How this works: These forecasts are based on 50,000 simulations of the rest of the season. Our player-based RAPTOR forecast doesn’t account for wins and losses; it is based entirely on our NBA player projections, which estimate each player’s future performance based on the trajectory of similar NBA players. These are combined with up-to-date depth charts — tracking injuries, trades, changes in playing time and other player transactions — to generate talent estimates for each team. A team’s full-strength rating assumes all of its key players are in the lineup. A team’s current rating reflects any injuries and rest days in effect at the moment of the team’s next game. Elo ratings — which power the pure Elo forecast — are a measure of team strength based on head-to-head results, margin of victory and quality of opponent. Read more »
Game metrics are on a 0-100 scale. Quality is determined by the harmonic mean of the teams’ Elo ratings; importance measures how much the result will alter playoff projections; the overall number is the average of the quality and importance values.
Download data. For historical team ratings, see the Complete History Of The NBA.
Design and development by Jay Boice. Statistical model by Nate Silver, Jay Boice, Neil Paine and Holly Fuong. Marc Finn and Andres Waters contributed research. Illustration by Elias Stein.