I calculated every 2010-11 NBA teams’ Defensive Win Score average and then ran each team’s average along side its wins to see how DWS and wins correlate, if at all. The correlation co-efficient I produced was -0.8218, suggesting a very strong negative relationship between DWS average and total wins acheived. (obviously, the negative correlation makes sense — if Win Score reflects the box score statistics that correlate with wins, then fewer of those statistics on the defensive end should translate into fewer wins for the opponent and more wins for the team.)
Having established that relationship, I then used data provided by 82games.com, as well as supplementary data from basketball-reference.com, to calculate each team’s Defensive Win Score average at each of the five traditional basketball positions. (“Defensive Win Score” is the team’s opponents collective Win Score average. DWS does not purely measure ”defensive” statistics in the classic sense, but I am calling the metric Defensive Win Score because that makes it easier to conceptualize).
I list each team’s positional averages in the table below (the numbers at the bottom are the overall positional averages for 2010-11. You will notice those averages are almost precisely the same as the averages found by Professor Berri when he examined data from 1977 to 2005, suggesting that his method for assigning positions to players — based upon physical criteria and deductive reasoning – mirrors very closely the actual ”eyeball” observations made by those who watch the games).
To assess the defensive performance of a particular team at a given position, simply compare the Defensive Win Score average achieved by the team’s opponent at the position to the NBA average listed in bold at the bottom of the positional column.
2010-11 NBA Defensive Win Score Averages by Position