Relatively poor results this past week, 44 of 94 games, 47% accuracy overall. Given our relatively strong performance in the NBA this past season, our early MLB results have been a bit disappointing. However, after doing a bit of learnin’, we realized an interesting point: Professional baseball teams are simply much more balanced than their NBA counterparts. A statistic illustrating this point is provided by the winning rates of the top teams: This past year, the Warriors had a win rate just over 80%, while the top team in the MLB last season, the Angels, had just under a 61% win rate. This difference explains why we can’t achieve 70% win prediction accuracy in the MLB. Instead, we’ll be shooting for just 56% accuracy this season — this may sound modest, but we expect this rate will be a challenge to meet. To work towards that goal, we’ve already begun to implement a number of improvements to our algorithm. New predictions are posted.

Point spread # games Accuracy
<= 1 23 43%
<=2 20 45%
<=5 32 41%
>5 19 63%

Share this post on: Twitter | Facebook | Email

Jonathan Landy Avatar Jonathan Landy 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 worked as a data-scientist at Square for four years and is now working on a quantitative investing startup.

comments powered by Disqus



MLB prediction project