Research
Does the maths work?
Every signal, correlated against what price actually did next. Hit rate, sample size, and whether it beats a coin flip.
0 of 6 signals have been tested
Smart-money positioning
—
Stage 1 — sets direction
NOT TESTED
no measurement recorded · no sample
horizon not recorded
LIVE · scoring
Stage 2 — confirm
NO READING
no measurement recorded · 266 occurrences
horizon not recorded
LIVE · scoring
Dealer delta exposure (DEX)
—
Research
NO READING
no measurement recorded · 121 occurrences
horizon not recorded
SHADOW · not scoring
Institutional write reversal
—
Stage 2 — confirm
NOT TESTED
no measurement recorded · no sample
horizon not recorded
SHADOW · not scoring
Volume confirmation (price/OI)
—
Stage 2 — 4th trap component
NOT TESTED
no measurement recorded · no sample
horizon not recorded
SHADOW · not scoring
Entry path
NOT TESTED
no measurement recorded · no sample
horizon not recorded
DISABLED · failed its test
What these are correlated to: forward return — what price did
over a fixed horizon after the signal fired — measured on real
captured sessions, never simulated. A hit rate is meaningless without its
horizon and sample, so both are shown on every row.
Why a signal can be right and still not score: SHADOW means it is
measured but deliberately kept out of the arithmetic until the evidence
earns it. Trader 7 fade mode is the precedent — built, tested over
23 real sessions, held 30% of the time, and switched off on that
evidence rather than on an opinion.
A sample under 20 is flagged thin however good the hit
rate looks. “NOT TESTED” is a finding, not a blank.