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 size8488
  • Correct1036
  • Mean stated confidence 32.0%
  • Confidence gap 19.8 pts
  • Symbols covered 1050 / 1064
  • Coverage of the field 98.7%
  • Horizons used1
Calibration — stated confidence against measured accuracy
Confidence n Accuracy
0–50% 8135 11.4%
50–70% 191 39.3%
70–85% 42 14.3%
85–100% 120 22.5%

Mistral Small Latest

mistral-small-latest
36.8%
95% CI 34.1% – 39.7%
  • Sample size1110
  • Correct409
  • Mean stated confidence 74.6%
  • Confidence gap 37.8 pts
  • Symbols covered 143 / 1064
  • Coverage of the field 13.4%
  • Horizons used2
Calibration — stated confidence against measured accuracy
Confidence n Accuracy
0–50% 21 52.4%
50–70% 392 45.4%
70–85% 249 38.6%
85–100% 448 27.7%

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