## Attributing Bucks Wins and Losses using Socialist Principles

Yesterday, or maybe two days ago, I attempted to attribute the Milwaukee Bucks wins and losses this season to individual Bucks players according to their Marginal Win Scores per 48 minutes.  Basically, MWS divides responsibility for wins and losses according to the difference between the beneficial statistical production per minute of the individual player and the beneficial statistical production per minute of his direct opponents.  In some ways, the MWS system can be too individualistic (especially on defense, where there is a definitive “help and recover” system), and I acknowledge that.  But I think its still a reliable and fair and American way to distribute responsibility for team success and team failure.

But what would happen to those win-loss numbers if I threw out Marginal Win Score and its individual offensive and defensive statistics and absolutely went to the other extreme?

In other words, what would happen if I simply considered everything that happened on the court as a group generated activity and therefore assigned 1/5th responsibility for the team’s wins and losses strictly according to the actual results the team achieved when each particular player was on the court?

So I went ahead and did that, and I post the comparative results below.

First, here’s specifically what I did.  I went to basketballvalue.com and looked up the offensive and defensive efficiency numbers that the Bucks have posted so far while each individual player was on the court (number of points divided by number of possessions).  Then I translated those raw efficiency numbers into expected winning percentages (using Professor Berri’s formula: (0.480 + (off eff) * 3.138 – (def eff) * 3.119).  Once I calculated the player’s specific winning percentage, I took his minutes and divided them by 241 to come up with his 1/5th games (because 241 is the average number of total player court minutes per Bucks game played this season).  Finally, I multiplied the number of 1/5th games played by the expected winning percentage.

Here are the comparative results.  The first two columns are the result of a Marginal Win Score analysis, the second two columns are the result of Socialist analysis.

 W% W__L socW% socW__L Bogut 0.789 5.7__1.6 0.479 3.5__3.9 Moute 0.651 3.9__2.1 0.464 2.9__3.3 Ilyasova 0.469 2.7__3.1 0.523 3.0__2.8 Dooling 0.411 2.3__3.3 0.637 3.6__2.1 Delfino 0.571 2.3__1.7 0.381 1.6__2.5 Boykins 0.721 2.1__0.8 0.535 1.6__1.4 Jennings 0.331 1.9__3.9 0.368 2.2__3.7 Maggette 0.349 1.8__3.4 0.414 2.1__3.1 CDR 0.519 1.7__1.6 0.431 1.4__1.9 Brockman 0.631 1.4__0.8 0.646 1.4__0.8 Salmons 0.161 1.2__6.3 0.395 3.0__4.6 Gooden 0.289 0.7__1.7 0.336 0.8__1.7 Sanders -0.061 (-0.2)__2.7 0.197 0.5__2.0 TOTALS 0.4485385 27.5__33.5 0.4466154 27.6__34.4 Actual: 23-38 Actual: 24-38

In a Socialist World, Dooling is the Bucks best performer

As you can see from the chart, the player that benefits the most from such a Socialist analysis is G Keyon Dooling.  In fact, he becomes the Bucks best performer.  The player whose winning percentage would take the largest hit is Andrew Bogut.  He would go from being the Bucks best performer to being just a run-of-the-mill Bucks player.

The other players who benefit substantially under Socialism are G John Salmons and C Larry Sanders, both of whom would take large winning percentage jumps.  The players who would suffer substantial fallbacks along with Bogut are Earl Boykins, Luc Moute and Carlos Delfino.

Interestingly, the rest of the Bucks numbers move up or down by just 0.3 tenths of a win.  Jon Brockman’s numbers are basically identical under both systems, and Drew Gooden’s numbers are nearly identical.

I’m really uncertain what all of that means, but in light of the Slate article cited in one of yesterday’s posts, an article that suggested one cannot distinguish basketball players using statistical production as a measure, I thought it would be interesting to see what would happen if we didn’t.

I suppose you could say that the absolute truth, if there is such a thing, must lie somewhere between the two poles.  But I guess that’s only if you accept that MWS represents the far edge of “individual” attribution, and that’s not neccessarily true.

So there you have it.  I guess I haven’t moved the ball one inch.  But that’s not what I’m here to do.  I’m here to try to provoke and hopefully entertain basketball fans who have a layman’s analytical bent.

I guess I’m not sure I did that either.  Oh well, bottoms up pro hoop fans.  We may be a diminishing minority, but we can always say we are fans of the one major professional sport that has its roots planted unequivocally in the United States.