LIVE
NEXT São Paulo GP · Sun 17:00 GMT · in 3d 12h LAST Mexico City — VER P1, NOR P2, LEC P3 WDC NOR 412 · PIA 387 · VER 366 WCC McLaren 799 · Red Bull 612 · Ferrari 484
Qualifying analysis

2026 qualifying predictions vs reality

5 races scored · simple position-based model · ← back to predictions

For each completed 2026 qualifying session, this page shows every driver's predicted P(pole) and the four inputs that produced it (team's recent form, team's track history, driver's recent gap to teammate, driver's track gap to teammate), plus the actual qualifying position they finished in. Each table is sortable by clicking column headers.

Qualifying prediction formula

Each driver gets four summary numbers, all on a 0–21 scale where higher = better (an Elo-style direction). Driver captures the driver's long-history qualifying strength, Team the car's long-history qualifying strength, Track the team-and-driver fit at this specific circuit, and Form the combined team-and-driver form over the last 12 races. Each is a weighted blend of its sub-metrics (Driver leans 80% on pace, Team 70% on ratings, Track and Form are 50/50), then inverted so 21 reads as "pole-quality" and 0 as "back of grid".

Score = 0.20 × Driver + 0.30 × Team + 0.25 × Track + 0.25 × Form
Driver = 0.2 × Driver Ratings + 0.8 × Driver Pace
Team = 0.7 × Team Ratings + 0.3 × Team Pace
Track = (Team Track + Driver Track) ÷ 2
Form = (Team Form + Driver Form) ÷ 2

Sub-metric meanings (each on the same 0–21 scale, higher = better; hover cells for per-driver values):

  • Ratings — current quali Elo, mapped linearly across the active grid so the top rating scores ~21 and the lowest scores ~1.
  • Pace — career average qualifying position, inverted (lifetime P3 → score 19; lifetime P15 → score 7). Heavily car-influenced for drivers; a long-history measure for teams.
  • Track — career median qualifying position at this specific circuit, inverted. The Track Score column averages the team-side (best of their two drivers' positions at this circuit) with the driver-side (this driver's position at this circuit).
  • Form — recency-weighted average qualifying position over the last 12 races (ρ=0.7 decay; most recent race ~30% of total weight), inverted. The Form Score column averages the team-side and driver-side form signals.

The 0.30 / 0.20 / 0.25 / 0.25 weighting splits the prediction across long-history car (30%), long-history driver (20%), circuit-specific (25%), and recent-form (25%) signals. Each sub-component links to a deeper view: Driver ratings, Driver pace, Driver form, Driver track, Team ratings, Team pace, Team form, Team track.

Model accuracy

How the combined-score formula above has performed against the 4 completed 2026 races (85 driver-events). Each race is scored using only data available before that weekend's qualifying.

3/4
Pole sitters predicted
75.0% hit rate · the model's predicted P1 actually took pole this often
2.24
Mean rank error
Average distance between predicted and actual qualifying position, across every driver-event
2.36
E[pos] error
|expected position − actual position| averaged over all 4 races. Uses the Monte Carlo expected-position output
0.857
Rank correlation (ρ)
Spearman ρ between predicted and actual full-grid order. 1.0 is perfect; 0.0 is random
2.00/3
Top-3 overlap
Of the predicted top-3, how many were in the actual top-3 (averaged per race)
8.75/10
Top-10 overlap
Of the predicted top-10, how many were in the actual top-10 (averaged per race)
0.198
Pole log-loss
Per driver-event. Uniform-prior baseline: 0.185. worse than baseline

Per-race breakdown

RoundRacePredicted poleActual poleTop-3 overlapRank MAEE[pos] MAEρ
R1Australian Grand PrixRussellRussell2/32.212.390.856
R2Chinese Grand PrixAntonelliAntonelli2/32.092.150.892
R3Japanese Grand PrixAntonelliAntonelli2/32.272.400.852
R4Miami Grand PrixVerstappenAntonelli2/32.362.510.827

Sample is small — 4 races, 85 driver-events. Numbers will firm up as the season progresses. The model's predicted top-3 is correct on average to within 2.0 of 3 places, and its full-grid ordering correlates at ρ=0.857 with reality.

R1

Australian Grand Prix

Actual pole: Russell
# DriverP(pole) E[pos] Score Driver Team Track Form Actual Verdict
1Russell20.8%3.618.0213.4319.4118.5019.551POLE
2Verstappen19.2%3.817.8618.3118.0418.2516.91
3Antonelli15.0%4.117.5313.3519.4116.5019.662±2
4Norris13.5%4.317.3514.6319.4116.2518.146±2
5Piastri12.1%4.417.2915.0719.4116.0017.825±1
6Leclerc8.9%4.816.9315.7016.5017.5017.834±1
7Hamilton8.3%4.916.8515.8916.5018.7516.137±2
8Hadjar2.2%6.515.4912.2618.0416.7513.763±3
9Sainz0.0%12.310.6013.557.4814.008.58
10Gasly0.0%12.510.5111.128.6610.5012.2514±2
11Lindblad0.0%12.910.288.869.3112.5010.359±4
12Lawson0.0%13.69.838.179.3110.5011.098±6
13Hülkenberg0.0%13.99.6410.517.2410.2511.2111±3
14Alonso0.0%15.09.0614.096.3311.505.8617±2
15Albon0.0%15.08.9910.037.4811.757.2215±0
16Colapinto0.0%15.38.816.018.669.5010.5516±1
17Bearman0.0%15.58.768.706.7810.009.9312±3
18Bortoleto0.0%15.68.678.487.249.509.7310±6
19Ocon0.0%15.78.589.246.789.009.8013±3
20Stroll0.0%18.66.756.476.339.754.49
21Bottas0.0%19.95.7113.331.607.502.7419±1
22Pérez0.0%20.64.9010.761.606.502.5518±3
Verdict: predicted vs actual quali position. POLE = right pole sitter. ±N = off by N positions.
R2

Chinese Grand Prix

Actual pole: Antonelli
# DriverP(pole) E[pos] Score Driver Team Track Form Actual Verdict
1Antonelli22.7%3.518.2213.3519.4119.2519.661POLE
2Russell20.9%3.618.1413.4319.4119.0019.552±2
3Piastri13.6%4.217.5415.0719.4117.0017.825±1
4Norris13.3%4.217.5314.6319.4117.0018.146±2
5Verstappen11.7%4.417.3618.3118.0416.2516.918±4
6Hamilton8.3%4.816.9715.8916.5019.2516.133±2
7Leclerc8.0%4.916.9315.7016.5017.5017.834±1
8Hadjar1.4%6.815.1812.2618.0415.5013.769±2
9Gasly0.0%12.010.6311.128.6611.0012.257±5
10Hülkenberg0.0%13.49.8310.517.2411.0011.2111±2
11Alonso0.0%13.69.6814.096.3314.005.8619±5
12Colapinto0.0%13.89.566.018.6612.5010.5512±2
13Sainz0.0%14.09.4113.557.489.258.5817±3
14Lindblad0.0%14.09.408.869.319.0010.3515±1
15Lawson0.0%14.49.208.179.318.0011.0914±0
16Ocon0.0%14.98.969.246.7810.509.8013±2
17Bearman0.0%15.18.828.706.7810.259.9310±5
18Albon0.0%16.38.1810.037.488.507.2218±2
19Bortoleto0.0%16.48.118.487.247.259.7316±0
20Stroll0.0%18.86.576.476.339.004.4921±2
21Bottas0.0%19.26.2113.331.609.502.7420±1
22Pérez0.0%20.74.7710.761.606.002.5522±1
Verdict: predicted vs actual quali position. POLE = right pole sitter. ±N = off by N positions.
R3

Japanese Grand Prix

Actual pole: Antonelli
# DriverP(pole) E[pos] Score Driver Team Track Form Actual Verdict
1Antonelli21.8%3.618.0913.3519.4118.7519.661POLE
2Verstappen18.6%3.817.9318.3118.0418.5016.9111±7
3Piastri14.0%4.217.5415.0719.4117.0017.823±1
4Russell13.4%4.217.5213.4319.4116.5019.552±2
5Norris13.2%4.317.4114.6319.4116.5018.145±1
6Leclerc9.9%4.717.0515.7016.5018.0017.834±1
7Hamilton7.0%5.116.7915.8916.5018.5016.136±1
8Hadjar2.2%6.515.4912.2618.0416.7513.768±2
9Gasly0.0%11.810.8211.128.6611.7512.257±5
10Lindblad0.0%13.39.848.869.3110.7510.3510±3
11Sainz0.0%13.69.7213.557.4810.508.5816±2
12Hülkenberg0.0%13.69.7010.517.2410.5011.2113±1
13Lawson0.0%14.19.398.179.318.7511.0914±0
14Alonso0.0%14.39.3114.096.3312.505.8621±7
15Colapinto0.0%14.98.946.018.6610.0010.5515±0
16Bortoleto0.0%15.18.808.487.2410.009.739±6
17Ocon0.0%15.48.719.246.789.509.8012±3
18Bearman0.0%15.68.518.706.789.009.9318±2
19Albon0.0%15.78.4910.037.489.757.2217±1
20Stroll0.0%19.06.386.476.338.254.4922±3
21Bottas0.0%20.05.5813.331.607.002.7420±0
22Pérez0.0%20.35.2710.761.608.002.5519±1
Verdict: predicted vs actual quali position. POLE = right pole sitter. ±N = off by N positions.
R4

Miami Grand Prix

Actual pole: Antonelli
# DriverP(pole) E[pos] Score Driver Team Track Form Actual Verdict
1Verstappen27.0%3.218.4318.3118.0420.5016.912±1
2Antonelli18.3%3.717.9113.3519.4118.0019.661POLE
3Norris14.1%4.117.5314.6319.4117.0018.144±0
4Russell11.9%4.317.3913.4319.4116.0019.555±1
5Piastri11.8%4.317.3515.0719.4116.2517.827±3
6Leclerc11.5%4.317.3015.7016.5019.0017.833±1
7Hamilton4.5%5.516.2915.8916.5016.5016.136±1
8Hadjar0.8%7.314.6212.2618.0413.2513.7622±15
9Gasly0.0%11.810.8811.128.6612.0012.259±3
10Sainz0.0%12.210.6013.557.4814.008.5813±1
11Hülkenberg0.0%13.29.9510.517.2411.5011.2110±3
12Lawson0.0%13.79.708.179.3110.0011.0911±3
13Colapinto0.0%13.99.566.018.6612.5010.558±6
14Ocon0.0%14.39.339.246.7812.009.8014±0
15Lindblad0.0%14.49.288.869.318.5010.3516±2
16Bearman0.0%15.58.638.706.789.509.9312±4
17Alonso0.0%15.98.4314.096.339.005.8617±1
18Bortoleto0.0%15.98.428.487.248.509.7321±5
19Albon0.0%16.38.1810.037.488.507.2215±1
20Stroll0.0%19.36.196.476.337.504.4918±1
21Pérez0.0%19.65.9010.761.6010.502.5520±0
22Bottas0.0%20.35.3313.331.606.002.7419±1
Verdict: predicted vs actual quali position. POLE = right pole sitter. ±N = off by N positions.
R5

Canadian Grand Prix

Upcoming · prediction only
# DriverP(pole) E[pos] Score Driver Team Track Form
1Antonelli22.5%3.418.1613.3519.4119.0019.66
2Verstappen20.8%3.618.0518.3118.0419.0016.91
3Russell19.8%3.618.0213.4319.4118.5019.55
4Piastri12.2%4.317.3515.0719.4116.2517.82
5Hamilton9.2%4.717.0415.8916.5019.5016.13
6Norris8.8%4.716.9714.6319.4114.7518.14
7Leclerc5.0%5.516.3015.7016.5015.0017.83
8Hadjar1.6%6.615.3012.2618.0416.0013.76
9Gasly0.0%12.410.5711.128.6610.7512.25
10Alonso0.0%13.010.1814.096.3316.005.86
11Sainz0.0%13.010.1613.557.4812.258.58
12Hülkenberg0.0%14.09.6410.517.2410.2511.21
13Colapinto0.0%13.99.626.018.6612.7510.55
14Lindblad0.0%14.19.598.869.319.7510.35
15Albon0.0%14.79.2410.037.4812.757.22
16Ocon0.0%14.79.219.246.7811.509.80
17Lawson0.0%15.38.838.179.316.5011.09
18Bearman0.0%16.08.448.706.788.759.93
19Bortoleto0.0%16.58.178.487.247.509.73
20Stroll0.0%18.86.696.476.339.504.49
21Bottas0.0%19.36.3313.331.6010.002.74
22Pérez0.0%20.84.9010.761.606.502.55
Verdict: predicted vs actual quali position. POLE = right pole sitter. ±N = off by N positions.

Source · Five Reds simple position-based quali model · Each race's prediction uses only data prior to that race's qualifying (leak-free)