# A2-V3P204 outside money test: FULL result, all 499 pools (8 Oct 2026, 23:05 Bangkok)
Verdict: PASSES the outside money test (kill condition not met). Ready for paper test (stage 4), with one caveat on the 5x exit.

Data: PP R1 signer table /opt/sw/ppdata/v3from.db, 499 of 499 pools, coverage 100%. Scripts on research-box: /root/fwa2/edge/outside204.py (raw: outside204_p493.out) and outside204_kill.py (raw: outside204_kill.out).

## Kill condition: can a non-fleet wallet really sell high in the pools where our rule hit 5x?
- Our rule (bt.py, +66 s, 5x TP, stop 0.5, 4 h): n=317, mean 2.87x. 40 trades (13%) hit 5x.
- In those 40 pools: 289 non-fleet positions; sells at >=2x: 65, >=3x: 14 (in 6 of the 40 pools), >=5x: 0.
- So sells at 3x+ do fill (not a honeypot). A 5x sell by a non-fleet wallet was never seen.

## The 5x exit does not carry the edge
Mean if every 5x exit only filled at 1x / 2x / 3x: 2.36x / 2.49x / 2.61x. Dumb baseline: 1.22x / 1.26x / 1.76x. Worst case still beats it.

## Who loses (outside money)
- Truly outside wallets: 668 positions, $65.8k in, 98% never sold and got 0. Those in our window (30-120 s): n=13, all got 0. They hold through the pull; our rule exits on time.
- Busy (multi-pool, non-fleet) wallets: 1,552 positions, 0.82x; 182 sold at 2x+, 18 at 3x+.
- Token code scan: 0 risky of 496 tokens.

## Caveat for the paper test
5x fills are unconfirmed by any real wallet. Paper test should run a 3x-take-profit arm next to the 5x arm, and record real fill prices on every 5x exit.
North star (D-036): these are pump-and-pull coins, NOT runners. This earns money but does not raise runner capture.