# FW Analyst 1 to Test Manager: $5 live micro-probe on fresh v3 launches (8 Oct 2026, 20:00 Bangkok)

Needs Ben's approval (real money, small). Purpose: settle the one question no paper replay can answer.

## Why
On paper, buying EVERY v3 WETH launch 30 s after its first swap (sell all at 20x, stop 0.3x, 4 h) makes $2,222/day at $50 (2,072 trades, 13 Sep - 8 Oct, tick-walk fills), four times any picked book. But 1,225 of those trades are fleet coins with 0-2 outside buyers and 88% get pulled. On lane1 the paper book looked great and real outside buyers lost 73-83%, because the fleet reacts to a real buy. FW A2's A2-V3P204 (+$1,215/day on paper) is the same kind of coin. A real buy is the only way to know.

## Probe
- Universe: every new Uniswap v3 WETH pool, fee 100, on Robinhood Chain.
- Entry: 30 s after first swap, $5 per coin, GMGN, same slippage rules as fmmc5. Skip if the pool holds < 0.1 WETH.
- Exit: sell all at 20x, stop at 0.3x (live value, as fmmc5), pull exit, 4 h time stop.
- Size of the test: first 100 coins (about 1-2 days at ~80 launches/day), max $500 at risk in total, max 20 open at once.
- Record per trade: the paper model's x for the same coin (harness replay the next day) next to the live x.

## What decides it
- **Pass (fleet coins are tradeable):** live mean x within 25% of the paper model on the same coins AND live pull rate not more than 10 points above paper. Then the dumb book and A2-V3P204 go to a proper paper/live ramp.
- **Fail (fleet reacts to us):** live mean x below 1.0 or pulls within 10 s of our buy on more than 20% of trades. Then all fleet-coin books are killed for good and we stay with real-outside-money books (live rule, A1-KITLIVE).
- Either way the cost is at most $500 and the answer is worth more than any paper study.