THE BOOK--Playing The Percentages In Baseball

Interesting stuff. It would of been nice to see some sort of adjustment for pace factor. I guess you can’t really project the Pace that a team is going to play at, unless you looked at the coaches history. We know that the Knicks under Mike Dantoni will have a lot more possessions this year than they did last year with Isiah Thomas as their coach.

The only statistician that I know of who does NBA player projections is John Hollinger. I’ve never seen anyone actually attempt to project counting stats (PTS, RBD, AST...) like we do in baseball. In my opinion, the sabermetric concepts are easier to grasp and much more intuitive than APBRmetric concepts. I don’t know much about the uberstats, such as PER and Tendex. In sabermetric terms, Tendex looks like Total Average. Approximate Value is Tendex^(3/4)/21.

I don’t see any regression in the description.  Did I miss that?

Also, I didn’t look at the stuff too closely, but the categories are not “per possession” or 100 possessions?  It is true that you can’t use a player’s stats “per minute” because of the pace factor.  Pace is easy enough to calculate (estimate) from the team stats.

Pace adjustments are similar to Park adjustments. Last season the Pistons had the lowest pace factor, but ranked 6th in offensive effeciency (pts per 100 possessions).

In 1990-91 the Nuggets led the league in scoring at 120 PPG. However, they ranked 21st out of 27 teams in offensive efficiency. They also allowed 130 PPG game and had the worst defensive efficiency rating of all-time. They also had the worst record (20-62) in the league that season.

Sure, they are just like park adjustments/factors, but much easier to compute and pretty much apply equally to everyone.  If you have PBP data, you can computer them exactly.  Without, you can still compute them pretty close to exactly.