How to Predict F1 Race Weekends with MATH?! 🤯👀 #formula1 #f1 #motorsport #grandprix

Added: 2026-05-26

[00:00:00]I built a mathematical formula to predict Formula 1 results, and here's exactly how it works. Now, I'm not claiming this is the best formula, but it's the formula I could come up with, and I think it's pretty good. This series is called Trust the Formula.

[00:00:11]Every race weekend I run two predictions, one for qualifying and one for the race, and they use completely different data. And I call the final number the Club Chicane Score. For qualifying, I'm asking one core question: How fast has this driver actually been when it matters on a Saturday? I take every grid position from the season, average them out, and then flip the number. I subtract from 23 because there are 22 cars on the grid.

[00:00:31]It essentially means that P1 becomes 22 and P22 becomes one. And then I divide by 22 to get everything on the same scale. And the Club Chicane Score higher always means better. This is our proxy for raw one-lap pace. And then I look at the last two weekends. Trending better than your season average, score goes up by 10%. If it's trending worse, it goes down by 10%. And this is our proxy for driver momentum. That right there is 60% of our qualifying score. The other 40 is simple: Where did they qualify here last year? P1 gets a 10 and P10 gets a one.

[00:00:59]Anything below that, a DNF, or a DNS gets a zero. And this is a proxy for how good a driver is at any particular circuit. Now, onto race predictions. The race prediction uses the same 60/40 structure. However, championship points replace the qualifying positions, and last year's race result replaces last year's qualifying. For race results, 60% of our prediction comes from championship points. This is our proxy for racecraft, reliability, and point-scoring consistency. And the same momentum nudge applies. Up or down 10% based on the last two average race finishes. Now, for the other 40% of our race results prediction, we use last year's race result. Similar to before, this is our proxy for a driver's history at a particular circuit. Now, after we've done all that math, we combine them, and every driver gets a Club Chicane Score between zero and one.

[00:01:42]Those are ranked, and then that's your prediction. Now, the system is flawed because it doesn't care about context, only the data. Every race weekend for it, I'm going to be running these numbers, and I think it'll be pretty interesting. So, be sure to subscribe or follow if you want to join me in this journey. We'll call it a journey of the data.