A look at efficiency and point differential

A very interesting take from Citizen BelgianClipper (he has dual citizenship in Belgium and Clips Nation). I couldn't agree more that the unbalanced schedule makes it dubious to get too upset about relative positioning in something like point differential. At the end of the season, when all Western Conference teams have played more or less the same schedule, point differential and Pythagorean Wins will correlate with overall team quality. Even in the second half of the season, it will start to get better. But a third of the way in? There's way too much noise in the data. The Sixers versus the Clippers is my favorite example here. The Sixers are tied for the league lead with eight victories by 20 points or more. But those eight games came against teams the Clippers have played a total of three times -- winning by 20 twice themselves. The fact that Philly has played a lot of really bad teams and trounced them doesn't prove anything. Anyway, have a look at BC's approach for trying to adjust for some of these schedule differences. after the jump. Steve

My dear friends,

A healthy statistical discussion about the TS% of Randy Foye with Erik O inspired me to some extent to dig into the statistical nature of the game. Now the direct inspiration came from our friends at Clipperblog (http://clipperblog.com/2012/02/12/defense-and-the-clippers/ ). Some of you might know that I’m not exactly blown away by the writing and the attitude of that website but this article annoys me for a whole different reason. This article does something what I have seen all too often lately: take some random statistics and take them out of context to make a point. Luckily commenter Tom F. pointed out some much needed background information:

The Clippers are the fourth most efficient offensive team in the NBA. They have played 14 out of the other 15 most efficient teams, including Portland, Utah, Denver and the Lakers twice. They didn’t play a team outside of the top 15 until January 22nd against the Toronto Raptors, meaning they played 13 games before playing a team considered a less than average offensive team.

This got me thinking:

Points per possession, the holy grail of efficiency doesn’t exist in a vacuum. You’ll fewer less points per possession against the Chicago Bulls than against our friends the Bobcats. This is also true for point differentials. I mean, it is a lot easier to have a nice point differential when play teams like the Wizard and Bobcats several times a year. So in this piece I try to address this to a degree.

Let’s try to gauge what our opponents should score against us and what they have scored. How did I do this? Well follow along on the tour.

1) The stats come from http://www.basketball-reference.com/leagues/NBA_2012.html (ESPN (Hollinger) and basketball reference differ quite a bit on points per possession: ESPN gives 103.3 B-R 105.6 but B-R is more easily converted to excel and access).

2) If you take the average the points per possession (both ESPN and B-R us a per 100 possessions stat) of our opponents you get 103.56 ORtg and 102.61 DRtg. League averages are 102.8 so this confirms that the Clippers have played an above average strength schedule.

3) Then I made a rough estimate of the average possessions of the Clippers and their opponents by dividing average point per game for and against by the respective ORtg and DRtg. Clips have 90.86 possessions per game and their opponents 90,91.

4) Now I compute a new stat which I call "expected average points per game" (EAPPG anyone?). This is done by multiplying the average possessions per game with the relevant average points per possession. As such our opponents would be expected to score 94.15 points and allow 93.23 for a point differential of -0.92 if the Clips were the perfect average team from an efficiency standpoint.

This leads to this table

The conclusion is not that shocking. The Clips are very good on offence and less than stellar on defence. I’m quite s(t)atisfied that I now have an Adjusted Point Differential that takes in account the efficiency of the opponent (3.60 for our Clips). Perhaps I will do the same exercises for the other teams in the league in the future but that would take some work. Like I say, it is a very crude approach.

edit:

But here the approach for all NBA teams

How to read this

the 76ers have a actual point difference of 9.11 but the average team would be expected to have a point diff of 1.41 against their opponents. Thus their adjusted point difference is 7.70. The last 2 stats are the difference between their actual points and points expected for and against. This has the nice effect of dividing the Adjusted Point differential between offence and defence. Thus the Sixers have an Adjusted Point Diff of 7.70 which is composed of 2.11 points of offence and 5.59 on defence.