# A2-LV forward paper test (Test Manager) - 12 Oct 23:20 UTC
Hunt: entry +60 s after v3 creation (cap ~700k), create-tx value >= 12.5 ETH from a first-time sender, target 3M-10M+ (A2 #638, fw/a2-iter39.md).
Arms: A = first-time sender, tp10 | B = same, tp20 | C = baseline, any sender, tp10. Stop 0.5, 4 h, 7% slippage, pull=0, skip if est. pool ETH < 8.
- Replay since 6 Oct (create values fetched so far; backlog fills ~200 pools per run): 69 entries
  - A: n 62 closed, mean 1.38x, median 0.45x, win 8, tp 6, stop 54, pull 0, time 2
  - B: n 62 closed, mean 2.06x, median 0.45x, win 8, tp 4, stop 54, pull 0, time 4
  - C: n 69 closed, mean 1.32x, median 0.45x, win 9, tp 6, stop 60, pull 0, time 3
- FORWARD only (pools created from 10 Oct 00:00 UTC): 35 entries
  - A: n 30 closed, mean 0.91x, median 0.45x, win 3, tp 1, stop 27, pull 0, time 2
  - B: n 30 closed, mean 0.99x, median 0.45x, win 3, tp 0, stop 27, pull 0, time 3
  - C: n 35 closed, mean 0.91x, median 0.45x, win 4, tp 1, stop 31, pull 0, time 3
- Skipped: {'rpc fail': 1, 'value<12.5': 1837, 'no tx value yet': 2}; create values cached: 1906 of 1908 pools
- Pass (A2): A mean >= 1.3 over 30 closed; kill < 1.0. About 5.7 qualifying pools a day. A2 replay: tp10 1.37, tp20 1.81 (fat tail, first half below 1: not proven).
- Limits: first-time sender checked against kitb.db (8 days) + cache only; mark-to-print exits; pool ETH estimated as create value + net buys. Only Ben switches live (D-030).