Andres ‘Dre’ Alvarez — of NerdNumbers — offered a post this week that generated some attention. His discussion of the merits of Derrick Rose led a number of people to suggest that he start watching basketball. In response to this argument, Dre offered the following at NerdNumbers (which is re-posted in its entirety below):
Just a note…what you are about to read is from Dre (in case that wasn’t clear). And if you don’t have a sense of humor…. well, this may not be for you.
DJ over at the Wages of Wins Journal was kind enough to reprint my article. I love it when this happens, as his blog has a larger audience then mine. However, some of his commenters are not as nice as mine. The first response to my Derrick Rose piece was
“Statistics are useless. Watch a game. Then you understand why he should undoubtedly be in the All-Star game in a starting position.”
It turns out this argument is nothing new. “Just watch” is a silly argument when on any given night there are one to fifteen games on and over 500 hours of game time by the time the All-Star game comes around.
Let’s say hypothetically that I found the time to watch the best performance though of each player in the league. What would the All-Star Starters look like then? Well I was pleased with the answer and thought it was worth a post. So with that here are your 2011 All-Stars, providing of course you had the foresight to actually watch them play on their best night and of course don’t do something silly like look at their aggregate stats over the season.
Table 1: East Starting All-Stars based on their best game.
On the forward front the East looks right. Al Horford was a surprise to play in front of Dwight Howard, given Dwight’s rank on the best games lists. However the guard situation becomes much more amusing. The Derrick Rose, Rajon Rondo and Wade decision is gone, with John Wall and Stephen Jackson having the best games so far this season. I can’t imagine why fans of three such successful franchises as Boston, Chicago and Miami aren’t watching teams more than 10 games back in their spare time.
Table 2: West Starting All-Stars based on their best game.
I’m happy with the West because a Nugget I liked “legitimately” makes the cut. However, the West actually doesn’t look as good compared to the East using this model, which ironically mimics the real world. Love is an excellent choice, but the rest of the starters have not consistently been the best players out West this season. Part of the issue with center is of course that many of the excellent centers out West (including Love) are listed at forward. Still, if the goal was exciting popular players, then the West hits at least two times.
I think many people vote for their favorite players (that is of course the point of the All-Star game) but feel offended when they find their favorite player isn’t everybody’s favorite. That said, no amount of clutch play, excitement factors, leadership and other such intangibles will do much to sway me from disagreeing with players like Kobe, Melo and to a degree Derrick Rose from starting over other players (or playing at all in Melo’s case). The other option of course is what I have done here, where I looked only at the stats I wanted to see to make my player look good. Most players have played 30-50 games at this point and using only one game to decide if they are All-Star worthy is silly. I’m willing to bet that most people usually only watch their home team play on a regular basis. Of course judging the entire NBA on a small set of games watched is probably silly as well, but apparently it is something people do.
*Afflalo would have had to have been a write in vote, which is allowed.
This article uses the Wins Produced and Wins Produced per 48 minutes (WP48) metrics. These use the player’s box score statistics, the team statistics, and league averages to estimate how the player contributes to winning. An average player has a WP48 of 0.100. For a regular starter this would generate around 6.0 wins for the team in a full season of play. By contrast a “superstar” player has a WP48 of 0.250 and in the same minutes would generate around 15.0 wins for the team.