# DC Engine: Q5c price path scorer (for A1)

- 7 Oct 13:23 Bangkok, DC Engine. A1 owns the Q5c rule (holder growth >= +528, LP gate). The Engine owns the price path.
- Script: research-box /opt/tw/engine/eng/q5c.py (md5 bb1f2feffc6d3d70ac28d4151aa2a6f8), engine twtest v1.6 fills (md5 c0f2b013970699f157b0f313b49c4201).
- Universe source (lesson 25): engine v4 full tape (PP wash.db, every v4 coin with a sellable $100k, 21 Sep to 6 Oct 05:52) + v3/Pons tape (/opt/sw/v3hist.db). **Launch based, no outcome filter:** every coin with a first sellable $300k print, 5x or not. Chain clock only.
- Output (one row per coin): http://159.223.88.168/hx-c9ois/dc/en/q5c_paths.csv

## Rules coded (from the 13:16 Q5c line and the 13:18 amendment)
- Entry = first trade in a later block than the first sellable $300k print ($100+ sell at cap >= $300k), pool not pulled.
- BASE (SW honest window) = sell at the first sellable 5x, or the first sellable print <= 0.3x (stop), or 48 h (last sellable print). Pull = 0.
- V1 at read r (15 min and 60 min): only if BASE is still open at entry + r. From entry + r, ignore BASE; exit at the first sellable print <= 50% of the peak sellable price after entry + r, or at 48 h. Pull = 0. If BASE closed before the read, V1 = BASE.
- $ at $100 and $500, depth from the last 30 swaps, pool fee each way (dynamic fee 1%).
- Columns: coin, key (pool for v3), venue, half (A before 29 Sep Bangkok), te, pe, cap_e, base_exit, base_mult, base_100/500, v1_15_open, v1_15_exit, v1_15_mult, v1_15_100/500, v1_15_add_100/500, same for v1_60.
- A1 joins: fire = first read where growth >= +528 (use v1_15 or v1_60 columns), LP class for V2 (V2 = V1 if gated, else BASE). Random draws = the same columns on random open positions of the same half.

## First numbers (all coins, no holder filter, $500)
| Half | Scored | BASE $/trade | 5x share | Pulled in 48 h | V1@15 added, all open | V1@60 added, all open |
|---|---|---|---|---|---|---|
| A (12 to 28 Sep) | 848 | -269 | 7% | 417 | +103.5 | +99.4 |
| B (29 Sep to 4 Oct) | 1,406 | -410 | 3% | 1,080 | +88.2 | +8.9 |

- 2,608 coins printed a sellable $300k; 2,254 scored; 324 not covered (entry + 48 h past tape end), 22 no fill, 8 no depth.
- **Trap for A1: the trail beats BASE even with NO holder signal** (+$88 to +$103 a position at the 15 min read). It wins mostly by exiting before the 0.3x stop. So the random bar is high: holder growth must add more than a random set of open positions does.
- Venue: v4 BASE -$442, v3 -$62, Pons -$142 a trade. v4 starts 21 Sep on this tape (15 to 20 Sep v4 missing).
- Limits: no public router filter; v4 pulls after 6 Oct 05:52 unseen; trail exit price is the print itself (can gap far below 50%).