# A2.34 Let the kit runners run longer (Analyst 2)

7 Oct 2026, 09:57 Bangkok (server clock). Pre-registered 09:55 in the status page, before any result.

## Answer
**KILL. Keep the live kit hold as is (48 h cap, 50% trail).**

- **The 48 h cap never sold a coin.** It sold 0 of 15 kit coins. The 50% trail sold 13, and 2 were still open when the data ended.
- **So 4 d, 7 d and no cap give the same result as live** at the 50% trail: +$0 a day in both halves. They cannot pass.
- **Wider trails lose in both halves, under every cap.**
  - 60% trail: A -$444, B -$739 to -$2,178 a day.
  - 70% trail: A -$1,050 to -$1,130, B -$1,516 to -$2,955 a day.
- **Runners still climbing when the 48 h cap sold them: 0.**
  - The longest hold sold on the trail 43.3 h after entry: 0xf15aab29c3d22cf2b469d4a9b0c0a72f31ba0711, at 11.69x. That is 4.7 h before the cap.
  - Median hold after 5x is 2.8 h. The longest is 37.2 h.

## Setup
- **What stays live:**
  - entry at gate + 30 s;
  - 0.3x stop and 5x trigger;
  - non-kit coins sell at 5x;
  - slot rule: 3 normal + 2 holds;
  - $1k ticket;
  - exits at the end of the next block, real depth, 1% fee.
- **What changes:** only the kit hold leg.
- **Cells:**
  - time cap from entry: 48 h (live), 4 d, 7 d, none;
  - trail: 50% (sell at 0.5 x peak), 60% (0.4 x), 70% (0.3 x);
  - the pull exit is kept in every cell.
- **Sample:** the 15 kit A/B coins among the 70 wide trades that hit 5x.
  - Half A has 9 coins, half B has 6.
  - The archive ends 6 Oct 10:58 Bangkok.
- **Random check:** 200 draws. Same coins, with the hold exit at a uniform random time between the 5x block and that cell's horizon.
- **Live book on these halves:** A +$2,870 a day (worst -$2,000), B +$9,631 a day (worst -$3,161).

## Cells ($ a day vs live, $1k)
| cap | trail | A | B | worst A | worst B | beats random A / B | without top 1 (A / B) | without top 2 (A / B) | verdict |
|---|---|---|---|---|---|---|---|---|---|
| 48 h (live) | 50% | 0 | 0 | -2,000 | -3,161 | 84% / 44% | 0 / 0 | 0 / 0 | live |
| 4 d | 50% | 0 | 0 | -2,000 | -3,161 | 88% / 72% | 0 / 0 | 0 / 0 | same as live |
| 7 d | 50% | 0 | 0 | -2,000 | -3,161 | 95% / 83% | 0 / 0 | 0 / 0 | same as live |
| none | 50% | 0 | 0 | -2,000 | -3,161 | 94% / 79% | 0 / 0 | 0 / 0 | same as live |
| 48 h | 60% | -444 | -739 | -2,000 | -3,161 | 61% / 30% | -444 / -1,112 | -444 / -1,112 | FAIL |
| 4 d | 60% | -444 | -2,178 | -2,000 | -3,161 | 70% / 43% | -444 / -2,178 | -444 / -2,178 | FAIL |
| 7 d | 60% | -444 | -2,178 | -2,000 | -3,161 | 83% / 47% | -444 / -2,178 | -444 / -2,178 | FAIL |
| none | 60% | -444 | -2,178 | -2,000 | -3,161 | 80% / 49% | -444 / -2,178 | -444 / -2,178 | FAIL |
| 48 h | 70% | -1,050 | -1,516 | -2,000 | -3,161 | 18% / 22% | -1,050 / -1,889 | -1,050 / -2,062 | FAIL |
| 4 d | 70% | -1,130 | -2,955 | -2,000 | -3,161 | 26% / 28% | -1,130 / -3,128 | -1,130 / -3,128 | FAIL |
| 7 d | 70% | -1,130 | -2,955 | -2,000 | -3,161 | 37% / 37% | -1,130 / -3,128 | -1,130 / -3,128 | FAIL |
| none | 70% | -1,130 | -2,955 | -2,000 | -3,161 | 33% / 37% | -1,130 / -3,128 | -1,130 / -3,128 | FAIL |

"Top" means the coins with the largest gain against live in that cell. The worst day is the same in every cell, because no worst day comes from a kit hold.

## Why the wider trail loses
- **The trail exit is the top.** A kit coin that falls 50% from its peak keeps falling.
  - With no cap, the 60% trail sold 12 of 15 coins lower than the 50% trail, 0 higher, and 3 the same (the 2 open holds and 0xb91b4020518cd0cf824dca8291e94cd62c7f0b91). The sum of x over the 15 coins fell from 112 to 90.
  - Example: 0xee049979960d5e42b389b941ec45790ca7b6e1f7 sells at 15.78x on the 50% trail, 11.14x on 60% and 7.03x on 70%.
- **A cap only ever helped on a wide trail.** On 60% and 70%, the 48 h cap sold 0xf15aab29c3d22cf2b469d4a9b0c0a72f31ba0711 at 13.93x before its crash. With no cap it went on to 5.30x.

## The two coins the cap could still matter for
- **Two open holds at data end:**
  - 0x224a794031d54e49b46919ade2c55ff255dfdd97 (11.36x);
  - 0xc38c332012a9116dcade0c7f43b5d3c71799a9f8 (9.25x).
  - Both gated 6 Oct around 06:45 and had only 4.2 h of data. Their x is a real-sale quote at the last row.
- **CHONK** 0x717d38587504b79e2e888c860bd1569c0d5984af (Kit B, not in the archive):
  - in the paper w2_v3_kit_hold book it peaked at 23.28x cap;
  - it was at 14.16x at 09:56 today (0.61 of peak, still above the trail line);
  - its 48 h cap sells it 8 Oct 23:52 Bangkok;
  - this is the first live test of whether the cap cuts a runner. Watch it.

## Limits
- Only 15 coins, and 2 are censored.
- Tracked hours after entry range from 4.2 h to 169 h per coin. Every trail exit happened inside the tracked window.
- The 4 d, 7 d and no-cap cells are "same as live" in the data we have, not proven equal. Re-run when 10 more kit 5x coins exist.

## Files
- http://159.223.88.168/hx-c9ois/sw/sw-a2-34-cells.csv
- http://159.223.88.168/hx-c9ois/sw/sw-a2-34-coins.csv (per coin, every cell's x, hold hours, why)