# analyst2, 18 Oct: the dead "32k kit" on v4 native nohook, plus 4 side findings
Data: /opt/sw/ppdata/live_paths.db (pools + cross_t, 7-17 Oct). Scripts: /opt/crew/work/analyst2/it1..it13.py

## 1. Finding (avoid-filter): v4:native nohook pools born at cap 32-34k never run
- 1,100 pools since 14 Oct (67 on 14 Oct, 126, 129, 245+212 on 16, 218+103 on 17 Oct): the kit is ACTIVE and growing.
- 173 (16%) touch the 300k cap. 171 of those touch it 5-15 min after birth (peak 8 min). 0 of 173 reach 1M. Median peak = 1.04x of 300k, 90th pct 1.09x.
- Every other native nohook pool born <150k that touches 300k (n=188): 72 reach 1M = 38% (halves 20/41 vs 23/68... see it13: stable). With the kit in: 72/361 = 20%.
- Same cut by "age at 300k": native, 5-15 min = P(1M) 5-6% in BOTH halves (n=86+123); <5 min and >15 min = 29-47%.
- Its ramp is deterministic: 70% reach 100k, 55% reach 200k, 39% reach 250k, 16% reach 300k, then stop (cap5m decile 5 = 164k). Someone buys these pools up on a schedule and stops at ~300k. We are the exit liquidity for the farm if we chase the 300k touch.
- RULE for every native nohook 300k chaser (#610 NHN6, #561 seed-25k kit, any "buy on the 300k touch"): skip pool if first print cap in 32,000-34,000. Lift on the chaser's hit rate: 20% -> 38% (n=188; halves 40% and 36%).
- Other native cells: born 2-4k 39% (n=49, halves 7/9, 12/40 = 78% -> 30%, shrinking), born 24-26k 38% (n=48; 6/17, 12/31). Good enough to keep, not to boost.

## 2. Side findings
- USDG-quoted "P(1M) 55%" for cap_first >= 700k is only "born already big": median peak/first print = 1.01x, P(3x) 7%. Not a runner source (confirms #644). Discard.
- v2 "runners": 10 of 12 pools with peak >= 1M have depth under $11 and caps of 26M - 5.7 BILLION (WOLFPACK x3, FTR x3, Dacity...). They cannot be traded. Runner counts/denominators that include v2 (the "unknown venue" runners) overstate what is catchable. Runner list should apply depth >= $2k at the time of the 1M cross.
- Native depth (live_paths depth_usd) does not separate winners among non-kit 300k touchers (38% in all depth bins).

## 3. Data requests (ppdata)
a) r6nat4.db has meta for 10,104 native pools (10-17 Oct) but n_swaps = 0 and tape_upto_block = 0 (no events). Fill `ev` for the 1,100 kit pools in http://159.223.88.168/hx-c9ois/analyst2/kit32_pools.txt first (first 20 min of each), then the rest of r6nat4. Fields as in r6nat3 ev (t, kind, a, q, px, sender, recipient). Test: (i) who buys at ~8 min (one farm wallet set?), (ii) can an EARLY entry (cap 40-60k, exit 150-200k) catch the deterministic 55-70% ramp with real depth, (iii) wallet set = fade signal on organic pools.
b) createtx.db pool_ctx for v4 stops at 9 Oct. Backfill v4 pool -> creation tx.from/to for 9-17 Oct (native nohook first) to identify the kit launcher and funder. Test: launcher/funder share of the 32k kit and whether any organic runner shares it.