We finished the NBA regular season strong, with correct guesses on 41 of 56 games this past week — 73% accuracy. Overall, our season average was roughly 65%. This is significantly lower than those returned by the bookies this year, whose accuracy overall sat just under 70%. The differences there likely have much to do with information relating to individual players (eg injuries) that would be difficult to take into account in our team-centric modeling approach. It also took us a number of weeks to come up with a solid way of choosing our model’s hyperparameters. With these issues identified/settled, we anticipate improvements on these numbers next year. We won’t be modeling the NBA playoffs, so this will be our last basketball post for the current season — we had a blast!
With regards to the MLB, in our first week, we guessed correctly on 52/93 games, giving 56%. A somewhat poor result, fueled in part by the lackluster Giants. New predictions are up, with better accuracy hoped for this week!
Jonathan grew up in the midwest and then went to school at Caltech and UCLA. Following this, he did two postdocs, one at UCSB and one at UC Berkeley. His academic research focused primarily on applications of statistical mechanics, but his professional passion has always been in the mastering, development, and practical application of slick math methods/tools. He currently works as a data-scientist at Stitch Fix.