The only two of the top five positive statistics (points, rebounds, assists, steals and blocks) or two negative statistics (fouls and turnovers) that unquestionably results in a score for either team are points and assists. For the rest, whether it's a steal or turnover or a block or a foul, can only increase or decrease the likelihood of a score, but none of them are sure things.
With that in mind, we can look at the combination of points and assists as a measure of the number of points each player is responsible for. Suppose you score 20 points and hand out 5 assists. Then you are responsible for a total of 30 points: 20 that you scored yourself and the 10 points other teammates scored off of your assists.
Points Responsible = Points + 2 * Assists
There are four problems with this approach. First, an assist for a three pointer counts the same as an assist for a two pointer. You might have actually been responsible for a total of 35 points: 20 points that you scored yourself and the 15 points off of 5 three pointers that your teammates scored off of your passes. So the right multiplier on assists is some number between 2 and 3.
Second, passes that lead to a foul and a missed field goal attempt do not count as an assist at all. Players who tend to give assists under the basket to players who get fouled would therefore not get the credit they deserve by this method.
Third, this method completely ignores field goal percentage and turnovers. Players who score a lot but shoot a low percentage and players who make a lot of passes, only some of which are good, would tend to be rewarded by this methodology.
Fourth, this method penalizes teams who play at a slower pace. Those teams that play an up-tempo, high-scoring game would tend to have higher points and assists all across the board.
How to address these three concerns? First, we will try different multipliers between 2 and 3 to see how it affects the rankings. Second, we will assume that among the caliber of players we are looking at, there is no bias among players who tend to give passes that lead to fouls vs. those that don't; in other words, we will assume it averages out. Third, we will address the shooting percentage and turnovers problem by noting again that those statistics do not necessarily result in a score for the opposition. You can miss all the shots you want and turn the ball over as much as you want but if your team puts up more points than the other guys, you win. Fourth, in a future exclusive, we can consider the result of points responsible as a ratio to the average points scored of the team.
For statistics through March 3, 2005, here are the leaders using a multiplier of two for assists. These are total numbers for the entire season as opposed to averages to reflect the total contribution by the player over the entire year to date.
And here are the top 10 if the multiplier is 3 instead of 2: notice how Steve Nash moves up in the rankings immediately because of his high assists totals.