The most one-sided teammate record since 2014
ALB 0 SAR
Alexander Albon beat Logan Sargeant in 43 of their 43 duels as teammates — same car, same weekend.
Driver Rating
SAME CAR. WHO'S FASTER?
Every driver measured only against the person in the same machinery
All-time · 2014–2026
Every driver rated since 2014, retired ones included, on their average skill across their career, car removed. The likely rank is where each driver lands in 90% of re-draws of the ratings within their margin of error. Where ranges overlap the order is noise — read the table in groups, not as a queue. Even the best version of this model calls only about two teammate duels in three: drivers in equal cars are close.
The 2026 cars
How much each car is worth once its drivers are accounted for, on the same scale as the driver ratings (0 = the average car). This is what the driver table has already subtracted.
| # | Team | Car |
|---|---|---|
| 1 | Mercedes | +548 |
| 2 | Ferrari | +433 |
| 3 | McLaren | +365 |
| 4 | Red Bull | +177 |
| 5 | RB F1 Team | +147 |
| 6 | Alpine F1 Team | +89 |
| 7 | Audi | +43 |
| 8 | Haas F1 Team | −156 |
| 9 | Williams | −304 |
| 10 | Aston Martin | −647 |
| 11 | Cadillac F1 Team | −650 |
Receipts
What the duel record says once you look at it properly. Every number here is checkable on the page it links to.
The teammate battle that never had a winner
OCO 38–37 PER
Esteban Ocon and Sergio Pérez split 75 duels almost exactly down the middle: 38–37.
One was faster on Saturday. The other won on Sunday.
QUALI MAG 6 MSC · RACE MAG 11 MSC
Kevin Magnussen had the edge in qualifying, but it was Mick Schumacher who finished ahead more often — same car, same seasons.
The gap at the top is not close
VER 336–82 vs teammates
Max Verstappen rates 105 points clear of Nico Rosberg — a bigger gap than the one between second and tenth.
These drivers cannot be told apart
ROS 1859 · NOR 1783 · RUS 1780 · LEC 1769
Once the margin of error is taken seriously, Nico Rosberg, Lando Norris, George Russell, Charles Leclerc are statistically level. Any ranking that puts them in a strict order is inventing precision.
Not the driver he was
BOT FORM 1549 · CAREER 1635
Valtteri Bottas rates 86 points below his own career level once recent seasons are weighted heavier.
Settle an argument
Pick two drivers. The model reports who wins in equal machinery, and their real record as teammates if they ever shared a garage.
How it works — and what it can't see
A driver's finishing position mostly measures their car. Put anyone in the dominant chassis and they look like a champion. So this rating splits every result into two parts: how good the driver is and how good the car is, and only the first part is the driver's rating.
Every qualifying session and every race since 2014 is used as a full finishing order, not just "who beat their teammate". Each driver's strength in a session is their skill plus their car's level that season plus how well that car suited that particular track. The model (a Bayesian rank-ordered logit, the approach published by van Kesteren & Bergkamp in 2023) finds the skills and car levels that best explain every order at once. Teammates pin down the drivers, because inside a garage the car cancels out; drivers who change teams link the whole grid together.
Skill is allowed to change from season to season, so a driver's rise or decline shows up, and races count half as much as qualifying because strategy, safety cars and team orders make race orders noisier.
Mechanical retirements are discarded. Losing to your teammate on the day your gearbox failed measures the gearbox. Crashes and collisions do count — those are the driver's.
It was tested, not just built. Trained on every season before a given year and then asked to call that year's teammate duels — which it never saw — it matched or beat the previous teammate-only model: on 2023–2026, 67% of 1,271 held-out duels called right (log-loss 0.604 vs 0.6053). That is a real edge over a coin flip and a modest one. Two drivers in the same car are genuinely close, and anything claiming to call these duels nine times out of ten is selling something. What the new model adds is the car: it can compare drivers who never shared a garage, and say what a driver would do in a different car.
The margin of error is part of the answer. Every rating comes from thousands of posterior draws, so each one is published with the range the evidence supports, and when two ranges overlap this page says "too close to call" instead of inventing a winner.
The honest limits, because a number worth sharing should carry them:
- It cannot see team orders, split strategies, or who got the upgrade first — an upgrade given to one car only is credited to that driver.
- Car and driver are only separable because drivers change teams. A driver who has only ever had one teammate is measured almost entirely against that teammate.
- Race and qualifying duels from the same weekend are correlated, so the ranges are, if anything, slightly optimistic.
- Drivers with fewer than 60 duels are left out of the all-time table, and marked provisional on the current grid (at least 20 needed to appear). One strong season is not a career.
- The rating uses finishing orders, not lap-time margins. The quali gap column shows the margin (median % of lap time to the teammate this season) as evidence — we tested using the margin instead, and it did not predict future duels any better.
Built on the public F1 results archive (Jolpica/Ergast), 2014–2026.
267 RACE WEEKENDS ·
4,534 TEAMMATE DUELS (1,904 RACE + 2,630 QUALIFYING) ·
13 SEASONS
· UPDATED 27 Sep 2026
The rest of the grid data
Real telemetry, head-to-heads, live title scenarios and a prediction record you can check. Free to look around.