I left off last time having come to the conclusion that value is a function of the amount of credit one gives to a player for his team’s wins. I operationalize this in the following manner:
player touches = fga+fta+as+tr+to+st+bk
team touches = tfga+tfta+tas+ttr+tto+tst+tbk
touch percent = player touches / team touches
win share = touch percent * team wins
If the NBA officially kept track of possessions, this would be easier, but since they don’t, and since I don’t know where some other people get their possession figures, I am making up my own version, which I call “touches”. Essentially, I add up the number of time I know for sure that a player literally touched the ball, and the number of times I know his team touched the ball.
The more important a player is, the more times his team gives him the ball, and the more skilled he is, the more times he gets the ball for himself. Now, this measure gives equal credit for steals and turnovers, missed shots and made, but I am only trying to measure the extent to which he had the ball while his team had the ball. Take the touch percent, and multiply it by team wins to allocate appropriate credit to each player for the team’s success.
Note that this is similar to finding the best player on the best team, except that it is better: First, after we establish who is the best player on the best team, how do we know who the second best player in the league? Is it the second best player on the best team, or the best player on the second best team, or what? This “win share” measure is more nuanced. We are multiplying two factors together, capturing relative player and team goodness within and across teams.
Thus we can see that a player on a bad team, if he is responsible for a huge degree of that team’s success, might even be better on a player on a good team, if his supporting cast is more helpful. In 2005, Antoine Walker (atl) was probably a better player than Rasual Butler (mia), and more valuable to his team, but in terms of win value, Rasual was actually better. Antoine just couldn’t do enough for the Hawks. Maybe if Michael Jordan, in his prime, was put on that Atlanta team, it would have been a different story. But, think of which team was earning more revenue, the Hawks or the Heat? Based to some extent on their star power, but largely also on their success, the Heat took in $132 million to the Hawks’ $92 million, 8th best and 8th worst in the league, respectively.*
So, now for the moment you’ve all been waiting for. Who should have been the regular season MVP for the 2007 season? Here is a list, sorted by winshares, of the top fifteen most valuable
players in the league:
Note that I have to agree with the MVP voters that Nowitzki deserved the honor. This list is heavy on the pho and dal, but they were the two best teams. Note, however, that LeBron James comes in a respectable third, on a 50-win team — he had to carry the load much more for his team than any of the Suns did for their team. You might be surprised to see Nash listed as the second most valuable Sun, but remember a) he missed several games, b) this is not a subjective rating, and c) Stoudemire is very good. By the way, a good way to read this is, for example, “Tracy McGrady was responsible for a little over ten wins, while Carlos Boozer earned just under nine.” McGrady, this regular season, was just a little more valuable.
* These are the 2006 numbers, but they’re probably similar to the 2005 numbers.
** Sorry about the image, but I can’t find a good way to put a table in my post.