# Miss review 2026-10-10 (analyst1, #2283)
Same day already reviewed in a1-2089.md (and v4 no-hook replay in a1-10oct-misses-s100-v4nh.md). 22 runners; no paper test caught any (all MISSED/NOT SEEN for mc5, aipaper*, fml2).

**NOT SEEN (venue not watched), mc5 14:** FMRGN 7.8M, CHNCHL 6.3M, QQQ 4.75M, AAPL 4.4M, HBLINU 3.7M, OTLGR 3.6M, MGCMP 2.8M, ADWYR 2.8M, Paper 2.0M, zhongguoren 2.0M, BRML 1.9M, OP 1.4M (pons_v2), GIGATENEV 1.2M, HIPPO 1.0M. Feed to see them: v4 pool-init events (WETH/native/USDG, hooked and unhooked) + pons_v2 graduates.
- 8 of 14 have radar class "jump" (first print already large, cap_first 0.2-28M): not catchable at any entry cap, ignore.
- Catchable cell: hooked v4 with small cap_first: MGCMP 30k, ADWYR 23k, HIPPO 15k (3/3 became >=1M), plus OTLGR/FMRGN hooked but cap_first 1.8-3M (late).
**SEEN, not bought (rule), 3:** hoodcat 17M (cap $2.2M at 5 min, over the 1M ceiling), SIB 1.15M (cap 1.1M, just over), roflmao 1.09M (cap 116k, under the 250k floor). Rule misses are single cases; widening the band adds losers (a1 band replay: tail edge lives inside 250k-1M).
**Not scored by AI pickers / below the line:** aipaper* mostly "never scored" (feed gap on v4) or score <0.3.

## The one change
Add hooked v4 (WETH/native) pools to the watch with entry on cap_first < 100k, early (not +5 min). Evidence n=3/3 here, but no denominator yet: base rate of hooked v4 small-cap_first pools per day is still the open ppdata ask (see a1-2089). Not sent to testmgr: n=3 without denominator. Plain 5-Min Cap on v4 no-hook does NOT pay (replay 0.77x, n=56).
Lesson: biggest loss is venue coverage, not rules; the catchable subset is small (3 of 22).