Wins Produced and the Utah Jazz

I have become very fond of the Wins Produced stat put forth by David Berri and the other economists writing at the Wages of Wins Journal.

I love it because I think players who score a lot of points tend to get overvalued. This should come as a surprise to nobody who has read my posts over the past three years. From Carlos Boozer to Big Al, I have long advocated looking beyond who happens to get the most shots (and thus score the most points) and examine other aspects of the game. Sometimes it's a sheer efficiency issue ... the team is stupidly giving an inefficent scorer all its shots (call this the Allen Iverson effect). Sometimes it's more complex ... Boozer scored quite efficiently. However he was (a) a terrible defender and (b) surrounded by teammates who could score at a similarly efficient rate. Thus he could have been replaced by a good defender and his shots spread to other players without any loss in team offense, i.e. Boozer wasn't really adding a lot to the team, despite what his 20 points per game seems to indicate.

Anyway, I think that sometimes a player's points per game stat overpowers everything ... occasionally to the detriment of the team.

While PER and Win Shares value those who score a lot of points, Wins Produced values those who (a) score effectively and (b) dole out more responsibility for wins to other aspects of the game.

So I like it better. I think it's the best all-in-one snapshot of a player's overall effectiveness that we have right now.

But there is a big problem with it: it's not on Basketball-reference.com, or any other data source that is updated daily. Instead the scores are posted only after each season (the 2012 results haven't even been posted yet).

So, I have created a magic spreadsheet that will figure out the stat for us this coming year. Every ten games I'm going to put the data for our Jazzmen, and we'll see what it churns out.

There will be a couple flaws. First, Wins Produced includes a metric for the value of the team's overall team defense. This is includes evaluating the stats for every team. I don't have time to do this. So, instead, I'm going to look at the Jazz's current ranking in defensive effectiveness and assume they deserve the same adjustment as those listed for the 2007-08 teams.

Second, Wins Produced includes an adjustment according to a player's position. A player who plays minutes at multiple positions would have his Wins Produced score adjusted according to the percentage of playing time he plays at the different positions. I am not going to do this. I'm just going to label a player at the position he plays most.

2011-12 Wins Produced

Now, since I know you're curious, here are the Wins Produced by our team last year. They are ranked according to Wins Produced per 48 minutes. For details about how it is calculated, you can see it here.

Here's what we're looking at.

Wins Produced per 48 minutes. A perfectly average player will have 0.099 WP/48.

Wins Produced. An estimate of how many wins that player helped the team gain.

Wins Produced by average player. How many Wins Produced an average player would have gained in the same amount of playing time.

% Above Average. How much more (or less) the player produced than an average player at his position.

Projected Wins Produced. How many Wins Produced the player's current level of achievement projects to over an 82-game season.

I included totals for several of the columns. We can see that Wins Produced predicts the Jazz to win 35 games last year. They actually won 36. Pythagorean W-L projected 34 wins. So adding up the Wins Produced gives us a total within reasonable error of the team's actual wins.

It also projects that had the season been 82 games, they would have won 43 games. That's just an interesting tidbit for looking at last year's team. But it will hopefully be more relevant this year. I think it will be really interesting to see the projections after 10, 20, or 30 games and see how well it ends up predicting what the team actually wins.

Looking at what the Wins Produced stats suggest about our team last year:

• Jeremy Evans' was insane in his limited playing time. But you already knew that. Nobody knows what he'll produce given regular, meaningful playing time. But again the suggestion here is that even if he has a HUGE drop, he's still a very decent player.
• Of the five guys who played the most minutes, Millsap was the best, followed by Favors and Hayward. Big Al ranks 4th. This shouldn't be a surprise when remember that the metric values scoring efficiency over scoring totals.
• Devin was performed slightly below average. This surprised me. While Devin is certainly NOT my favorite PG to ever suit up for the Jazz, I thought the team undervalued what he contributed last year.
• Aside from Hayward (who was really good), our wing situation was awful last year. The Jazz also did a terrible job giving out playing time to these wing players. Josh and CJ got the most minutes, yet they were simply horrific. Again, although Burks wasn't great (his performance was significantly below average last year), given the other options at wing it was insane to keep him on the bench.

Anyway ... whether you're a devotee of Wins Produced like me, or whether you think it's just one metric of many, or whether you hate it completely, it's now part of the Dunk. I'll be updating my magic spreadsheet all season long, and we'll see how it goes.