Thread Tools

Display Modes

**Showtyme** · 08-27-2006, 11:31 AM

By the guy that wrote Tipping Point. His original book review on the book "Game Theory":

http://www.newyorker.com/critics/boo...529crbo_books1

His own blog about it:

http://gladwell.typepad.com/gladwell...ball_by_t.html

Some interesting notes from the review and his blog:

Quote:

In “The Wages of Wins” (Stanford; $29.95), the economists David J. Berri, Martin B. Schmidt, and Stacey L. Brook set out to solve the Iverson problem. Weighing the relative value of fouls, rebounds, shots taken, turnovers, and the like, they’ve created an algorithm that, they argue, comes closer than any previous statistical measure to capturing the true value of a basketball player. The algorithm yields what they call a Win Score, because it expresses a player’s worth as the number of wins that his contributions bring to his team.

Quote:

Berri, Schmidt, and Brook argue that the arbitrary algorithms of basketball experts elevate the number of points a player scores above all other considerations. ... Every other factor, like turnovers, steals, assists, blocked shots, and personal fouls—factors that can have a significant influence on the outcome of a game—seemed to bear no statistical relationship to judgments of merit at all.

Quote:

Here’s what I think the real value of the Wages of Wins system is, though. It gives us a tool to see those instances where our intuitive ratings of players may be particularly inaccurate. In my New Yorker piece, I focused on how the algorithm tells us that Allen Iverson isn’t nearly the player we think he is. But here’s a more interesting finding. The best player, by this measure, hands down, over the past five years has been Kevin Garnett. No one else comes close. I had the authors update their numbers for this season, and Garnett is again number one (with Jason Kidd second, Shawn Marion third and LeBron James fourth).

I don't agree with a lot of it, but it's definitely going to shed all new light on things.

BG7 · 08-27-2006, 11:44 AM

Ben Wallace was 5th in Wins Produced fwiw.

Diophantos · 08-27-2006, 12:16 PM

I haven't read this book, but I've read a few reviews on it and taken a look at the authors' blog.

They have an approach that is a little interesting, but I'm not sure I buy their overall methods. Their basic statement, as regards basketball, is that most people overrate scoring and underrate aspects like rebounding. Unfortunately, I think they take this a little too far, putting down players like Allen Iverson while saying things like "Jerome Williams was one of the strongest players of his generation," or something like that.

I think their fundamental mistake, according to some of the reviews I read, is that they heavily penaize players who score high volumes at relatively low efficiencies: guys like Iverson. After all, shouldn't they give up those shots to their higher efficiency teammates?

While this is true, to a point, it is not completely true. Low-volume, high-efficiency scorers are often incapable of creating their own shots with any degree of regularity, and rely on superstar teammates like Iverson to create opportunities and eat up the "difficult" posessions, allowing them to capitalize on the easy opportunities the defense provides.

This is why they may overrate players like Jerome Williams and Ben Wallace, and underrate guys like Iverson.

Like I said though, I haven't read the book, so if anyone who has thinks I'm off base, they should tell me. The authors do have some nice ideas, I just think their methodology may be slightly off. It's interesting to talk about, at any rate.

**Showtyme** · 08-27-2006, 10:54 PM

Quote:

Originally Posted by Diophantos

I haven't read this book, but I've read a few reviews on it and taken a look at the authors' blog.

They have an approach that is a little interesting, but I'm not sure I buy their overall methods. Their basic statement, as regards basketball, is that most people overrate scoring and underrate aspects like rebounding. Unfortunately, I think they take this a little too far, putting down players like Allen Iverson while saying things like "Jerome Williams was one of the strongest players of his generation," or something like that.

I think their fundamental mistake, according to some of the reviews I read, is that they heavily penaize players who score high volumes at relatively low efficiencies: guys like Iverson. After all, shouldn't they give up those shots to their higher efficiency teammates?

While this is true, to a point, it is not completely true. Low-volume, high-efficiency scorers are often incapable of creating their own shots with any degree of regularity, and rely on superstar teammates like Iverson to create opportunities and eat up the "difficult" posessions, allowing them to capitalize on the easy opportunities the defense provides.

This is why they may overrate players like Jerome Williams and Ben Wallace, and underrate guys like Iverson.

Like I said though, I haven't read the book, so if anyone who has thinks I'm off base, they should tell me. The authors do have some nice ideas, I just think their methodology may be slightly off. It's interesting to talk about, at any rate.

I think that you're right about that, but I think it's interesting the difference in the EFF ratings for certain players to the Win Scores. The author addresses the fact that statistics are too discrete to really be indicative of anything without the total picture, but then the Win Score does exactly that: it tries to ignore the Gestalt effect of the team and instead focuses on "how many wins" a player contributes to a team. I'd be really interested in their methodology.

Basically ANY statistic about a player singularly is going to be ignoring the Gestalt effect. And the truth is, there's no real way to measure that except the way 82games does, by 5-man units and best tandems and so forth. It's an intelligent way to look at stats.

Because what GMs should be MOST interested in knowing is more than just, "can this player perform at this level with these skills and bring such and such to this team", but rather, "will the team around him provide an environment where the total team is better as a whole, and if not, what other players should i acquire to make it work". Contrary to NBA marketing, it's ROSTERS, and not individual superstars, that are the concern of GMs and winning franchises.

**Premier** · 08-28-2006, 10:17 AM

I have not yet read the book [nor do I plan to]. The book has not received acclaim from the respect members of the statistical community. In reviews, the fundamental problems with the statistic's formula are discussed and these flaws are pretty significant. I'll address the problems in the order in which they appeared in Kevin Pelton's excellent review on 82games.com, which is a much better read than the following post.

In Pelton's review, he mentioned that scoring rate was not considered in the statistic and the authors used total points scored, which is not an accurate reflection of scoring ability. As Pelton said, "So by using total points instead of scoring rate, the authors of Wages unintentionally introduce information about a player's passing, rebounding and defense." Outside variables, seperate from the scoring variable that was being isolated, pollute any conclusions derived from any minute-based scoring statistic.

As for Wins Produced, the authors use team defense in the formula, a statistic that should not have any value in an individual linear models. Tony Parker's individual rating is strengthened because he was fortuante to play on a good defensive team, even though Tim Duncan and Bruce Bowen, not Tony Parker, were far more important defensive players. Also, the authors ignore individual aspects of basketball, primarily creating shots, which explains low ratings for players that control the ball more often, which results in an higher amount of turnovers [and subsequently, poor efficiency ratings, also due to lesser shooting percentages from situations in which they are forced to take a difficult shot, being the primary ball handler]. Apparently, no consideration is given to usage rate - how much possessions one uses [per 40 minutes] - so players like Kyle Korver [a crude reference, as I don't know his overall win produced rating] who are generally efficient, being stagnant shooters, will rate higher than players that are less efficient, but one may argue more important, as they the main ball carriers [Allen Iverson, for example].

I'll look more into it, but honestly, this book does not interest me at all.

I'm going to move this back to General [after moving it to Statistical Analysis] temporarily for more traffic.

Busta · 08-28-2006, 11:58 AM

these stat things are so stupid its ridiculous. Especially the +/-, that stat just doesn't work for basketball maybe for Hockey.

It doesn't really look at minutes played, quality of teammates, the fact that they're playing the second unit. And other things such as the fact that the team wont win without some of those players that dont have great +/- stats

The Mad Viking · 08-28-2006, 12:04 PM

These guys can't carry Rosenbaum's laptop.

82games is light years ahead of them.

Before someone starts doing "groundbreaking" research, they would do well to look around and see who else has already completed the excavation.

TDAO · 08-30-2006, 10:06 PM

Quote:

Originally Posted by Premier

In Pelton's review, he mentioned that scoring rate was not considered in the statistic and the authors used total points scored, which is not an accurate reflection of scoring ability.

I know this thread is a couple of days old, but I want to clarify this point. This issue isn't with the Wins Produced statistic, but with the study the authors did correlating scoring and salary. Using points instead of points per minutes made this relationship appear stronger than it really is.

4BiddenKnight · 08-30-2006, 11:22 PM

Quote:

Originally Posted by Busta

these stat things are so stupid its ridiculous. Especially the +/-, that stat just doesn't work for basketball maybe for Hockey.

It doesn't really look at minutes played, quality of teammates, the fact that they're playing the second unit. And other things such as the fact that the team wont win without some of those players that dont have great +/- stats

Tell me, how the hell does the +/- stats work in the NHL? I'm still clueless about it.

Thread Tools Display Modes

Thread Tools

Display Modes