Latest Updates
- June 13, 2024: Initial launch of the Glicko model as something of a "proof of concept". Ratings are based on the 2023-2024 season to give a starting point and to show what I'm planning to do with some of the charts and tables. I would expect that accuracy of the model will rapidly increase once I start getting more data on new/returning drivers. The Win Probability Over Time chart will especially improve as we get more data.
Latest Driver Ratings
Who's Favored to Secure Pole, the Win, a Podium, and Points?
How Many Points is Each Driver Expected to Score?
Win Probability Over Time
Methodology
Long-time acquaintances of Jeff know he is a huge fan of the Elo and Glicko rating systems. Elo is a simple way to grade competitors based on a series of head-to-head results (Glicko is the same, but less simple). After initially using Elo ratings to create a race outcome projection model, the current model uses the Glicko system. Unlike most sports, Formula One is not a head-to-head sport; multiple drivers take to the track at the same time or in the same session. In order to make this work, each session or race is treated as if it was a round-robin one-on-one tournament. A driver who finishes second out of 12 cars is viewed as having gone 10-1, losing to the first-place finisher and defeating the rest.
All drivers are assigned Glicko ratings going into each qualifying session and race, which represents the model's estimate of their form at that particular moment. After each event, the driver's rating will changed based on the result. In general, finishing high helps you gain Glicko points, while finishing low costs you Glicko points. If a driver doesn't finish a race, Glicko acts as though that driver never entered the race, but the driver's One-Lap Reliability is affected (this metric helps inform the model's expectation that a driver will finish a given race).
Each driver starts with a Glicko rating of 1500 (the best racers can achieve ratings of 2000+).
Special thanks to Justin Moore's posts on his IRL F1 prediction model, 538's post debating the best F1 driver of all-time (Justin Moore was also involved with this), and Glicko creator Mark Glickman's white paper on his Glicko system.