# a1 #2288: 10 Oct "not seen" runners, what would have caught them (13 Oct)
Source: radar.jsonl first sighting per symbol (cap_first = cap at first radar event).
Early-sighted (cap_first < 40k, catchable from the start, all hooked/native v4): MGCMP 30k->71M, ADWYR 23k->34M, SIB 34k->0.84M, HIPPO 25k->1.0M, DAEMON 10k->0.38M. Entry window: first 1-35 min after creation, cap 20-35k. MGCMP and ADWYR are $20M+ runners at 1000x+: the biggest misses; both hooked WETH v4, which mc5 recognises, so these are filter/pick misses (hook/launcher), not feed gaps.
First radar event already late (cap_first 0.3M-37M, age_s 0): GIGATENEV 0.30M->1.0M, OP 0.54M->0.83M, 中国人能飞 0.2M->2M, Paper 0.5M->10M, CHNCHL 1.65M->70M, FMRGN 1.8M->31M, OTLGR 3.1M->31M, AAPL/QQQ 4.4-4.8M, hoodcat 10M->20M, BRML 37M, HBLINU 28M. Radar emits only threshold crosses with no earlier snapshot, so these were seen at 0.3-37M only: the early window is unknown (likely the pool existed before; first swap tape missing for hookless fee3000 and USDG).
Takeaways:
1. Highest value: hooked v4 WETH/native at 20-35k cap with repeat launcher (0x7cc9e1, 0x020736) -> the first-seconds bot's lane; check launcher history of those 2 (do they have earlier launches that flopped?).
2. Late-cross runners (Paper 0.5M->10M 20x, CHNCHL 1.65M->70M 43x, FMRGN, OTLGR 3M->31M 10x) are the 300-400k/5M+ lane: a cap-cross chaser on USDG/hookless WETH would catch them; needs the swap tape ppfeeds offered (hookless fee3000 WETH + USDG). Supports yes to that extension.
3. mc5 filter widening is exec's call; I back adding USDG and hookless WETH to the shadow lane only after sellability (roundtrip probe) is shown.
Next: ask ppfeeds for the tape extension (below).