Pace
95.0
Race
96.4
Tyre
67.3
Quali
88.5
Cons
70.2
Disc
75.9
Pace
94.5
Race
87.2
Tyre
78.9
Quali
91.4
Cons
71.5
Disc
69.8
Pace
94.1
Race
75.9
Tyre
70.6
Quali
86.4
Cons
72.2
Disc
75.5
Pace
89.4
Race
86.6
Tyre
60.1
Quali
86.3
Cons
74.0
Disc
71.1
Pace
89.9
Race
86.2
Tyre
60.6
Quali
87.0
Cons
68.7
Disc
73.6
Pace
87.5
Race
85.3
Tyre
72.4
Quali
89.0
Cons
63.4
Disc
69.4
Pace
82.1
Race
82.6
Tyre
65.3
Quali
78.2
Cons
69.4
Disc
70.0
Pace
77.2
Race
66.4
Tyre
76.6
Quali
78.6
Cons
69.7
Disc
72.8
Pace
71.1
Race
72.0
Tyre
79.4
Quali
72.5
Cons
74.6
Disc
72.6
Pace
75.6
Race
63.5
Tyre
67.0
Quali
75.6
Cons
71.0
Disc
66.7
Pace
68.1
Race
74.0
Tyre
70.2
Quali
70.9
Cons
71.3
Disc
72.1
Pace
69.9
Race
66.3
Tyre
73.5
Quali
68.7
Cons
72.9
Disc
62.6
Pace
66.3
Race
57.2
Tyre
67.6
Quali
72.2
Cons
68.8
Disc
71.8
Pace
66.1
Race
56.1
Tyre
72.3
Quali
68.1
Cons
70.6
Disc
70.6
Pace
64.0
Race
59.0
Tyre
75.2
Quali
62.9
Cons
63.0
Disc
59.8
Pace
57.6
Race
58.6
Tyre
66.4
Quali
67.2
Cons
67.7
Disc
64.6
Pace
52.6
Race
58.8
Tyre
76.3
Quali
65.5
Cons
61.5
Disc
70.3
Pace
57.8
Race
58.3
Tyre
59.2
Quali
58.6
Cons
63.1
Disc
68.3
Pace
49.4
Race
64.5
Tyre
64.7
Quali
58.6
Cons
62.2
Disc
62.9
Not ranked: raced 3 of 20 rounds, below the third of the season needed to rank against the field.
Pace
49.9
Race
61.6
Tyre
75.9
Quali
51.5
Cons
61.3
Disc
71.5
Not ranked: graded in 13 of 20 rounds, below the 75% needed to rank against the field. The other 7 ended in early retirements or non-starts (7 mechanical).
Not ranked: raced 1 of 20 rounds, below the third of the season needed to rank against the field.
Pace
43.5
Race
37.4
Tyre
70.0
Quali
52.2
Cons
70.0
Disc
72.8
Not ranked: raced 1 of 20 rounds, below the third of the season needed to rank against the field.
Pace
40.9
Race
46.5
Tyre
58.4
Quali
55.3
Cons
52.3
Disc
70.3
Not ranked: graded in 14 of 20 rounds, below the 75% needed to rank against the field. The other 6 ended in early retirements or non-starts (6 mechanical).
Overall grades, 2019 to 2026, across 169 driver seasons. The median is 69 and 7% reach elite.
Below average
Below 6013%
Average
60 to 69.941%
Strong
70 to 79.922%
Excellent
80 to 89.917%
Elite
90+7%
Six sub-grades, each weighted differently: Pace (40%), Racecraft (20%), Qualifying (25%), Tyre Management (5%), Discipline (5%), and Consistency (5%). Pace, Racecraft and Qualifying are graded on fixed scales (real pace gaps and a frozen scale (Pace and Qualifying set on 2023 to 2026, Racecraft on 2019 to 2026), so the same performance earns the same grade whoever else raced that season), so most drivers land 60–75, standouts reach 80+, and 90+ is genuinely elite. Discipline and Consistency are absolute scales, set so a typical driver reads 70 like every other grade (a clean record is the norm, so Discipline only really moves for drivers who pick up penalties), and Consistency measures race-to-race steadiness on its own terms. Sprints count too, at a quarter of a Grand Prix's weight, in Pace, Racecraft and Qualifying.
Fuel-corrected lap times, blended 60% how far the driver's pace is from the field median and 40% the gap to their teammate. The field share mostly reflects the car; the teammate share is what separates driver from car. You can be in the fastest car and still grade low on Pace if your teammate is keeping up everywhere.
Three things, measured lap by lap: passes made on track (60%), defending a position while a car is close behind (20%), and the opening lap (20%). Defense is half time spent holding a car off and half whether the car behind got by when a model said it was likely to. Passes count in proportion to the chances a driver had, so leading from the front is not marked down. Pit-stop gains do not count, and each race is scored on one fixed scale with extreme single-race scores clipped, then blended 65/35 with the gap to the teammate in the same race, so beating a teammate in the same car counts for more than the car alone would suggest. The teammate share shrinks when the gap is very large, because that says more about the teammate than about the driver. Starting P10 and finishing P4 through real passes grades much higher than starting P3 and finishing P6 without a fight, and a pole-to-win drive earns credit for the pressure it held off, not for the result alone.
Tyre Management rewards drivers whose tyres hold their pace and whose stints run long. 70% of it is how much lap time each lap of tyre age costs the driver compared with the field on the same compound in the same race, estimated from every clean lap of the season (a separate wear rate for each race and compound, and running in dirty air allowed for). The other 30% is how much longer the driver's stints ran than the field's median on that compound. Because it is measured against the field, it also reflects a team's strategy and how kind its car is to tyres, not only the driver, and it does not separate teammates. About 44% of the differences it shows across a full season is consistent, so each season's reading is pulled toward a neutral 70 by how much evidence there is, and each grade carries a range of roughly 7 to 9 points. This grade is provisional while we see how it holds up.
Pace, Qualifying, Racecraft and Tyre all use the teammate in the other car as a reference, but only partly: Pace is 40% teammate gap and Qualifying leans on the teammate more further back on the grid. A driver who consistently outqualifies a strong teammate earns more qualifying credit than one who barely edges a weaker car.
Yes, but only so far. About 70% of the variation in overall grades follows the car's pace, because part of Pace, Qualifying and Racecraft is how a driver ranks against the field. A driver who clearly beats their car is credited for it: Alonso's 2023 grade of 81 in the fourth-fastest car, and Verstappen's 90 in 2025 in the fourth-fastest car, sit well above what those cars predict. Drivers in the slowest cars have typically graded between 55 and 70. Read a grade next to the team's Car Pace grade, and treat the gap to the teammate as the cleanest read on the driver alone.
How reliably you know what you'll get from a driver on a Sunday: the race-to-race swing in their performance level, not how good that level is. A steady midfielder and a steady front-runner both score well; a driver who is brilliant one week and anonymous the next scores low. Mechanical DNFs are dropped, but a driver-error crash stays in, since a shunt is the ultimate inconsistency. It's capped by the driver's own level, so 'reliably slow' can't outrank 'reliably fast', and it needs at least five races before it produces a real number. It is also the least certain grade after Tyre Management: the race-to-race spread is partly the car suiting some tracks and not others, so we keep its weight at 5%.
On-track conduct sanctions, weighted by how serious the offence was: causing a collision or dangerous driving costs far more than a track-limits penalty. Stewards' investigations that ended in no further action count as a light nudge: a driver investigated for contact repeatedly and cleared every time isn't quite spotless. And a driver who raced wheel-to-wheel all year without incident grades above one who stayed clean simply by never being in a fight. Technical infringements (a plank-wear disqualification, a fuel-flow breach) and unsafe releases from the pit box are the team's, not the driver's, and are excluded.
No. Mechanical failures are excluded from every sub-grade. If the car failed, that session is skipped entirely for scoring purposes: the grades only reflect what the driver could actually control. A crash is treated differently. If a driver ends their own race in an incident, that round is scored as a fixed drop below their own usual level: 20 points on the overall grade for a crash that was their fault, 12 where no fault was established, with Racecraft dropping further. It stays in the Consistency calculation and carries a direct penalty to the season number. If the crash came in the second half of the race, Pace, Tyre and Qualifying keep their real readings from the laps driven; in the first half they are left out, since a handful of laps is not a real reading. A driver graded in fewer than 75% of the rounds they entered (for example after many early mechanical retirements) is shown below the ranked field as NR, with their grades and the reason, and still counts in tier tables and best-of lists.
After every race weekend, once all session data is fully ingested: typically within 24 hours of the race. The 'Last updated' stamp in the header shows exactly when grades last changed.
Constructor grades measure the five things a team controls, independent of who's driving: Car Pace (35%), Race Execution (20%), Reliability (20%), Pit Stops (15%), and Development (10%). Mechanical DNFs count against Reliability but are excluded from Race Execution: a brake failure doesn't tank the execution grade.
Every sub-grade is read from the team's own result in real units (percent of lap time, seconds per stop, places gained), not by ranking teams against each other. The scales were set on 2023 to 2026, where 70 is a typical team-season and the best seasons we have measured reach the mid-90s. 2019 to 2022 are graded on the same scales, with pit stops measured by pit-lane time only and less complete engine penalty records, so those seasons carry a little more uncertainty. Because a grade does not depend on who else is in that year's field, adding or removing a team never moves yours, and the same margin earns the same grade in any season. Each grade still says how a team did against the competition of its own era, since cars and regulations change, so Red Bull's 2023 and Mercedes' 2026 can be set side by side but are not the same car. A very poor result in one area bends toward a floor instead of falling to zero. The overall is the weighted average of the five grades shown, except that each point above 70 counts 1.7 times, so a team that leads in several areas stands out. A single race's grade is that race's own result and does not average to the season grade, because a failure costs a lot in its race but only a small share of a season.
Race pace and qualifying, 65 / 35, as how far the team's car is ahead of or behind the best other team, in percent of lap time. Race pace is the faster of its two cars with traffic and tyre age taken out, and qualifying is its best lap. Measuring against the best rival, not the middle of the field, means a dominant car reads as dominant and a close fight reads as close, independent of race execution or reliability. Top speed is shown in the breakdown but not scored: it reflects how much downforce a team runs, which lap times already capture, and in 2023 to 2025 it ran slightly against results.
Two things of equal weight. The first is the pit wall's calls, three team-controlled decisions, each scored against the alternative: the undercut or overcut at each green-flag stop, the pit-or-stay decision under a safety car versus the cars that chose the other way, and whether the number of stops matched the optimum from a tyre-degradation model fitted to that race. Forced stops and mechanical retirements are stripped out, and a chaotic race (red flag, wet, a rash of safety cars) counts for less because luck drives the result as much as judgment. F1 pit walls are all excellent, so this half is closely matched between teams and we scale its spread by a fixed amount rather than a statistical stretch, and race conversion is counted at a discount because it is noisy. The second is race conversion: for every car that finished, we compare its finishing place with the place its grid slot and the team's recent Car Pace (the last three Grands Prix, so a car that is being developed is judged on the pace it has now) imply, ranking both among the cars that finished, and average the team's cars. A team whose cars gain places beyond that is credited and one that loses them is marked down. Retirements are left to Reliability and the speed of the car is left to Car Pace, so neither is counted twice. A place lost to a driver's own conduct (a collision, track limits, a flag breach) is read as if the penalty had not been served, because that belongs to the driver's Discipline grade. A place lost to the team's operation (an unsafe release, pit-lane speeding, a missed penalty service, a car out of position) stays with the team.
Season-level only, and it asks one thing: how much has this team closed on, or lost ground to, the rest of the field this season? Each race weekend we measure the team's gap to the other teams in qualifying and in race pace (with traffic and tyre age taken out), as a percentage of lap time, and take the trend across the season. A team that starts far behind is expected to gain some ground, since it has more to find, so we credit only the gain beyond what a team starting that far back typically makes; a dominant team that held its lead earns a small credit for holding it. Being fast is Car Pace, not Development. The reference is the field: a team developing as fast as its rivals shows no gain. The grade is on a fixed scale, so a season where nobody moves stays flat, and it fades in over the first rounds because a trend from a handful of races is mostly noise.
A team's median stop time in each Grand Prix, measured against the other teams that weekend. We use the stationary (jack) time, the crew's own work, when most of a race's stops have one on record, and the time from pit entry to pit exit otherwise, because the public timing feed only lists a share of the stops in some races. The two agree closely over a season. We use the median, so one blistering stop or one fumble does not swing it, and a stop over 10 seconds (almost always a repair after damage) is counted but not scored.
It hits Reliability directly. Each mechanical DNF and any drop in completion rate counts against that sub-grade, but it's excluded from Race Execution, Car Pace, and Development, so one blown engine doesn't drag down the team's entire profile.
After every race weekend, once all session data is fully ingested: typically within 24 hours of the race. The 'Last updated' stamp in the header shows exactly when grades last changed.