gap intelligence: model #19 vs market (consensus with #17)

← lab · 21,280 model-vs-market gaps dissected · survival AUC 0.6787 · ran 7/6/2026

What this is: Dissects every disagreement between our models and the market: who was right in which situations, whether two engines agreeing means anything, and whether a classifier can predict when OUR side of the gap wins.

Every model-vs-market disagreement dissected: who was right, where, and can we predict it? model advantage = market log-loss minus model log-loss in the bucket (positive = our number was closer to reality there).

Gap size	Direction	n	Model advantage	Reality	Model said	Market said	Stable
large	model_lower	1,616	-0.04316	0.4666	0.3368	0.4679	⏳ yes
large	model_higher	1,833	-0.04063	0.3944	0.5228	0.3884	⏳ yes
medium	model_lower	3,404	-0.01534	0.386	0.3134	0.3848	⏳ yes
medium	model_higher	2,922	-0.01351	0.3765	0.4506	0.3788	⏳ yes
small	model_lower	3,405	-0.00381	0.3471	0.3185	0.3528	⏳ yes
small	model_higher	2,911	-0.00195	0.3748	0.4001	0.3661	⏳ yes
tiny	model_higher	2,399	-0.001	0.3443	0.3633	0.3535	⏳ yes
tiny	model_lower	2,460	-0.00038	0.3394	0.3273	0.3372	⏳ yes

Gap-survival classifier holdout AUC: 0.6787 · top features: rel_gap, price_open, prob_market, gap

Hurricane test over 4,560 shared predictions — Brier: engine A 0.19489, engine B 0.19611, engine mean 0.19337, market 0.19016. Averaging wrong models beats both members, but the market still beats the ensemble.

Reading the columnswhat each number actually means

Model advantage	market log-loss minus model log-loss in the bucket; positive = our probability was closer to what actually happened
Gap size	\|model p − market p\|: tiny <2pp, small <5pp, medium <10pp, large <20pp, huge beyond
Survival AUC	can gap features predict when our side of the gap is right? >0.5 = learnable
Hurricane test	are individually-wrong engines nearly-right on average? Brier: lower = better

Spec · the reproducible recipe

{
  "kind": "gap_intel",
  "model_run": 19,
  "second_run": 17
}