The brain is graded on data it never trained on (walk-forward + random split). The honest question isn't just "beats a coin flip" — it's "beats the market's natural upward drift." That gap is the real timing skill, and right now it's small and not yet statistically significant.
Walk-forward accuracy
—
market drift (base rate) —
timing skill vs drift —
sample (n)—
Held-out accuracy
—
Brier—
calibration (ECE)—
p-value—
Each trading day we log every directional call + its entry price, then grade it against the real 5-trading-day move. No cherry-picking, no hindsight. This is the receipt that decides whether paid options-flow data is worth it — it fills out over the coming weeks.
Verdict: loading…
A separate, isolated forward-test (pass 292): each weekday we capture the daily call/put $-flow direction per liquid name from the free CBOE chain, grade the 5-day move, and test it against the market-drift base rate at 95%.
See the live flow board →Loading…
How to read this. Backtest = how the brain scored on history it never saw, split into the market's natural drift vs real timing skill (only skill above the base rate counts as edge). Forward test = the live, no-excuses receipt: did today's calls actually work? "Beats coin flip" needs a one-sided z > 1.64 (95% confidence) — which needs enough graded calls, so early on it correctly says "too few." A real edge here is small by nature (markets are efficient): even 53–55% directional with positive expectancy only counts as edge if it clears the base rate (drift), not just a coin flip. Today the backtest is mostly drift, so the live confluence test below is what we're watching: if the confluence leg (brain + insiders agreeing) beats the brain-only leg over time, that's the proof that fusing more signals — including paid options flow — adds real edge.