Compare AI models

Two models, one window, one asset class, one horizon. Accuracy is shown with a 95% confidence interval, because a rate without one cannot be compared to another rate.

Rule-Based Analysis

rule-based-analysis
12.2%
95% CI 11.5% – 12.9%
  • Sample size8267
  • Correct1008
  • Mean stated confidence 32.1%
  • Confidence gap 19.9 pts
  • Symbols covered 968 / 978
  • Coverage of the field 99.0%
  • Horizons used1
Calibration — stated confidence against measured accuracy
Confidence n Accuracy
0–50% 7916 11.4%
50–70% 190 38.9%
70–85% 42 14.3%
85–100% 119 21.8%

Mistral Small Latest

mistral-small-latest
34.8%
95% CI 31.9% – 37.7%
  • Sample size1010
  • Correct351
  • Mean stated confidence 74.6%
  • Confidence gap 39.8 pts
  • Symbols covered 130 / 978
  • Coverage of the field 13.3%
  • Horizons used2
Calibration — stated confidence against measured accuracy
Confidence n Accuracy
0–50% 19 52.6%
50–70% 358 43.9%
70–85% 231 36.8%
85–100% 402 24.6%

Over the last 30 days the two confidence intervals do not overlap, so this sample does separate the models.

Accuracy counts a prediction correct when the direction it stated matches the direction the price moved past a 1% threshold over the stated horizon. Coverage is the share of symbols scored in this window that the model expressed an opinion on — a high rate over four symbols is not the same claim as the same rate over ninety. Full method and disclosures