Record

85.2% headline hit rate across 396,862 backfilled calls. This is a BACKFILL, not a live-tracked record — see the note below.

Measured odds. No trade calls. what this means →

The rule

At TF-bar close t: range_N = max(High,N)-min(Low,N) over last N TF bars. position = (Close_t - min(Low,N))/range_N. prior_move_atr = (Close_t - Close_(t-N)) / mean(High-Low,N). Extreme-high setup (position>=0.97, prior_move_atr>=+0.50) -> CALL UP (continuation). Extreme-low setup (position<=0.03, prior_move_atr<=-0.50) -> CALL DOWN. Winning lookback: N=40.

ASYMMETRIC RACE, v3. HIT = a later TF bar's High/Low TOUCHES entry +/- d_target*ATR_N(t). MISS = a later TF bar's CLOSE is beyond entry -/+ d_inv*ATR_N(t). HIT has precedence over MISS within the same bar. Neither within 40 periods -> EXPIRED (reported separately, excluded from the odds fit).

Bars compose on the same 18:00 ET session grid futures uses.

EXPIRED (neither target nor invalidation hit within the 40-period cap): 0.005% of calls (20) — excluded from the odds fit and from every tier's hit rate.

Live — calls published before their outcome

Live calibration, by tier

Every live call we published, graded against the probability it stated at publication. That probability is frozen when the call goes out and never recomputed. A call counts only if it was published before the bar that decided it closed; calls published after their outcome stay visible on their own pages but are not counted here.

Since 2026-09-28 (first counted call): 565 resolved calls.

Until 2026-09-29 15:16 UTC our live engine could build a longer-timeframe bar before it was complete, so some calls were opened on a price that was 15 to 30 minutes old. Those calls stay visible on their own pages, marked with a flag, but they are not counted in the numbers here. Every call published since the fix is counted as before.

Including calls from the period when bars were composed early (until 2026-09-29 15:16 UTC), kept for the record. Back to counted calls only

Actual = HIT / (HIT + MISS); expired calls are excluded, as in the backfill. The 95% interval is a Wilson score interval on the actual rate. Tiers with fewer than 30 resolved calls are marked not enough data yet.

Backfill — the rule replayed over history, not live calls

Stated vs. actual, by tier

Stated vs. actual, by year

Stated vs. actual, by market

Stated vs. actual, by timeframe

Payoff

What the published rule actually pays, per call, gross of costs: a hit realizes exactly the target (+0.05 ATR), a miss realizes the deciding close. High hit rate, small wins, larger losses — the expected value is slightly negative, and a call only breaks even above the win rate in the last column. Spread and commission make it worse.

Fixed-horizon side number

1-period close-vs-entry outcome, reported ALONGSIDE the v3 asymmetric-race status (target touch / invalidation close) -- 'hit is not whether it paid'. Independent field, not a substitute for the main HIT/MISS/EXPIRED status.

Download

Every backfilled call, unfiltered: record-backfill-fx.csv (25.2 MB, CSV)

All times are US Eastern (ET). "date" is when the call became known: the close of the trigger bar, whose closing price is the entry. "resolved" is the close of the bar that decided HIT or MISS.

By market, year by year

Stated vs. actual hit rate for tiers S, A and B, year by year for each market. This is a rules-based backfill, not the live record.