an absolute certainty.
Back to top
View user's profile Send private message
Guy
Joined: 02 May 2007
Posts: 104
PostPosted: Thu Feb 03, 2011 3:32 pm Post subject: Reply with quote
Quote:
Disagree that it's implausible. Not saying it's an absolute certainty.
Yes, it's totally implausible. First, .299 hitter on last day of season are probably something like .285 hitters in real talent. So for them to hit .340 would require opposing pitchers to be about 55 points worse than average, say about .325. I doubt that a single pitcher allowed to pitch for a major league team, even in game 162, is that bad. In any case, there is no possibility whatsoever that the pitchers in game 162 are anything close to being that bad.
Trust me on this -- there isn't a chance in the world this could happen. And, as I indicated, we actually know that it doesn't happen, which should count for something....
Back to top
View user's profile Send private message
iamawesomer
Joined: 01 Sep 2008
Posts: 14
PostPosted: Thu Feb 03, 2011 4:28 pm Post subject: Reply with quote
Jeff Fogle wrote:
Shin-Soo Choo is a career .297 hitter, but he has splits that suggest he'll better than that in favorable situations and even high focus situations.
.310 at home for career
.285 on road for career
.280 in first half of season
.312 in second half of season
.373 ahead in the count
.217 behind in the count
.323 in high leverage situations (as defined by baseball reference)
.283 in low leverage situations
(loosely defined as "importance" of the situation by br)
.324 vs. groundball pitchers
.266 vs. flyball pitchers
Pretty much everyone hits better at home, much better ahead in the count, against groundball pitchers (higher babip) and he has all of a little more than 300 AB in high leverage situations.
Back to top
View user's profile Send private message
John Hollinger
Joined: 14 Feb 2005
Posts: 172
PostPosted: Thu Feb 03, 2011 5:19 pm Post subject: Reply with quote
The thing about the NBA 3-point line is you needed a generation of players to come along who had developed that skill. The original NBA 3-point shot was intended to be rare and beyond the range of what a typical jumper was in that day. Players just practiced the shot, whereas prior to that time it made no sense.
I played HS basketball when they put in the 3-point rule, and what was interesting was that the players understood its value almost immediately (the college line, which we used, is close enough that good HS shooters were already shooting from roughly that range), but the coaches, having never been in that environment before, took a long time to get it.
Back to top
View user's profile Send private message AIM Address
Jeff Fogle
Joined: 11 Jan 2011
Posts: 24
PostPosted: Thu Feb 03, 2011 5:29 pm Post subject: Reply with quote
Well...I'm not awesome...I'll grant that...
There are researchers in the markets who have established to their satisfaction that the world is different enough in "get away" games to profit from. The last day of the season is the ultimate get-away day. When the mainstream sabermetric world does some studies on that...we can talk more about what's plausible for a hitter to do in a game that barely matters to anyone but him, just by showing up and being himself, when everyone else is setting up their Tuesday tee times. He'll be in a favorable hitting situation, and hitters do better than their norms in favorable hitting sitations...
PS to JH's note...I played HS ball in 79 and 80...before the line...to my utter chagrin. Agree with everyone pointing out timeline issues...
Back to top
View user's profile Send private message
Guy
Joined: 02 May 2007
Posts: 104
PostPosted: Thu Feb 03, 2011 7:26 pm Post subject: Reply with quote
Jeff Fogle wrote:
When the mainstream sabermetric world does some studies on that...we can talk more about what's plausible for a hitter to do in a game that barely matters to anyone but him, just by showing up and being himself, when everyone else is setting up their Tuesday tee times. He'll be in a favorable hitting situation, and hitters do better than their norms in favorable hitting sitations....
You don't need much of a study. Retrosheet assembles the box scores for each day of the season. For example, here is the last day of the 2010 season: http://www.retrosheet.org/boxesetc/2010/10032010.htm. On this date, the 30 teams produced 230 hits in 277 half-innings, or an average of 7.5 hits per 9 innings. But for the season, teams average 8.8 per 9. So hitters hit significantly worse, not better, in the final game. It's just one date, but it represents over 1,000 plate appearances, so it's a decent sample. If you want to check out a few more games, be my guest -- it only takes a few minutes. I guarantee you that you will find that the average final day batting average does not even approach .290, much less .340. And if you care only about the .299 guys, well Phil Birnbaum has already addressed that.
There's just nothing here......
Back to top
View user's profile Send private message
Jeff Fogle
Joined: 11 Jan 2011
Posts: 24
PostPosted: Fri Feb 04, 2011 1:07 am Post subject: Reply with quote
Yes, you'll find hitters are generally worse on all getaway days. During the regular season, pitchers are typically in their normal form on rotation in terms of effort whether it's a getaway game or not. They're taking their turn and giving it their all. Hitters swing more freely (which is important for strikeout props). Sequential offense takes a back seat to getting things over with and moving on (not for every single team but it's a good general rule).
On the last day of the season though, nobody cares in a way that's consistent with regular season norms. Or, hardly anyone. This still favors pitchers (because hitters are taking their cuts and everyone wants to go home)...but, there's less must-win intensity from everyone. Should there be a hitter who has individual motivation to achieve a goal or a contract bonus threshold or something...he maintains his normal intensity in world where hardly anyone cares (which is worth knowing for market players looking at individual player props)...then he's theoretically in a more favorable hitting situation than normal.
Teams race through a 3-1 game, but he's there for his moments. Or, maybe he's lucky enough to draw Phil Coke or Justin Germano or Dustin Moseley (or middle relievers after those guys go out) in a way that makes his life even easier.
I'm not suggesting there are a zillion hitters who will be in this position. But, it could influence a study of the specific few over time who had been in that position.
I hope we can agree to disagree at this point. Your position is very clear. I appreciate you taking time to present the case.
Back to top
View user's profile Send private message
Mogilny
Joined: 05 Aug 2010
Posts: 12
PostPosted: Fri Feb 04, 2011 6:48 am Post subject: Reply with quote
Regarding the defense wins championships in soccer mentioned earlier - I think the popular belief nowadays is that offense wins championships because of the 3-1-0 point system. Low scoring teams plays more draws which is costly when a win gives 3p and a draw only 1p compared to 2p and 1p earlier. I often read lines like "a great defensive team might win a knock out-cup but in order to win the league you need a great offense".
Note, I'm not saying this is the correct theory, only the one I hear most commonly from "experts" and the general fan. And there is a chance my selective memory is playing a trick on me too. Smile
Back to top
View user's profile Send private message
EvanZ
Joined: 22 Nov 2010
Posts: 177
PostPosted: Fri Feb 04, 2011 7:36 am Post subject: Reply with quote
LeBron scored 51 last night. He must be a lurker here. Laughing
_________________
http://www.thecity2.com
http://www.ibb.gatech.edu/evan-zamir
kjb
Joined: 03 Jan 2005
Posts: 857
Location: Washington, DC
PostPosted: Fri Feb 04, 2011 10:48 am Post subject: Reply with quote
John Hollinger wrote:
The thing about the NBA 3-point line is you needed a generation of players to come along who had developed that skill. The original NBA 3-point shot was intended to be rare and beyond the range of what a typical jumper was in that day. Players just practiced the shot, whereas prior to that time it made no sense.
I played HS basketball when they put in the 3-point rule, and what was interesting was that the players understood its value almost immediately (the college line, which we used, is close enough that good HS shooters were already shooting from roughly that range), but the coaches, having never been in that environment before, took a long time to get it.
Good points. The 3pt shot was introduced to high school for my senior season. I went to several basketball camps that preceding summer, and it was interesting to see the coaches trying to figure out what to do with it.
At 5-Star, my coach (Bill Bayno) at first had us running a fairly traditional offense, and he didn't want us shooting the three. About halfway through the week (and after several big losses) he had a change of heart. We started pushing the pace, shooting spot-up 3s in transition, playing drive-and-kick. And we started winning. Part of that was because the strategy fit the players pretty well; part of that was that it was a novel tactic at the time. Of course, we got bounced from the playoffs in the first game, so maybe it didn't work that great after all.
I think the transition to the three may have been tougher for established coaches who had success. Denny Crum comes immediately to mind as a coach who eschewed the 3pt shot while teams surpassed him by shooting it.
Back to top
View user's profile Send private message AIM Address Yahoo Messenger
schtevie
Joined: 18 Apr 2005
Posts: 392
PostPosted: Fri Feb 04, 2011 11:01 am Post subject: Reply with quote
Picking up the arguments...
Jeff Fogle wrote:
If you accept the premise that it takes years of practice to truly master long range shooting, then it's tough to judge 3-point strategies in the NBA in the 1980's based on how we see the world now.
This premise is irrelevant to the issue at hand. The issue is why was the NBA so damn slow in integrating 3 point shooing into its offenses. The fact of the matter is that there were players in 1980 who were excellent long range shooters (more on Larry et al in a minute) and their demonstrated abilities were such that the rewards of 3 point shooting should have been/were clear (unless one is prepared to believe some rather unkind things about the acuity of coaching staffs of the day).
Jeff Fogle wrote:
The NBA adopted the line for the '79-80 season.
It wasn't universal in the colleges until the 86-87 season.
High schools match the colleges
So, you didn't have a set of players who had been shooting the shot their whole lives until the 90's. That may have influenced the confidence coaches had in using that as an important part of the offense.
Again, the ABA experience belies your claims. The history of that league - what was surely generally familiar, no? - is of a pool of talented scorers, the best of whom were converting 3 pointers such that the eFG% was greater than the relevant averages and many of whom were proficient enough that the returns were higher than many, many of the shots taken in a typical offense. If you build the 3 point line, they will come. This is the essential point.
In this light, the lack of "confidence" of coaches makes no sense at all, the shooters were already there.
Jeff Fogle wrote:
Plus, Larry Bird was impossible to guard in his 20's! He was an amazing shooter, earned 400+FT attempts most years where he hit 88% or so, and he had a ton of assists. You want a guy like that driving to the basket because odds are good you'll get something...often getting something that attaches a personal foul to one of the opposing bigs too. Can see why the emphasis for Bird would be attacking the basket until his body couldn't take it any more. High trey attempts over a period of years didn't begin until he was 29 or so.
http://www.basketball-reference.com/pla ... dla01.html
I would urge you to return to this page and revisit the facts, because I think they tell a rather different story than the one you relate. And more generally, as went Larry, so went the league. (Actually Larry lagged the league, which I add for alliterative pleasure, and because it aids the narrative.)
Larry Bird, as we all know, entered the NBA with the 3 point rule. And he, like the league, apparently enjoyed the novelty. Larry shot many more 3 pointers that year (as a share of FGA) than in any of the four following years. An inexplicable consequence, really, given that he hit 40% of them. But the issue raised is whether he cut down because he found better things to do with his possessions. And I think the clear answer is "no". Against a baseline expectation of general improvement with experience, the distinct improvement awaits the 1984-85 and 1985-86 campaigns, where his long-range game returns to the importance of his rookie year and non-coincidentally, his FTA/2FGA improves. (Nothing like the threat of being able to step back for a three to open up the floor and draw fouls.)
But this aside, there is the more interesting issue of Larry as representing, as well as lagging, the league. That his 3 point talents weren't utilized in his early career, to the detriment of the Celtics, is one thing. But this technological retrogression was a league-wide phenomenon. After the novelty (and success) of the first year, 3 point shooting largely disappeared from the league, exceeding the initial year's 3.1% of FGAs only five years later (3.5%). Same with Larry, but worse! He started off with 9.8% of his FGAs, and only six years later did he regain (and surpass) that level (12.1%).
The point is that Larry is illustrative in terms of being the revolutionary who wasn't, but the ostrich-like tendency to look away from the potential of the 3 point shot was a league-wide phenomenon.
Mike G wrote:
Yes, and in the ABA, toward the end of the 9-year experiment, 3-pt shooting had been pretty much declared dead as a viable strategy. About as useful as the 2-point conversion in football.
What can I say? If the point is that this is what was the conventional wisdom, I guess I agree. But it was wrong. Full stop. A fact as plain as day for anyone back then who could multiply by 1.5 and had a basic intuitive sense of the relative worth of the range of observed scoring opportunities. Reality is a bitch sometimes.
John Hollinger wrote:
The thing about the NBA 3-point line is you needed a generation of players to come along who had developed that skill. The original NBA 3-point shot was intended to be rare and beyond the range of what a typical jumper was in that day. Players just practiced the shot, whereas prior to that time it made no sense.
Why you hate Larry Bird?! Don't hate Larry Bird! The story of Chris Ford is also interesting from that same first year of the three point shot. He killed the shot to the tune of shooting 0.427, and then in the last two years of his career that followed, he suffered the consequences of the Dark Ages alongside his far more storied teammate.
But clearly he should have had a sense of the potential value, no? No evidence of that in his stretch as coach of the Cs in the early 90s however, not that he was blessed with a lot of raw material to work with (but did he make it a priority?) And then Larry was coach of the Pacers which had a lot, a lot of long range talent. It is a superficial reading, but it isn't clear to me that he pushed the curve with it, in terms of maximizing attempts.
Conventional wisdom is a tyrant.
John Hollinger wrote:
I played HS basketball when they put in the 3-point rule, and what was interesting was that the players understood its value almost immediately (the college line, which we used, is close enough that good HS shooters were already shooting from roughly that range), but the coaches, having never been in that environment before, took a long time to get it.
Doesn't make much sense, does it?
Back to top
View user's profile Send private message
Mike G
Joined: 14 Jan 2005
Posts: 3458
Location: Hendersonville, NC
PostPosted: Fri Feb 04, 2011 1:06 pm Post subject: Reply with quote
schtevie wrote:
Mike G wrote:
...in the ABA, toward the end of the 9-year experiment, 3-pt shooting had been pretty much declared dead as a viable strategy. About as useful as the 2-point conversion in football.
What can I say? If the point is that this is what was the conventional wisdom, I guess I agree. But it was wrong. Full stop. A fact as plain as day for anyone back then who could multiply by 1.5 and had a basic intuitive sense of the relative worth of the range of observed scoring opportunities. Reality is a bitch sometimes.
The last ABA player to attempt as many as 50 threes in a season and make over 35% was George McGinnis in 1975.
http://bkref.com/tiny/dB0j1
Sort by season.
Of a total 49 ABA player-seasons with >50 3FGA, the median 3fg% was .314 .
Multiplying by 1.5 gives us .457 as an eFG% on threes, when taken by a league's premier shooters.
This was a 9-year experiment undertaken by several hundred players and a few dozens of coaches.
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
schtevie
Joined: 18 Apr 2005
Posts: 392
PostPosted: Fri Feb 04, 2011 1:54 pm Post subject: Reply with quote
Mike, I believe that confusion reigns on multiple levels, and hopefully it is not mine.
First of all, the threshold you consider relevant, is it? Let's take the last year of the ABA, just as the closest bridge year. No cherry-picking. The 2 point fg% that year was 47.3%. So, the effective 3 point equivalent would be 31.5%, right? And that is only if you consider the threshold for three point shooting to be the average. It clearly isn't. The threshold is the opportunity cost, the worst alternative, which as we know (we do, right?) is much lower than that. The same class of shots that are still the worst today, mid to long range jumpers, contested more than not.
But let's take 31.5% as the threshold anyway, setting the bar high to see if we can clear it. And let's take the 50 attempts reference point as well, though I am by no means arguing that the ABA was enlightened in integrating 3 point shooting into its offenses. My only point in invoking the defunct league is that there was a history to ponder than should have been familiar to the NBA.
So, ABA, 3fg% > or = 31.5% and 50 attempts min., basketball reference here we come:
1975-76: 6 players
1974-75: 9 players
1973-74: 11 players
1972-73: 11 players
1971-72: 12 players
1970-71: 12 players
1969-70: 8 players
1968-69: 11 players
1967-68: 8 players
So, you are looking at about one guy per team on average, but many of these players shot a lot higher percentage than 31.5% and had a lot more than 50 attempts. I don't know why your B-R query left you short, but there are the data.
And for comparison, in the NBA's first 3 point year, there were 12 players taking 50 attempts (where there analogous threshold was 32.5%), roughly one player for every two teams, compared to about one per team, on average, in the ABA.
Back to top
View user's profile Send private message
Mike G
Joined: 14 Jan 2005
Posts: 3458
Location: Hendersonville, NC
PostPosted: Sat Feb 05, 2011 9:04 am Post subject: Reply with quote
OUCH! My bad. Player Season Finder set to Forwards only.
Try again -- top 100 ABA 3fga seasons lists those with >118 3fga.
Sorting by 3P%, the median is .320
Sorting by Season, just one player (Billy Shepherd, '75) hit over .354 after 1974.
Just 4 of the 100 'prolific' 3-shooters are in the list from the final ABA season.
Quote:
I am by no means arguing that the ABA was enlightened in integrating 3 point shooting into its offenses.
Louie Dampier (Ky), from '69 onward always managed to hit at least .348 from the arc (.358 career). He had 550 attempts/yr before Issel came, then cut that in half, and when Gilmore arrived dropped further every year.
Even as his 3fg% rose to near .400, his 3 attempts dropped below 10% of his FGA. Coach Hubie Brown apparently didn't think much of it, except perhaps as a distraction.
Of course, not everyone has a Gilmore in the painted area, nor also an Issel in the wings. But everyone gets more trips to the FT line from shots inside the arc. And these shots skew quite a bit the effectiveness of the average 2 vs the avg 3.
If a team's TS% is .540, then shouldn't a 3fg% of 2/3 that rate be the threshold? This would be .360, and it would be a long time before teams stocked players with that ability -- and/or with adequate additional basketball skills.
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
schtevie
Joined: 18 Apr 2005
Posts: 392
PostPosted: Sat Feb 05, 2011 3:38 pm Post subject: Reply with quote
Mike G wrote:
.....But everyone gets more trips to the FT line from shots inside the arc. And these shots skew quite a bit the effectiveness of the average 2 vs the avg 3.
If a team's TS% is .540, then shouldn't a 3fg% of 2/3 that rate be the threshold? This would be .360, and it would be a long time before teams stocked players with that ability -- and/or with adequate additional basketball skills.
Um. No. Definitely not.
This is not how the concept of a threshold works, given the realities of offensive basketball.
On average, ex post, one expects the average productivity/efficiency to be realized. But that has essentially nothing to do with actual choices faced, ex ante, on a given possession, where the threshold concept is operative.
You start a possession (and the shot clock) hoping for luck and good fortune, that a good shot attempt around the basket can be realized, one that yields, in expectation, a way above average shot. This happens relatively infrequently however (otherwise all shots would be above average, and....) and offenses are compelled to take a large fraction of below average shots.
Given that the basic physical dimensions of the basketball court haven't changed over the relevant history, nor has the reality that defense extends out from the basket, contemporary data about the relative returns of shots from different locations can inform our views of the ABA and 1980s NBA past (http://hoopdata.com/teamshotlocs.aspx and http://www.82games.com/locations.htm).
Accordingly, the relevant threshold for a 3 point attempt was and is approximately that of a mid-range to long 2, and these aren't (and likely weren't) fouled all that much (so no large FT bonus should be added). The bottom line being: the relevant 3 point threshold was and is well below the average TS%. And as such, there has always been an "excess supply" of 3 point shooting ability in the league.
THE 3 POINT SHOT HAS EVER AND ALWAYS, AT A MINIMUM, BEEN AN ABOVE AVERAGE, BELOW AVERAGE SHOT IN ITS POTENTIAL.
As this discussion continues, I guess I am obliged to revise my expectations slightly. Maybe coaches back then really couldn't multiply by 1.5 and really didn't have a sense how unproductive jump shooting was. But a lack of the ability to keep score in a game where score is everything seems such an odd assumption.
Back to top
View user's profile Send private message
Mike G
Joined: 14 Jan 2005
Posts: 3458
Location: Hendersonville, NC
PostPosted: Mon Feb 07, 2011 6:15 am Post subject: Reply with quote
The dimensions of the court are the same as ever, but the effective reach of the defense has seemingly expanded.
In the ABA, and in the post-merger NBA, the long 2-pointer was always available. As was the 3-pointer. But it also wasn't hard to get a shot from 16-18 feet. Without the 3 as an incentive, the vast majority of players' shooting ranges drop off dramatically somewhere before 20'.
In the mid-80s, it was rare to see perimeter passing as far as 25' out, and now it's standard; even bigs who are never going to shoot from there are out there, because that's where the offense is happening.
If milling around the 3pt line is taking you out of the offense, you had better be a damn good 3pt shooter. Lots better than 32%. You aren't going to set many picks, get many offensive rebounds, get many assists out there.
At some point -- the Bad Boy Pistons era, perhaps -- the open 16-18 foot shot had become so rare that offenses had 2 choices: drive it inside, or set up at the arc. Thus the 3-pt era was created. Not so much from great insight, as from necessity.
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
schtevie
Joined: 18 Apr 2005
Posts: 392
PostPosted: Mon Feb 07, 2011 3:34 pm Post subject: Reply with quote
I very much look forward to Ed K reprising his chapter reviews, but until then, let me try to better establish the fact that opinions on the shape of the earth should not be allowed to differ when it comes to the history of the adoption of 3 point shooting in the NBA. There can no fundamental ambiguity. Taking a broad brush (and apologies to the unappreciated best practitioners of their day) NBA coaches, players, all relevant decision makers were, collectively, horrible innovators. Many, many points, and dramatic competitive advantage for the innovators who weren't, were foregone.
Invoking changing (improving) defenses over the relevant history doesn't impact this fundamental determination. Arguing that there were supply side failures is a non-starter. The relevant shortcomings were all on the demand side. For ignorance or strategic risk-aversion (aka keeping one's head down) big mistakes were made, persisting over decades.
I pursue this pedantically, not because NBA empirical history should necessarily be of interest to those interested in stats (though one would be a better person accordingly) but because the belief that we have ever and always been in a quasi-optimal state of best practice is false and influences current thinking. It was screamingly false regarding the 3 point shot way back when, but all that has changed to the present day is a question of degree. The 3 point story doesn't end in the 80s. And if this, what else? This should be the informed prior.
OK.
The threshold. It is probably very fair to say that the reasonable opportunity cost of a 3PA throughout NBA history has been about 0.400. Why? As noted, the fundamental reality of defense extending out from the basket hasn't changed since the game began. This is a very useful fact. Roland posted some extreme charting summaries (previously linked) for the 2004-05 season (I believe). These tell a very clear, and ultimately intuitive story. Comparing a straight average of the three categories that comprise his zones labeled 6 through 10, we have a FG% of .426. And we also note that these long (and some not-long 2s) had a straight average "play" efficiency of 0.394 (Roland labels this column "Pts per 100 poss" but I am pretty sure he means per 100 plays, according to the Oliverian nomenclature.) This latter concept is very close to the actual decision threshold. If I am to take a longish 2, I expect a certain augment for being fouled (in the instance, including being merely in the zone, not just attempting a shot...I believe that this is a correct interpretation of the data) as well as a certain decrement for the possibility of turning the ball over and getting no points. So, in 2004-05, 0.400 was a useful opportunity cost.
And this is essentially corroborated by the more contemporary shot location data at hoopdata, that shows that from 10 feet out to the 3 point line the average FG% is about 0.400.
Now, is there any reason to believe that this benchmark has changed over the years? I don't think so, but I would be interested in hearing alternative views. Going back to 1979-80, when the 3 point shot came to be in the NBA, the average offensive efficiency was basically the same. The reality of defenses extending out was the same. The basic rationale that high quality shots are the ones worthy of fouling was surely the same. Two differences that might be noted is that FG% was higher and FT/FGA was lower back in the day, but that is consistent with the story of later defensive tightening on high quality shots (i.e. "no layups" allowed) and suggests nothing about long 2s. These weren't fouled much in 2004-05, and it is highly implausible to believe that they were fouled more (when fouling was generally less) back in the early NBA shot clock era.
So until persuasively argued otherwise, a threshold of 0.400 it is. Bump it up a bit if you choose, no matter.
Next issue: was slow adoption due to demand or supply?
I repeat the NBA's initial year's success rate on 3s: 0.280 or a eFG% of 0.420. "Ah", the skeptic might say, "Right at the threshold. See. Quasi-optimality prevailed." And in a sense, yes. The average use was just at (above) the threshold. Call it a baseline. But it is important to realize that in the initial, half-assed (usage rate compared to the ABA) "implementation", there were inevitably buzzer beating heaves, not consciously integrated as part of an offense. But call it a baseline, and we move on to year two, 1980-81, the beginning of the mini-Dark Age of the 3 point shot.
Attempts dropped from 2.8 to 2 per game, and the success rate dropped to 0.245, making the realized eFG% 0.37, now discernibly below the alleged threshold. But what of the marginal change? Tracking this, you see that what was culled from the previous year's offense were the 3 point shots having an eFG% of 0.566! Technological retrogression indeed.
And this story, of "marginal" yearly changes being well above the 0.400 threshold, indeed often well above overall offensive efficiency, is the crucial fact in understanding the history of the 3 point shot. Despite realized changes being incredibly effective, nothing but the most tentative steps in implementation.
Year "Marginal" eFG% (* denotes shorted 3 point line)
1981, 0.566
1982, 0.571
1983, 3.750
1984, 0.750
1985, 0.581
1986, 0.353
1987, 0.526
1988, 0.886
1989, 0.508
1990, 2.250
1991, 0.267
1992, 0.750
1993, 0.556
1994, 0.448
1995, 0.608*
1996, 0.811*
1997, 0.320*
1998, 0.609
1999, 0.535
2000, 0.559
2001, 0.000
2002, 0.547
2003, 2.100
2004, 0.300
2005, 0.750
2006, 0.917
2007, 0.532
2008, 0.609
2009, 12.000
2010, -27.000
Pretty much no matter the year (the few odd numbers mostly reflecting small usage changes and rules changes) and encompassing what is generally acclaimed to be the era of improved team defense, there was adequate supply of "3 point shooting ability", defined as the combination of raw shooting talent and the ability to find its effective use in the offenses of the day. (And completely unmentioned so far in the argument is that demand could have easily called forth additional supply. There was never any impediment, and all incentives would have been perfectly aligned, for coaches to have instructed their players in the off-season to work intensively on their long-range shooting.)
Eyeballing this data, we are talking 0.3 points per shot transferred from a long 2 to a considered 3 pointer (and likely more in the sense that one first culls the contested long 2s by mediocre shooters from the offense, not the wide open ones by the more skilled). A huge opportunity, always available, and decisively taken advantage of by none.
And still, 30 years on, and its not yet clear that the optimal usage rate of the 3 pointer has been achieved in the league.
Back to top
View user's profile Send private message
Mike G
Joined: 14 Jan 2005
Posts: 3458
Location: Hendersonville, NC
PostPosted: Tue Feb 08, 2011 9:31 am Post subject: Reply with quote
schtevie wrote:
...
I pursue this pedantically... because the belief that we have ever and always been in a quasi-optimal state of best practice is false and influences current thinking. It was screamingly false regarding the 3 point shot way back when, ...
Perhaps nobody really believes in such a thing?
Nevertheless:
In 9 years, the ABA had 92 team-seasons, not counting teams that folded early in some year.
The median 3PA is 350.
The 46 teams shooting more than 350 3's in a year had a median Win% of .470; 9 teams over .600 and 13 teams under .400 .
Those attempting fewer than 350 had a median W% of .540, with 15 over .600 and 9 under .400 .
http://bkref.com/tiny/sFZ5j
In the years 1980-85, there are 137 NBA team-seasons, with a median team 3PA of 187-188.
Of 68 teams shooting 188 or more threes, the median W% is .476
Of 69 trying fewer than 188, the median is .537.
Where in this broad brush view is evidence that shooting more 3's is a good strategy?
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail