Baseball From Beantown

    Rejoice! Baseball is back! The 2026 MLB season kicked off last Wednesday, and new data is already being plugged into Whiff+ and other models which I will get into later. Anyways, here's a quick update on Whiff+ and some other projects I've been working on. Whiff+ Expansion     Whiff+ has been expanded to include every major pitch type, using the same methods I went over in the original Heat+ article. All of the non-fastball models are very much in their newest stage, and I haven't taken much time to go in and refine them. That being said, they are correlating well with whiff rates.       Sorry for the blurryness of that image, I'm looking for a new site to host this blog. Anyways, correlations are strong and show that high Whiff+ numbers correlate very well with whiff rates. Splitters have the best correlation, at an R value of .742. Slider is the lowest at .56. All are between .56 and .75.           I have a...

    About a month ago, I released my first writeup on my initial stuff model Heat+. Since then, in between class and spring break, I've made some major upgrades and fundamental changes to the model that have greatly improved whiff rate prediction power. I've also renamed the model, from Heat+ to Whiff+ because I think that sounds better and is more descriptive of what the model tracks. In terms of future changes, I would like to implement a location aspect to my model, as fastballs at the top of the zone are much more effective.  Fundamental Change      The biggest change made comes from what outcome the model is being trained to look for. Stuff+, Heat+ and PitchingBot are trained on Run Value, meaning it looks at the true outcome of the pitch. Whiff+ is trained Whiff Probability, meaning it looks at whether or not the batter made contact with the pitch. This removes a lot of external factors and focuses more on the physical qualities of the pitch and its a...