Chip Crain – of 3 Shades of Blue – has often had nice things to say about The Wages of Wins. So it was with some surprise when I read the following response from Chip to my recent column on the Memphis Grizzlies:
Most people know I respect Dave Berri’s work over at Wages of Wins but this time it is more difficult than normal. His piece titled The Memphis Mirage is accurate but doesn’t tell the whole story in my opinion and is based on some information that is questionable at best as to it’s relevancy to the NBA. Still everyone is free to make up their own mind.
When I read these words I immediately e-mailed Chip to find out what he meant “by doesn’t tell the whole story in my opinion and is based on some information that is questionable at best as to it’s relevancy to the NBA.”
This e-mail led to a response from Chip, and then after a few more e-mails, Chip and I decided to post our dialogue. Chip is much faster than me, so his additional comments were posted yesterday. What follows is what I offered to the discussion (and you should read Chip’s comments before you read what I have to say).
Let me begin by summarizing Chip’s issues with my post. I think Chip made four observations:
1. The link between attendance and wins – reported in The Wages of Wins — is different for Memphis.
2. The Wages of Wins metrics – like Win Score – has trouble forecasting the performance of rookies.
3. Memphis has more flexibility at the guard position than my post indicated.
4. Our time horizons are different.
Here are my comments on each point:
Attendance and Wins in Memphis
Chip agrees that wins are generally the driving force behind attendance in the NBA, but somehow Memphis is different. This statement led me to look at the attendance data from Memphis (reported at ESPN.com from 2002-03 to the present).
2007-08: 12,770 per game attendance, 22 wins
2006-07: 14,654 per game attendance, 22 wins
2005-06: 15,793 per game attendance, 49 wins
2004-05: 16,862 per game attendance, 45 wins
2003-04: 15,084 per game attendance, 50 wins
2002-03: 14,910 per game attendance: 28 wins
2001-02: 14,415 per game attendance: 23 wins
Let me first note that the Memphis Grizzlies have never drawn particularly well. In recent years the average NBA team has drawn more than 17,000 fans per game. Memphis came close to that figure in 2004-05, but in general falls far short of what a typical NBA team achieves at the gate.
Although Memphis does not draw well, it does appear that attendance in Memphis – like the rest of the NBA – is related to wins. When Memphis has done well, or expected to do well (as The Wages of Wins notes, attendance is driven by both current and past success), attendance in Memphis has gone up. When the team has done poorly, or was expected to do poorly, attendance suffered.
Forecasting Memphis
In my original post I linked to a story that indicated that season ticket sales have increased in Memphis. Given what we have seen about attendance and wins, it appears that some Memphis fans expect the Grizzlies to improve in 2008-09. I argued, though, that there is some reason to doubt that this team is actually going to get better.
The players who produced wins for Memphis this past season – Pau Gasol and Mike Miller – are now playing elsewhere. The NBA veterans Memphis is employing next year were – with the exception of Kyle Lowry (who was about average) – below average last year.
And last year’s performance was not an anomaly.
Table One: The Career Performances of the Veteran Memphis Grizzlies
Table One reports the year-by-year career performance of each veteran NBA player currently on the Grizzlies roster. An average WP48 [Wins Produced per 48 minutes] is 0.100. Of the 41 seasons listed in Table One, only 10 were above average. And with the exception of Kyle Lowry, only one above average season – Greg Buckner in 2005-06 – has occurred since 2004-05. In sum, this is not a great group of veterans.
Of the veterans on this team, Chip mentioned specifically the play of Antoine Walker. Walker has been an NBA All-Star in the past. Plus he is only 31 years old, so one might argue that he has something left.
One should note that Walker will actually be 32 when the season starts. And when we look at Table One, we see that Walker is actually best described as an “unproductive star.” He is generally thought of as a star because he can score. But he is not an efficient scorer. And he also has problems with respect to hitting the boards. Consequently Walker has only posted an above average WP48 once in his career, and has generally been well below average. Hence, it seems likely that he will play poorly in 2008-09.
My sense is that Chip would basically agree that the veterans on this team are as I describe. In other words, he agrees that past performance in the NBA is a fair indicator of future productivity. He is not as sure, though, that we can forecast an NBA’s player future performance from his productivity in college.
Chip is correct that the link between college and pro performance is not as strong as the link between current and past NBA numbers. Nevertheless, there is a link. In general, players who perform relatively poorly in college – as O.J. Mayo did this past season – do not become productive NBA players. Of course this is just a tendency. It’s possible that Mayo will become the productive player people currently envision. His college performance, though, suggests otherwise.
Overpopulation at Point Guard
In my post I also noted the issue of where Mayo would play. Currently Memphis has five players listed as point guards (Mayo, Lowry, Mike Conley, Javaris Crittenton, and Marko Jaric). Chip argued that Crittenton and Mayo can shift to shooting guard, hence clearing up some of this logjam. I have no problem with that argument.
Long term vs. Short term
And I think, Chip really has no problem with my assessment of the Grizzlies for the 2008-09 season. Neither Chip nor I think Memphis is going to be very competitive next season. Chip, though, argues that Memphis is on the right path. Although the team will suffer this year (see his reference to Losing to Win in 2007), in the long-run the Grizzlies are going to be very good.
Chip bases this forecast on the observation that young players do improve. Certainly this is true. But I would add that often un-productive young players develop into un-productive old players. In other words, although young players can become more productive, the effect is not so strong that all truly unproductive young players will transform into above average performers (in fact, that is not mathematically possible).
My sense is that Chip understands this point. But he is also encouraged by the fact that Memphis has gone from being a team that was clearly not trying very hard to win this past season to a franchise that is making moves with the intention of winning in the future. Certainly I agree that this is the intention. I am just not sure, though, that the data support the contention that what Memphis is trying to do will ultimately be successful.
Such an argument probably is not welcomed by Memphis fans. As the 3ShadesofBlue motto states:
As a rule we disbelieve all the facts and theories for which we have no use. –William James
I can’t imagine that a Memphis fan has any use for analysis that suggests this team is not on the right track. Unfortunately, I think that is the direction where the data points. Perhaps after more losing and more high draft picks, Memphis will assemble a productive team. Right now, though, I don’t think this has happened. Nor do we have much evidence that it is going to happen sometime soon.
So that is my contribution to this discussion. All in all, Chip and I had a very nice — and cordial — discussion about his team. Hopefully — for the sake of Chip and his fellow Memphis fans — I am wrong (but I still think I might be right).
– DJ
The WoW Journal Comments Policy
Our research on the NBA was summarized HERE.
The Technical Notes at wagesofwins.com provides substantially more information on the published research behind Wins Produced and Win Score
Wins Produced, Win Score, and PAWSmin are also discussed in the following posts:
Simple Models of Player Performance
What Wins Produced Says and What It Does Not Say
Introducing PAWSmin — and a Defense of Box Score Statistics
Finally, A Guide to Evaluating Models contains useful hints on how to interpret and evaluate statistical models.
Wages of Wins Journal 
Brandon
July 10, 2008
Looks like you may have lost this argument. Good post nonetheless.
ilikeflowers
July 10, 2008
Hopefully Memphis won’t acquire any more vets to take the place of the youngsters over the next two years, then this can be a case study that’s settled in 2010.
Jake
July 10, 2008
Two huge trades in the ECF – Brand to the 76ers and JO to the Raps – no comments Mr. Berri? I think Philly is a legit contender as long as they add a shooter (ben gordan? eddie house? brent barry?)
Chris Herrington
July 10, 2008
A couple of points:
1. I don’t distrust your research that shows winning driving attendance more than star players (though I saw a different study that more precisely — and believably — suggests that winning IN THE PLAYOFFS is the real key). But, by pitting those two factors against each other, you miss the real thing that drives attendance spikes: Hope. Expectation (of winning, of excitement). When Portland has a huge ticket spike after winning the Oden lottery, the ticket sales are not connected to winning, but to the expectation of winning (and excitement). When attendance drops over time despite the presence of a star, it’s not just a reaction to losing (a present and past tense element) but a reaction to a loss of hope for future contention.
2. I’m perplexed by the bit here where you say the Grizzlies have “five players listed at point guard.” First, I’m not sure what “list” you’re looking at, but more importantly — who cares? Basketball is a fluid game, with few players locked into one “position.” I don’t think Javaris Crittenton (a 6’5″ slasher) played a minute of point guard for the Grizzlies post trade; Jaric has always played all across the backcourt and wings; and Mayo played mostly at the 2 for USC last year. In basketball, it doesn’t matter in the least what “positions” players are “listed” as. What matters is how the skills and physical attributes of individual players can mesh in a team concept. The numbers can help you analyze this, but not divorced from watching players on the court.
I’m a big believer in advanced statistical analysis for sports, but not separated from a grounding that comes from watching the games. I haven’t read the book, but from my cursory readings of this site, I’ve never gotten a strong feel for the game as played on the court. The best writers who use advanced stats (James, Hollinger, the entire Baseball Prospectus crew, etc) also reflect a palpable understanding and appreciation for the games.
3. No-one seriously thinks Antoine Walker is going to be a major factor for the Grizzlies this season.