# hk50: v4 native other-hook (non-fcd1), buy at first 50k cap touch (first cap<=60k) (#639) paper test (Test Manager) - 22 Oct 19:32 UTC
Spec: ledger #639. Rows with src=replay are tape replay before registration (18 Oct 06:00 BKK); only src=fwd counts for the forward n>=15 check. Hunt: entry at 50-60k cap, target the 1M tier (D arm tp3.3 is primary; tp5+ gains = jump artefacts to re-check). Stop 0.5x, 6 h, fee + 3% slip each side, pull and quiet = 0. Duplicate: hkq40 hits the same pools.
- Other-hook native pools in tape: 419; with supply 278; waiting 227; open (arm legs) 0; skipped {'first cap >60k': 13, 'jump': 21}
- arm A (tp5): on time n 86, mean 2.90x, <1x 33, tp 53, pull/quiet 30 | late (excluded) n 72, mean 2.76x, <1x 30, tp 42, pull/quiet 25
- arm B (tp3): on time n 86, mean 2.56x, <1x 8, tp 78, pull/quiet 5 | late (excluded) n 72, mean 2.25x, <1x 15, tp 57, pull/quiet 10
- arm C (tp10): on time n 86, mean 0.35x, <1x 83, tp 3, pull/quiet 79 | late (excluded) n 72, mean 1.34x, <1x 62, tp 10, pull/quiet 56
- arm D (tp3): on time n 86, mean 2.78x, <1x 9, tp 77, pull/quiet 6 | late (excluded) n 72, mean 2.47x, <1x 15, tp 57, pull/quiet 10
- Kill (A1): arm A mean < 1.2 over >= 30 trigger pools. Sim: 2.25 (tp5), 1.86 (tp3), 2.49 (tp10). Caveat: sell side for outsiders unproven; cap uses a stale ETHUSD.
