Real Ray
25 Oct 2005, 01:04 PM
http://www.82games.com/comm54.htm
There's an article in the current print edition of Wired about a guy named Dean Oliver. I found the link for a longer piece on his views.
He's now a consultant with the Sonics. In a nutshell (from the Wired piece):
"Oliver thinks possession efficiency is basketball's version of on-base percentage. 'Teams that score a lot of points don't necessarily win games and teams that prevent opponents from scoring a lot don't necessarily win either. But if you convert a greater percentage of possesions into points than your opponent does, you win games.' By tracking a team's per-possession efficency whenever a given player is on the court, Oliver thinks he has the truest measure of a players worth."
Or has he notes in the link piece:
you’re into the technical details (if you’re not, just skip this paragraph), the method I’ve chosen to evaluate teams for their degree of balance is one that looks at the individuals playing at least 20 minutes per team game (i.e., 1640 in the season) and their distribution of possessions. For each of those players meeting the time requirement, I calculate a standard deviation of their percentage of possessions used, with that standard deviation centered around an average value of 20%.
That Dallas team, for instance, had an "Individuality Measure" of 0.020. It implies that the top guys deviated from 20% of team possessions by only about 2% per man, quite small. The Philadelphia Iversons had an individuality measure of 0.076 and the Minnesota "team" had an individuality measure of 0.082, so each team’s players deviated from 20% possession use by about 8%. These numbers seem to capture the subjective perception of individuality pretty well and have an objective side to them that is logical, allowing the analysis of the NBA through time.
The sabermetric thread is so big now that I was lazy and didn't look to see if this has already been posted-apologies if it has.
There's an article in the current print edition of Wired about a guy named Dean Oliver. I found the link for a longer piece on his views.
He's now a consultant with the Sonics. In a nutshell (from the Wired piece):
"Oliver thinks possession efficiency is basketball's version of on-base percentage. 'Teams that score a lot of points don't necessarily win games and teams that prevent opponents from scoring a lot don't necessarily win either. But if you convert a greater percentage of possesions into points than your opponent does, you win games.' By tracking a team's per-possession efficency whenever a given player is on the court, Oliver thinks he has the truest measure of a players worth."
Or has he notes in the link piece:
you’re into the technical details (if you’re not, just skip this paragraph), the method I’ve chosen to evaluate teams for their degree of balance is one that looks at the individuals playing at least 20 minutes per team game (i.e., 1640 in the season) and their distribution of possessions. For each of those players meeting the time requirement, I calculate a standard deviation of their percentage of possessions used, with that standard deviation centered around an average value of 20%.
That Dallas team, for instance, had an "Individuality Measure" of 0.020. It implies that the top guys deviated from 20% of team possessions by only about 2% per man, quite small. The Philadelphia Iversons had an individuality measure of 0.076 and the Minnesota "team" had an individuality measure of 0.082, so each team’s players deviated from 20% possession use by about 8%. These numbers seem to capture the subjective perception of individuality pretty well and have an objective side to them that is logical, allowing the analysis of the NBA through time.
The sabermetric thread is so big now that I was lazy and didn't look to see if this has already been posted-apologies if it has.