Get Now!

Wednesday, December 26, 2012


by Georgevine Moss

Learned some basics about basketball player evaluation methods recently and so ended up playing around with some stats...

First, I made up my own equations, one that evaluates a player based on points (pts evaluation) and one that evaluates the player based on other factors (game evaluation). And, finally, adding these two equations made up the complete evaluation (total evaluation).

Then I picked two teams: the LA LAKERS and the NY KNICKS. Why these two teams? Well, there is a very scientific and logical explanation for that...I saw some pictures of famous people on the court rooting for the KNICKS recently and that reminded me of some other person really liking the LAKERS. Follow the links if you're dying to know who they are. Also, there seems to be this debate, orange too bright?

The following step was to gather some stats for the players of these two teams. I found statistics (box scores) from 2010 and used those to do the evaluation. Now, since it is overall accepted that NBA players' paycheck is mainly based on how good a scorer the player is, I also tracked down some info about players' salaries from 2010 and 2012.

What did I come up with? Read on.

For 2010, the NY KNICKS was a worse team that the LA LAKERS.

Adding the performance of 20 players yielded a score of 2.52 for the KNICKS, while adding the performance of 16 players yielded a score of 2.92 for the LAKERS.

This was confirmed by looking at the 1/1/2010 - 4/14/2010 period results.

•1051 wins and 1826 losses for the NY KNICKS, a WIN/LOSS RATIO of 0.58
•2161 wins and 747 losses for the LA LAKERS, a WIN/LOSS RATIO of 2.89

So all of you who read this and actually follow these teams so far are thinking...duh! But hopefully the following unbiased and based solely on statistics info might offer something you've never thought about before...Here it goes.

For 2010:

•Best player for the LA LAKERS was Trey Johnson with a total evaluation score of 0.50 (he played for only one game, but the equations actually try to take into account both games and minutes played so this number should still be valid. So, based on this one game and the whole 13 minutes played he was the most efficient player. Agree, disagree?)

•Best player for the NY KNICKS was Raymond Felton with a total evaluation score of 0.26 (he played for 54 games and 2074 minutes, which was the 3rd position in terms of time played after Amare Stoudemire who played 2870 min (eval. score of 0.16) and Landry Fields who played 2541 min (eval. score of 0.14) Any objections?

For the NY KNICKS:

•Based on his total evaluation score (0.26), Raymond Felton with a 2012 salary of $3,480,453 is underpaid compared to both Amare Stoudemire (eval. score: 0.16, 2010 salary: $16,486,611, 2012: $19,948,799) and Carmelo Anthony (eval. score: 0.17, 2010 salary: $17,149,243, 2012: $19,444,503)

For the LA LAKERS:

•Based on his total evaluation score (0.18), Kobe Bryant with a 2010 salary of $ 24,806,250 and a 2012 salary of $27,849,149 is overpaid compared to both Pau Gasol (eval. scor: 0.21, 2010 salary: $ 17,823,000,2012: $19,000,000) and Andrew Bynum (eval. scor: 0.21, 2010 salary: $13,700,000).

2012-2013 STRATEGY

•For 2012-2013, based on the salaries list I found the NY KNICKS team has these 3 players who also played in 2010 (and therefore I have stats for): Raymond Felton (0.26), Carmelo Anthony (0.17), Amare Stoudemire (0.16).

In 2010, Raymond Felton played in 54 games for a total of 2074 minutes, while Carmelo Anthony played in 27 games and 977 min and Amare Stoudemire played in 78 games and 2870 minutes. Based on their evaluation scores Carmelo Anthony (points eval: 0.03, game eval: 0.14) should be playing more than Amare Stoodemire (points eval: 0.01, game eval: 0.15) especially for when points are needed. And Raymond Felton (points eval: 0.01, game eval: 0.25) should be playing more, as in all the games.

These 3 players make a combined total evaluation score of 0.59.

•For 2012-2013, based on the salaries list I found the LA LAKERS team has these 4 players who also played in 2010 (and therefore I have stats for): Kobe Bryant (0.18), Pau Gasol (0.21), Steve Blake (0.14), Devin Ebanks (0.15).

Devin Ebanks played in 20 games and 118 minutes total in 2010. Based on my performance equations his points evaluation was 0.03 and his game evaluation 0.12, while Steve Blake played in 79 games and 1581 minutes total and his points evaluation was 0.00 and his game evaluation 0.13. Based on their eval. scores Devin Ebanks should be playing more.

Kobe Bryant and Pau Gasol both played in 82 games with Pau Gasol playing for more minutes. Based on their respective evaluation scores, things should be kept as is.

These 4 players make a combined total evaluation score of 0.68.

Here are 2 tables with all the players' stats:

Do you want to know the equations behind the evaluation scores? Do they appear to be any good? Please, chime in the comments.

Also, if you liked this blog post you might enjoy this:

The Pot Calling the Kettle Black, Are NBA Statistical Models More Irrational than "Irrational" Decision-Makers? by David Lewin and Dan T. Rosenbaum, Sept 2007










SUPER IMPORTANT DISCLAIMER: All this is just for fun. Of course you can't evaluate teams like that.

No comments:

Post a Comment