Yesterday the Milwaukee Bucks fired their head coach, Terry Stotts. Little explanation was given initially for a move that appears on the surface to be pointless. With only 18 games left in the regular season, changing coaches would seem to have no impact on the final outcome this team will achieve in 2006-07. And although this team has probably not played as well as Milwaukee fans would like, should we be blaming the coach for this team’s failures? Or, to put it another way, did this team simply fail because of too many injuries and too little talent (two issues beyond the control of the coach)?
Consider what these players have done so far this season. The summation of Wins Produced for these players stands at 25 after 64 games.
Table One: The Milwaukee Bucks after 64 games
Now let’s ask what would have happened if the veteran players on this team (we will assume rookies are what they are) played the same minutes, but their Wins Produced per 48 minutes [WP48] was the same as it was in 2005-06? If we calculate Wins Produced based on last year’s productivity numbers, we see that Wins Produced for this team now sums to 24.9. In other words, this team is right where we would expect given how good these players have played in the past and how many minutes these players played this season.
Now it’s not the case the productivity for every player didn’t change. Ruben Patterson and Maurice Williams have improved. Charlie Bell has regressed (although that might be – as SI.com suggests – Bell has been spending some time playing out of position at small forward). The net impact of these changes, though, is virtually zero.
From what I understand Stotts didn’t pick these players or cause their injuries. So blaming him for the outcome achieved seems a bit unfair.
Now the team is in the hands of Larry Krystkowiak (who may have been given the job to prevent him from taking the University of Utah position). Is he going to be able to prevent injuries? Can he make these players play substantially different from what we saw in the past?
Although he might be able to alter performance, I am not sure that is likely to happen. My research with Mike Leeds and Mike Mondello suggests that many coaches do not impact player performance.
Now the team will add a lottery pick this year that might help. It might also add or subtract other players. As the roster changes, we can expect the results might also change. Krystkowiak will get credit if this roster improves and the team wins. He will also be fired if that doesn’t happen. I am not clear, though, if he has the power to make either result happen.
A Comment on an Iverson Story
The Bucks story points to a bigger picture, but before I get to that I am going to note a story about Allen Iverson that is related. And let me start by apologizing. Readers in this forum have to be tired of hearing about Iverson. I am certainly getting tired of talking about him (Marty is getting really tired hearing me talk about him).
Still, Ian Thomsen wrote a column yesterday at Sports Illustrated.com that begs for a comment (and it is related to the story I am telling here).
Thomsen has noted that the 76ers have improved since Iverson departed. But rather than draw the obvious conclusion (Iverson may not be as good as people originally believed), he has managed to construct a story that credits Iverson for the 76ers improvement. Yes, the team was only 5-10 with Iverson and 5-19 when he was traded. Since the trade the team has gone 20-21. Iverson, though, according to Thomsen, taught these players “how to play.” Therefore, it is Iverson who is ultimately responsible for this improvement.
I have to admit, that’s an original story. Team plays badly with player then improves when he departs. Still, the departing player is the reason for the improvement.
Let me note that the key players on this team are Andre Iguodala, Samuel Dalembert, and Andre Miller. Coming into this season these three players had a combined 13 years of NBA experience. In 12 of these years, these players were above average players. In other words, all three have an established history of being “good” players.
When we look at the trade we see that these players were above average before the trade. And they were above average after the trade. What can we conclude from this data? I think the simple story is that Iguodala, Dalembert, and Miller are all above average performers.
Thomsen’s argument ignores the simple story. Rather than simply say “the 76ers play better with Miller because Miller is more productive than Iverson”, Thomsen feels compelled to argue that it is the teaching from Iverson that have finally paid off (although why his teachings never paid off for ten plus season is not explained).
And the Big Picture…
It has been noted in this forum that player performance is impacted by injury, his level of experience, the productivity of teammates, the stability of a team’s roster, and coaching. All of these factors can have an impact — that is statistically significant — on player performance. In other words, getting back to Milwaukee, the Bucks could improve as Andrew Bogut gains experience. Or maybe Krystkowiak is one of those coaches who can alter performance. Or maybe if the team just stays together it will improve.
All of this is possible, but counting on this ignores that what a player does in the past is the primary determinant of what that player will do in the future. In other words, for the most part, players are what they are. Changing coaches is not what generally changes results. And working with inspiring teammates does not often change results.
What does change results – and I think the 76ers have found this to be true – is changing the players who wear your uniform. And when the Milwaukee Bucks learn from this lesson, they too will see the team finally make progress in the standings (assuming they have more productive players wear their uniform).
In Other News…
By the way, Jon Burkett at MVN.com, has been commenting on the 76ers all season. Burkett is also a reader of The Wages of Wins. Today he posted a comment of mine (from the comment section in this forum), which is my explanation for why playing Steven Hunter has an apparent positive impact on the 76ers ability to win.
– DJ
Wages of Wins Journal 
Okapi
March 15, 2007
Iverson seems to have served as a good test case of some of the hypotheses implicit in “Wages of Wins.” Very prescient analysis on your part.
The Franchise
March 15, 2007
Stotts was fired in order to retain Krystkowiak, according to SLC media. Apparently, the U. of Utah had offered him the job, and had scheduled a follow-up meeting to discuss it in more depth. Milwaukee, who wanted him to be the coach in the future, saw the need to act immediately.
Katie Gold
March 15, 2007
You have done great work in the past on proving the theory that player salary and points scored are linked. Have you done any analysis on the drivers of coaches’ salaries? That would be very interesting and relevant to the points you are making in this article. If you’d like some help, let me know.
Owen
March 15, 2007
Can anyone tell me how to use my netnewswire rss program to track comments on this site?
Sorry for the diversion
dberri
March 15, 2007
Katie,
First someone has to provide me data on coaches salaries. I have one year of data, but no more. If you have that data, please send it along.
Paulo
March 16, 2007
I think the problem lies in the way the players react to certain coaches. Take for example, the Knicks. It was pretty much the same team — sub Channing Frye’s drop for David Lee’s improvement (which on the surface, seems feasible, correct me if I’m wrong) — and you can see some sort of a difference. See also Avery Johnson and Don Nelson. Yes, the differences aren’t profound in terms of win totals, but the attitude of the players have been notable. Particularly the Mavs.
Yes, Terry Stotts is a victim of injuries. Even Phil Jackson (see current Lakers) can’t work around those. But on the other hand, one can say that he has not been able to motivate his players to work harder. Using Stotts as an example, Charlie Villanueva showed us last year that he was capable of doing under a coach who held his players responsible, emphasized work ethic, AND had experience in the pro level (Sam Mitchell) despite not having the X and O reputation. Terry Stotts had no reputation of being so.
So, what’s my point? It has to be a good blend of coach and players. Larry Brown worked well with the Pistons because of the hard-working players, but not New York because of the prima-donna attitudes (Starbury, Franchise, Curry) and the youth of the team.
Dave, is it possible that you publish some notable coaching change stats? Just the basic run of the mill win/loss deltas would probably suffice (I would love to do so, but I don’t have access to the statistics, if you can point me to one, please do so), with the premise that no significant player change was made (IE. last year’s Knicks, 04-05 Mavs, 04-05 Lakers, Byron Scott/Lawrence Frank Nets). I know that the results will be inconclusive, but the results will definitely be interesting.
don
April 2, 2007
Check Villenuava’s WinsProduced/48 last season. He performed basically the same in Milwaukee under Stotts, as he did in Toronto under Mitchell. The Wages of Wins model holds pretty well for him, not seeing coaching as that significant in player performance.
One caveat, however, is that Villenuava’s WinScores with the Bucks and Stotts were significantly better than his previous year in Toronto. But since his injuries and premature return they have sunk to where there is now no difference betweeen the two years.
I appreciate the objective approach here that helps demysify player performance and nonsense about professional coaches being primarily responsible for team’s winning percentages. Haven’t seen one coach, since first watching Mo Stokes and Jack Tyman, who shot the ball, got the rebound, made the turnover, or got to the line; or who stopped the opponents from so doing. The sports media and owenership are just scapegoating with all the coaching changes. And Stotts’ situation establishes that very clearly.