One number for how good a driver is, on the same scale for everyone we index.
Every driver here gets one rating. F1, WEC, IMSA, a club weekend at Mid-Ohio. If we've got the results, it all feeds the same number. Same idea as a chess rating or iRating: you take points off the people you beat and hand points to the people who beat you. How much depends on who they were. Beat someone way above you and it's worth a lot. Beat a field way below you and it's worth almost nothing, because you were supposed to.
We run every race we have in order, oldest first, and let the whole thing play out. That's 38,983 drivers across 149 series right now. The bottom of the scale is 400. The top is somewhere near 9,000, and getting there means beating everybody, everywhere, for about ten years.
One thing to be clear about up front: this is a measure of who you beat, not a trophy count. It doesn't know what a championship is worth or which win was the famous one. A driver with a thin record we haven't finished indexing will read low, and that's a gap in our data rather than a verdict on them.
Here's roughly what each level looks like. The lines aren't hard. Somebody sitting right on the edge of a band could honestly go either way, but this is the general shape of it.
| Rating | Level |
|---|---|
| 8,000+ | Generational The best in the world. Winning Formula 1 races outright. |
| 6,500–7,999 | Elite professional Wins at the very top. The front of Formula 1 and IndyCar. |
| 5,000–6,499 | Established professional A full-time seat at the top level, or a factory sportscar driver. Wins in the strongest fields anywhere. |
| 4,000–4,999 | Professional Runs at the front of a national championship, or the sharp end of a good feeder series. |
| 3,000–3,999 | Entry-level professional On professional grids without running at the front of them. The midfield of a national championship, the pro-am classes of international endurance, and the feeder single-seaters. A driver more than two seasons past their debut here is simply a professional: entry-level describes the way in, not a career spent at this level. |
| 2,500–2,999 | Semi-professional Paid for some of their drives, and racing in fields that regularly include full professionals, without a professional seat of their own. |
| 2,000–2,499 | Competitive amateur The front of club racing and the entry end of pro-am. Quick, and usually paying for it themselves. |
| 1,350–1,999 | Club Club, regional and vintage racing. A real licence and real results, in fields that never see a professional. |
| 0–1,349 | Developing At the back of club fields, or not enough races on record yet to have moved off the starting level. |
You're scored against every other car in your race, one pair at a time. Finish ahead of one and you take rating off its drivers. Finish behind and you pay. The most a single race will move you is about 300 points, split across everyone you raced. And if you retire from the lead you still score behind everyone who finished. Doesn't matter why the car stopped.
Winning a race you were always going to win is worth very little. The further below you the field is, the less it pays, down to almost nothing against a club grid. That's there to stop somebody farming an easy series into a big number. It only applies to genuinely amateur opposition though. Beat a real professional and you get paid in full, however far ahead of them you already were, because that result actually says something.
Losing is never discounted. If you lose to a field you were supposed to beat, it costs you the full amount. That's deliberate, and it's how iRacing works too: the favourite carries the risk. So a stack of easy wins won't buy you what one hard weekend in a strong field will, and one bad day against people you should have beaten is expensive.
In endurance racing three or four drivers share one finishing position. We price the car at the average of its crew, since that's what everyone else on track was actually up against. But the same result doesn't mean the same thing for everyone in it. A podium is worth a lot more to the amateur in the car than to the pro sharing it, and a bad day costs the pro more. iRacing does shared cars the same way. The higher you're rated, the more you've got riding on it. Where the timing data tells us who actually drove, we use it to drop people who were entered but never got in the car.
This is the least precise thing we do, and it's worth knowing about. One finishing position has to cover everyone who drove, so a fast driver sharing with slower ones reads lower than they really are, and there's no way around it from a results sheet. A driver can build a lead over a long stint and still take the loss when someone else puts the car in the wall on their first lap. We don't try to guess who was at fault. Over enough races it comes out in the wash, but for a professional whose record is mostly amateur endurance, the number will sit lower than their pace deserves.
Your first race sets your starting point. Debut in Spanish F4 and you start around 3,250, because that's what Spanish F4 is. Debut in IMSA's GTD class and you start near 4,150. We set these per series and per class, so a GTE-Am gentleman driver and a Hypercar factory driver don't start in the same place just because they showed up the same weekend. Move up to a bigger series later and you bank part of that step when you get there, but only if your results had already cleared the level you were leaving. Buying a seat in a faster series doesn't do anything on its own.
On a driver's page, the first season on record shows its gain counted from 1,350 rather than from wherever they started. Otherwise a rookie who came in at 2,900 and finished the year at 3,100 would look like they'd gained 200 points, and the number they actually arrived at would be invisible. Every season after that shows the real race-by-race movement.
Nothing protects the rating once you've got it. A season at the back of a strong field costs whatever it costs, and there's no floor propping you up because of what you used to be.
Those starting levels aren't guesses we made once and left alone. We keep re-checking them against drivers who actually move between series, thousands of them at this point, and we move a series when the evidence says we had it wrong. That's how we catch a grid full of professionals sitting at a club-level number, or an amateur class quietly starting at its series' pro number.
An FIA Platinum or Gold grade gets you a small bump on entry and nothing else. Everything after that has to show up in results.
A rating built on two races is mostly just the starting level talking, so we pull new entries toward the middle until they've got about five starts. It's also why a genuinely quick driver can look low here when we've only got a small piece of their career. That's usually what's going on when a number looks off, and it sorts itself out as we fill the record in.
Retire and the rating fades a step each year until it hits a floor. Champions end up near the level they actually raced at instead of sliding to the bottom of the list. Everything you won still shows on your page. The rating just stops pretending you'd go do it again tomorrow.
Which means a retired driver's number isn't their peak, and you shouldn't read it as one. It's a current estimate that's been marked down for time away. If you're comparing a great from thirty years ago against somebody racing this weekend, you're comparing two different things, and the active driver has the advantage.
Every series page has a Strength of Field, which is the average rating of the stronger half of its roster. That roster is the last two seasons the series ran, so you're seeing who actually races it now instead of some star who turned up once years ago.
For a series that's finished, it's the last two seasons it ran, priced at today's ratings. The 2017 and 2018 European F3 grids had Lando Norris and Alex Palou in them, so that page reads really strong. That's the alumni talking. It means those kids turned out great, not that European F3 was tougher than Formula 2.
Every rating comes from official results, replayed in full. We don't make anything up. When a number looks wrong it's usually because there's a chunk of that driver's career we haven't indexed yet, and we're still adding.