How Much is Carmelo Anthony Worth?

In my first post of the NBA Betting Market Team Rankings, I threw out a prediction of the line and over/under for Saturday's Knicks/Thunder matchup.

I picked the game at random as the intent was to just show how the rankings work. However, due to Carmelo Anthony's Thursday night ankle sprain which has ruled him out for tonight's game, I can also illustrate how these rankings can be used to evaluate the impact of injuries and assess implied player worth.

In regards to my prediction of the line and over/under for the Knicks/Thunder game, here is how it panned out (see the bottom of the post linked above):

Predicted Line: 8.0 points in favor of OKC
Actual Line: 11.0 points in favor of OKC

Predicted Over/Under: 204.5 points
Actual Over/Under: 197.0 points

I don't expect pinpoint accuracy when predicting lines and over/unders, but that's a pretty big miss (when back-tested, the model's average miss was about 1.5 points).

So here's how we can assess the "market value" of Melo's abilities:

The actual line came out 3 points higher in favor of the Thunder, compared to the BTM rankings. Since the rankings were calculated prior to Anthony's injury, the implication is that the market thinks he is worth about 3 net points over the course of a game.

Looking at the over/under miss adds a little more insight. The over/under is 7.5 points lower than the prediction. Combine that with the line miss, and you can draw the following conclusion (with a little algebra):

Offense: without Carmelo Anthony, the Knicks are expected to score 5 fewer points.
Defense: without Carmelo Anthony, the Knicks are expected to allow 2 fewer points.

Which nets out to the 3 point impact on the line and sums up to the 7 point impact on the over/under. This is consistent with the general consensus on Carmelo which is that he is somewhat of a defensive liability, but that his offensive talent outweighs his defensive shortcomings.

This is also somewhat consistent with Carmelo Anthony's Adjusted Plus/Minus Rating of 4 points from the 2010-2011 season (courtesy of basketballvalue.com). I chose not to compare to his adjusted plus/minus for 2011-2012 as I imagine those rankings are pretty volatile this early in the season (it's 8.5 points so far).