Chuck's Driver Rankings & Stats

Frequently Asked Questions

How are ratings calculated?

Ratings are computed using a modified version of the Elo rating system which has been used for decades to build world chess rankings.

For each race, finishing in front of another driver is considered a victory and finishing behind a defeat for the purpose of the Elo system.

A small fraction of each of these wins and defeats is then used to calculate a single rating diffenrential to apply to this race.

Further more, each differential is split equally between every driver in the case of endurance races, meaning their ratings moves half-speed if they share driving duties for a race

All drivers start their careers with an rating of 1800 points, meaning that any driver with a rating above 1800 is above the average and vice-versa

To create some basic numbers, the system also loops 100 times on the oldest 500 races before computing the current ratings.

My favorite series isn't included, how come?

Two major reasons might explain the absence of your favorite series :

• To build a somewhat precise comparison basis, I had to pick some criterias for series to be included in this system. In no particular order, they are : the series must run on closed courses, with all cars at the same time on the track (no rally), race four wheels vehicules (no superbikes, motocross) and run on asphalt only* (no sprint cars)
• The other reason might as well be that I didn't have time, resources or interest to research a particular series or year. In such a case, I'd be more than happy to include the series if you send-in race results at charles@cafefroid.com

*An exception was made for the ARCA dirt races as they are a small part of an asphalt majority series.

Doesn't your system penalize good drivers running with bad cars?

Yup, the only metric I use is race results. This system measures success, not talent, which is a much more subjective thing.

On the other hand, good drivers usually find their way up to great teams, so things should be fairly accurate in the long run.

Wouldn't your database be much cooler with starting positions, laps lead, teams etc.?

Yeah, of course it would. But I made some choices to focus a bit more on the task at hand : finding if a driver is better than another

Since there are so many variations to the starting grid building process, those numbers would have not been significant at all in this context

DriverDB.com is an excellent source for these kind of informations, so do the official series sites

Where did you get those flags?

They are from a nice (and free!) set made by famfamfam