Planning my next project... a UFC/MMA analytics platform!
New project idea, working title @lien_mma_corner: a UFC/MMA analytics platform. The question I want it to answer is simple to say and a lot harder to build - given two fighters, use everything we may know before the fight to say who’s statistically more likely to win, and why, not just spit out a number.
The plan is a full pipeline: pull in historical UFC fight data and fighter stats, engineer features from only what was actually known before each fight happened, train a model on that, and keep the reasoning visible alongside the prediction instead of a black box. It’s an analytics and research project, not a betting tool, and scoped small enough to build over a few weekends rather than a finished prediction engine on day one. Python again, same as always, with a real database behind it this time. More once there’s actually something running.