Next Race --
Model Accuracy --
Season 2026
Races Analyzed --

Optimal Team Pick

Estimated Points --
Budget Used --
Remaining --

Drivers

Constructors

Top 5 Configurations

Premium Builds (>$95M)

Balanced Builds (2 Stars + 3 Budget)

Driver Rankings

Rank Driver Team Gap Price Est Pts Pts/$M Tier

Constructor Rankings

Constructor Driver 1 Driver 2 Combined Pts Price Pts/$M

Driver Bias Correction

How each driver systematically over/underperforms their practice pace. Positive = finishes better than practice suggests.

Driver Avg Bias Races Trend

2025 Season Backtest

How the model would have performed picking teams for every 2025 race using only practice data available before qualifying.

-- Capture Rate
-- Points Earned
-- Best Possible
-- Races
Race Picked Pts Best Pts Capture Team

Calibration Curve

Average actual fantasy points earned per predicted finishing position, learned from 2025 data.

How It Works

01

Data Collection

Practice session laptimes are pulled from TracingInsights for FP1, FP2, and FP3. Both single-lap pace and long-run degradation are captured.

02

Session Weighting

A grid search over the full 2025 season finds optimal weights for each practice metric. FP3 best lap (40%) is the strongest predictor, followed by FP2 best lap (25%).

FP3 Best Lap
40%
FP2 Best Lap
25%
FP3 Long Run
10%
FP1 Best Lap
5%
FP2 Long Run
5%
03

Calibrated Point Estimation

Instead of a formula, the model uses empirical calibration from 2025. For each predicted position, it looks up the average actual fantasy points earned historically, then applies per-driver bias correction.

04

Budget-Constrained Optimization

Brute-force search over all valid team combinations: 5 drivers + 2 constructors within $100M. Evaluates ~1.4M combinations, pruned by cost constraints, to find the maximum expected point total.

2026 Fantasy Scoring

Drivers

Race P1-P1025, 18, 15, 12, 10, 8, 6, 4, 2, 1
Quali P1-P1010, 9, 8, 7, 6, 5, 4, 3, 2, 1
Position gained+1 per position
Overtake+1 per overtake
Fastest lap+10
Driver of the Day+10
DNF / NC-20
Sprint P1-P88, 7, 6, 5, 4, 3, 2, 1

Constructors

Both drivers Q3+10
Both drivers Q2+5
One driver Q3+3
Pit stop < 2.2s+10
Pit stop 2.2-2.49s+5
Pit stop 2.5-2.99s+2
Fastest pit stop+5 bonus
Record pit stop (<1.8s)+15 bonus

Team Management

Budget$100M
Roster5 drivers + 2 constructors
Free transfers2 per weekend
Extra transfer-10 pts each
Boost chipDoubles 1 driver

Known Limitations

  • Wet race weekends where practice is dry break the model's predictions
  • Multi-car incidents and penalties are unpredictable from practice data
  • No qualifying data — quali positions are estimated from practice pace
  • Sprint weekends with reduced practice sessions change scoring dynamics
  • British GP-type chaos races achieve only ~55% capture rate