# Be where the runners are (Ben, 10 Oct 2026)

Ben: "If all the runners are in pools we are not even watching, we need to put a lot of focus there."

## What the nightly review shows (7 to 9 Oct, 3 days)
- **66 runners ($1M+) were missed by every bot.**
- **By pool type:**

| Pool type | Runners missed |
|---|---|
| Unknown pool type ("?"), not in any of our feeds | 23 |
| v4 without hook, WETH | 14 |
| v4 without hook, USDG | 4 |
| v4 without hook, ETH | 2 |
| Pons v2 | 3 |
| Uniswap v3 | 3 |
| v3 without hook, WETH | 3 |
| v3 by launcher (4 launchers) | 4 |
| Bankr | 1 |

- **The live 5-Min Cap rule only watches v3 mc5 pools.** 45 of its misses were "not a pool type it watches".
- **Our paper tests mostly sit on hooked v4 pools.** They are not where most runners came from in these 3 days.

## Work
1. **Runner map, 30 days (Analyst 2, first).** For every $1M+ runner since 13 Sep, record:
   - pool type and venue, launchpad or launcher kit, quote token
   - time from launch to $1M, and the first-minute pattern

   Then show the share of all runners per pool type next to the share of our trading and testing per pool type (live, every paper test). The table shows where we are blind. Update it weekly.
2. **Unknown pool type (PP Feeds + Data Engineer, now).** 23 runners came from pools our feeds do not label at all. Find out what they are (factory, launchpad, router) and add them to the feeds. A runner we cannot see, we cannot trade.
3. **Shift focus (both analysts, from the next thinking run).** At least half of every iteration run goes to the pool types with the most runners that we do not watch. Right now that is v4 without hook (WETH/USDG/ETH) and the unknown types. For each: what does a runner look like in its first minutes there, and can a version of 5-Min Cap or an S2 rule catch it?
4. **Live rule extension (Analyst 1 + Test Manager).** Replay the 5-Min Cap rule on v4 no-hook WETH pools over 30 days. If it holds after fees and depth checks, it becomes a paper test the same day, and later a live candidate for Ben.
5. **Keep score.** The morning update gets one line: "Runners yesterday: N · in pools we watch: M · caught: K".