Calibration scoreboard
A prediction is well-calibrated when its stated probabilities match observed frequencies. Lower Brier and lower expected calibration error (ECE) are better. Measured on a pooled walk-forward holdout, 72,268 predictions across 24 investors. Data as of Mar 2026.
Base rate (new buys)
5.51%
Brier (raw → calibrated)
0.0492 → 0.0473
ECE (raw → calibrated)
0.0204 → 0.0011
Brier improvement
4.0%
Method: isotonic (pooled walk-forward holdout). Calibration cuts the expected calibration error from 0.0204 to 0.0011, but the underlying signal is weak, new-buy events are rare (5.51% base rate), so a small Brier improvement is all that is honestly on the table.
Leak-free walk-forward backtest: top-10 predicted new buys ranked from a broad ~2,000-name universe known at the rebalance date, equal-weight, 1-quarter holds, complete quarters only, gross (before costs). Small samples, per-investor edges are noisy and not statistically significant.