# Miss review 2026-10-17 (analyst1)
Source review/misses.csv, born 17 Oct, deduped by token+venue, real runners only (ath 1M..1B; ath >1B = phantom: SUSD 990B, 0x1cc4a7 2.4T).

| group | n | status | cause | would catch |
|---|---|---|---|---|
| v3 kit:B (organic/small-born, cap_first 3k-10k) | ~13 | mc5 MISSED: "cap over $1M at 5 min" x6 (2-2.4M at +5m), "under $250k" x5, "in band but skip" x3; aipaper* never scored (NOT SEEN) | entry rule: born small, band checks at +5min miss both ends | entry on first cross of a low cap (<=30k) with launcher kit:B tag; same as ledger KITB-ORG (16 Oct) |
| v3 kit:other / pp:v3/launcher | ~9 | same pattern; DOPE/DOPESTER cap 1.9-2.0M at +5m | too late at +5m | same lens, earlier entry |
| v4:native unhooked (ELON x2, MUT8 x2, EMPIREX x2, FTR, COVENANT, PRIORS, WTZ, coms, 1) | ~12 | mc5 NOT SEEN (venue pool_uniswap_v4 / pons_v2 not watched); aipaper scored low | feed gap + repeated symbols (serial relaunch) | SERIAL-SYM (16 Oct) |
| USDG hook pp:v4 (PONS 305M x2, CASHCAT, AI, ORBIO, HOOKR, NVDA, STANDARD) | 9 | NOT SEEN | USDG-quoted pools; ledger #341/#145 show these are born-high/pinned artifacts, 305M PONS equals the 7 Oct PONS phantom | none; do not chase |
| Paper v4:weth hook fee8388608 10M, Tour RobinHood 1.6M | 2 | NOT SEEN | unwatched hook venues | low count |

Technical failures: 0 (all rule/feed). Rate limits: none visible. Paper test that caught any: none (only fml2 shows MISSED on 1).

## One change
The biggest catchable block is v3 kit:B/kit:other (~22 runners, 1-2.4M at +5min): the +5min cap band is the cause (6 are 1.4-2.6M by then, 5 are still <250k). Change: for tag kit:B (organic) add an arm entering at first cross of 30k-150k cap instead of waiting for +5min. Caution: ledger #651/#654 found kit ramp pools are pulled 99.8% (median 437s) so the replay MUST include pull/outside-sell sim and the 30-day loser rate before paper.
Second: SERIAL-SYM for v4:native (12 runners, 6 are repeated symbols).