Our NBA Playoff model calculates win probabilities for each series matchup based on each team’s regular season statistics. We used multiple regression techniques to develop a variety of models. We arrive at our consensus winners by taking the average prediction from a handful of our most successful models. For a more detailed breakdown of these NBA models, and to see our 2015 & 2016 brackets, check out our NBA Playoff Bracket Methodology. Our model has picked 78% of series winners correctly, along with 15 of the past 27 champions. In other words, on an average year, it gets 11.7 of the 15 series winners correct and has a 56% chance of getting the champion correct. Our predictions for the 2017 playoffs are shown below:
Continue reading 2017 NBA Playoff Bracket →

Our data is built off of all regular season games since the 1989-1990 regular season, giving us 28 seasons in all. It was fun to test our models performance over the Jordan-dominated 90’s, the Shaq and Kobe 00’s, and to the modern pace-and-space era. There is no doubt that the style of play in the league has changed drastically over this time. This poses a challenge for modeling the most recent seasons, as most of our training data will not reflect the boom in outside shooting that we are seeing in the game today.
Continue reading NBA Modeling Methodology →