Skip to main content

Track record

Accuracy

How the forecasts have actually scored against real results — measured across every scored class-round, with nothing left out. Endurance grids are large, so a car landing within a few places of its prediction is a strong result.

Evidence

11 Aug 2026, 07:01 UTC

Over 4 paired rounds, the model too few rounds to say (last race order).

BaselineRaceRoundsModelBaselineGainVerdict
Last race orderHYPERCAR43.534.02+0.49too few rounds to say
Last race orderLMGT345.215.04-0.17too few rounds to say
Last race orderLMP215.112.11-3.00too few rounds to say
Season form orderHYPERCAR43.533.57+0.04too few rounds to say
Season form orderLMGT345.214.45-0.76too few rounds to say
Season form orderLMP215.114.11-1.00too few rounds to say

Paired bootstrap on the round-by-round difference: 0.494, 95% CI [-0.159, 1.147] in positions gained. Positive means the model is closer to the real finishing order than last race order is; the interval covers zero, so no difference has been demonstrated.

forward evaluation — each round was forecast before it ran and scored afterwards; not a backtest. Historical replays and the forward record are computed separately and never merged.

  • Only 4 round(s) have been scored. Nothing here is a claim yet.
  • Metrics are only comparable within this series. A position error over this field size means nothing next to another series' number.
Class-rounds scored
9
predictions graded
Mean position error
5.08
places off, on average
Podium hit rate
22%
predicted podium correct
Winner hit rate
0%
predicted winner correct

By class

Hypercar4R

Mean err.

4.21

Podium

33%

Winner

0%

LMGT34R

Mean err.

5.87

Podium

17%

Winner

0%

LMP21R

Mean err.

5.33

Podium

0%

Winner

0%

Round by round

RoundCarsMean errorPodiumExactWithin 3Winner
R1Imola174.352/33/177/17
R2Spa Francorchamps134.770/31/135/13
R3Le Mans143.501/31/147/14
R4Sao Paulo174.241/31/1710/17

Does the forecast beat guesswork?

Combined win + podium forecast error across 9 scored class-rounds — lower is better — measured against two naive baselines: simply repeating the last race’s order, and ranking cars by their season points.

Model forecast error0.209lower is better

Last-race form

baseline error 0.221

0.013Beats it

Season form

baseline error 0.207

+0.001Matches it

“Matches it” means the model is statistically no worse than that baseline over the rounds scored so far. On a young season with few rounds, matching a strong season-form baseline while beating last-race form is an honest, expected result.

Calibration

Probabilities are calibrated on 4 real rounds of results, per class. Calibrated on real WEC results, per class stratum.

Continuous evaluation

Every round is re-scored against simple baselines as results come in, so the track record above updates honestly over the season — no cherry-picking, nothing dropped.