NBA training camp starts next week, and as teams prepare for preseason, it's time to make projections for how all 30 will perform this upcoming season.
My annual projections combine box score numbers from my SCHOENE projection system, adjusted plus-minus data and my guesses at rotations for every team based on current injuries.
After a dismal year for the projections, which correctly predicted just nine of 30 teams relative to their over/under win totals last season, I spent much of this summer revisiting the model. The big change was weighting the luck-adjusted RAPM portion of player projections, based on data from Krishna Narsu, using minutes played over the previous three seasons. This adjustment, and other smaller ones, produced better out-of-sample results in testing.
One key change is the wins projections are no longer so compressed toward .500. As we are looking at mean wins, they still tend toward average because of the possibility that teams could dramatically under or over perform based on injuries and breakout seasons, but four teams are projected to win at least 50 games on average after only the eventual champion Boston Celtics were last season.
Let's take a look at who joins the Celtics at the top, which Western Conference finalist might fall down the standings and the rest of the projections.
Jump to:
ATL | BOS | BKN | CHA | CHI | CLE
DAL | DEN | DET | GS | HOU | IND
LAC | LAL | MEM | MIA | MIL | MIN
NO | NY | OKC | ORL | PHI | PHX
POR | SAC | SA | TOR | UTA | WAS