Model Validation
Backtesting Results
How well does our model perform? We backtest against every UK general election since 2015, using only data available at the time. No hindsight, no cheating.
100%
Largest Party
Elections Tested
4
2015, 2017, 2019, 2024
Avg Seat MAE
20.5
Mean absolute error per party
Largest Party
4/4
Correctly identified
Avg Brier Score
0.13
Lower is better (0 = perfect)
Majority Correct
Majority Missed
Our Prediction
Actual Result
Seat MAE
Vote Share MAE
Brier Score
In 80% CI
Seat Error by Party
Methodology
Our backtesting uses a strict out-of-sample approach: we run our current model on historical data, using only information that was available at the time. House effects are re-estimated for each election using prior data only.
Important caveats:
- The 2024 boundary changes mean pre-2024 elections use notional results
- Reform UK's current surge is unprecedented - no historical analog exists
- Model parameters may be slightly overfit to recent elections
- Past performance does not guarantee future accuracy