Bucks Diary

Sunday, December 07, 2008

The Skiles Effect... proven through number crunching


By my calculations, the Bucks defense last season produced a measly 23.4 half wins, and the offense produced a slightly better 26.8 half wins, for a grand total of 25.1 wins produced (the team actually finished with 26 wins).

Thus far this season, the offense has (incredibly) gotten worse, while the defense has improved immensely.

The NBA average Team Win Score per 48 minutes currently stands at 41.2. The Milwaukee Bucks average Team Win Score per 48 minutes is 33.6. Per player that averages to -1.52. If one does not consider defense or "Opponent Win Score", the Bucks ought to have 2.6 wins and 18.4 losses, and ought to be headed toward a season record of 20.1 wins and 61.9 losses. Obviously, that is not the case.

As I have argued in the past, you must consider the other half of the court, the defensive end, to get a complete picture. By my reckoning, defense ought to count for "half" of the team's overall win total (hence the "half win" construct). And it seems to work out.

The Milwaukee Bucks average Opponent Win Score per 48 stands at 38.9, which works out to +0.46 per court player. This represents an immense improvement... of almost shocking proportions... over last season and the season's before that. And its the sole reason the Bucks are not completely underwater already this season.

Last season the Bucks defense created, as I stated earlier, 23.4 half wins, which obviously works out to only 11.7 wins over the course of the entire season. By contrast, the Bucks defense this season can already be credited with more than half that total... 6.1 wins (the offense has produced, as I wrote above, 2.6 wins, for a total of 8.7 wins produced).

The defense is on course to produce roundabout 47.8 half wins, or 23.9 wins... more than double the defensive wins produced last season. Stunning. But it was what I rather optimistically predicted would happen... based upon historical precedent... when the Bucks wisely hired Coach Skiles.

Unfortunately, the offense is obviously worse than last season, which has thus far dragged down the team's win trajectory. The team's current win trajectory stands at 33.9 wins +/- 2.5 wins. Thus I can say with 95% certainty the team will win between 31.4 and 36.4 games if its Team and Opponent Win Score averages remain where they are.

By contrast, if the team were merely operating at last season's poor offensive efficiency, the team's win trajectory would be headed toward 37.3 wins, +/- 2.8 wins. That means they would be headed with 95% certainty to a record of somewhere between 34.5 and 40.1 wins. In that case, my prediction of +11 wins, made in the above linked June 27 post, would have been spot on.

I guess the point is, the Skiles Effect is stunningly real... but we need the offense to get clicking if we are to get to the playoffs.

2 Comments:

At December 7, 2008 at 12:48 PM, Anonymous Anonymous said...

For some reason I see us with a lottery pick this season. Lets say we were to get #10 or so, do you think Stephen Curry would be a good selection? I think we need a star player who can really sharpshoot and he has that written all over him.

 
At December 8, 2008 at 10:20 AM, Anonymous Anonymous said...

"The NBA average Team Win Score per 48 minutes currently stands at 41.2. The Milwaukee Bucks average Team Win Score per 48 minutes is 33.6. Per player that averages to -1.52. If one does not consider defense or "Opponent Win Score", the Bucks ought to have 2.6 wins and 18.4 losses, and ought to be headed toward a season record of 20.1 wins and 61.9 losses. Obviously, that is not the case."

Have you talked to Berri about this? I'm curious what his response would be.
Is my recollection correct that the "Opponent Win Score" metric is your own idea?

Great work, I really enjoy reading your blog.

 

Post a Comment

<< Home