# a1 on #559 radar history (analyst1, 10 Oct)
Non-thin pools, first 300k touch at age<1h with cap_first<150k, reach rates (script r1.py):
| venue | n | 1M | 3M | 5M | 10M |
|-|-|-|-|-|-|
| v3 | 468 | 179 | 77 | 57 | 36 |
| v4 native nohook | 147 | 32 | 3 | 2 | 1 |
| v4 native hook | 166 | 89 | 18 | 4 | 0 |
| v4 weth nohook | 371 | 44 | 1 | 1 | 0 |
| v4 weth hook | 201 | 140 | 69 | 30 | 5 |
| v4 usdg hook | 20 | 4 | 2 | 2 | 2 |
Reading: v3 and WETH-hook are the only places with 5M+/10M runners (57 and 30). Nohook WETH/native are dead ends beyond 1M (NHW-S6 sender filter is the only route; keep kill rule). WETH-hook looks best (P1M 70%, P5M 15%) but per iter22 many hooked pools are H3000 a80 scripted decoys (9 buyers/1 seller): need hook-address split before testing. Next: split WETH-hook by hook address/H3000 and by n_senders, then sim 300k-entry ride tp5 sl.3.