rev 4 · 2026-08-17 night. Five strategies, one shared core (rev 4 core required by all). The $325 funded port is NT8-verified 2026-08-17 (726/726 on Arman's own Analyzer). Three more funded variants added — see the second table below. NEW: P1 Fast reBreak M1be $325 — funded port (LTF02582991480001), core verified 721/721 trades identical to engine variant 14767 (the funded-pro-25k battery cell) and 314/314 / $23,418.05 exact against the $400 funded book. The eval port is unchanged (222/222, $32,106.10 exact) but needs the rev-3 core — the core file is shared and there must be exactly one copy of it.
sha256 P1FastCore.cs = 4dd7b84e2f47ae2db5549eaf660334c4d0b39329beeee2e0156c9213ebf3e493 (rev 4, C# 6-safe; RearmMode/DayCapUsd/TakeCount)
sha256 P1FastLucidFunded25k.cs = 07a2afc641a8bf4a186d5732352ca7bafaf64142e93f67f80f002b689c8d5947 (rev 4: +TakeCount/ConfCountBits params + input validation, defaults unchanged)
sha256 P1FastLucidFunded25kRminusC2.cs = f523822ccfdb9f46164a853f0cf25f6c5907e8f06a5d05b28487da8e50c50085
sha256 P1FastLucidFunded25kRminusC3.cs = 617893d5bb8da577ab50bffd4fbb84fec8584d663e00f8540d801559c5a32838
sha256 P1FastLucidFunded25kCap700.cs = c18a2eca32a2547b7a0586602610653419eb768b7f2d1d4acb3892c4d495a83a
sha256 P1FastLucidEval25k.cs = f93ce7d82e9d15d1c89b9dc9d6cad2da59c47d2f40dfb021b8f13eda84390fc1 (rev 4: RealizedDayPnlForCap plumbing, logic unchanged)
The multi-year 1-minute MES data file (MES.Last.txt, 2019 → 2026-08-16, the verified engine series,
re-exported 2026-08-17) is not hosted here — transfer it privately from the Mac: trading/nt8/out/MES.Last.txt.
trading/nt8/NT8-GUIDE.md §6.| Window | Trades | Engine ceiling (raw) | Tick-rounded (what NT8 should be near) |
|---|---|---|---|
| 2023-08-01 → 2026-07-09 (battery page window) | 314–318 | $23,820 free / $18,563 @ $0.74/side | $20,655 free / $15,328 @ $0.74/side |
| 2023-08-01 → 2026-08-13 | 320–324 | $24,090 / $18,765 | $20,928 / $15,533 |
| 2024-01-01 → 2026-08-13 | 286–289 | $22,542 / $17,900 | $19,684 / $14,991 |
| 2026-01-01 → 2026-08-13 | 63–64 | $8,453 / $7,702 | $7,503 / $6,736 |
| 2026-06-15 → 2026-08-13 (Tradovate front contract) | 15 | $511 / $350 | $504 / $342 |
Per year (raw, $0.74/side): 2019 −$1.5k · 2020 −$3.8k · 2021 −$1.8k · 2022 −$6.2k · 2023 +$52 · 2024 +$844 · 2025 +$9.4k · 2026 +$7.7k — the edge is 2025+, 2019–2022 lose. Fingerprints of a correct run: largest loss ≈ −$325…−$350 · breakeven exits ≈ $0 (−$9…−$12 with commission) · wins ≈ $570–640 · max 20 contracts · some days with 2–3 trades / repeated same-direction entries (2026-06-29 two longs at 7479, 2026-08-07 two longs at 7763.50) · flat by 15:56 ET · never more than one open position. Recent trades 2026-06-15+: 06-17 L SL, 06-17 S SL, 06-19 L SL, 06-25 S BE, 06-29 L SL, 06-29 L EOD-win, 06-30 L TP, 07-01 L TP, 07-08 S BE, 07-14 L SL, 07-15 L SL, 07-21 L TP, 07-24 S BE, 08-07 L SL, 08-07 L TP. Not TV-verified at $325 (the $400 sibling is). Optimistic fill model — ceilings, not live expectations. Sim ≠ live.
| Strategy | 2023-08 → 2026-08 trades | commission-free | @ $0.74/side | full 2019 → 2026 (free / $0.74) |
|---|---|---|---|---|
| P1FastLucidFunded25kRminusC2 | 602–608 | ≈ $6.7k (ceiling $16.1k) | ≈ −$5.5k (ceiling +$4.0k) | −$7.5k / −$38.2k |
| P1FastLucidFunded25kRminusC3 | 531–535 | ≈ $24.3k (ceiling $30.4k) | ≈ $13.6k (ceiling $19.8k) | +$16.4k / −$9.8k |
| P1FastLucidFunded25kCap700 | 311–314 | ≈ $27.4k (ceiling $30.3k) | ≈ $21.1k (ceiling $24.1k) | +$19.7k / +$4.5k |
Harness gates vs the engine books: c2 1432/1440 identical (+10 classified structural), c3 1233/1233, cap700 707/707. The R_minus books trade 1,200–1,400 times, so commissions swing them by ~$26–30k over the full window — compare against the column matching your commission setting. Day-cap parity: run commission-free (the overlay counts $0.50/side). Both R_minus cells are recorded on the battery page as failed kill bar 1. Not TV-verified. Sim ≠ live.
Below: the original eval-port notes (still valid).
Documents\NinjaTrader 8\bin\Custom\Strategies\ (exactly ONE P1FastCore.cs, ever).P1FastLucidFunded25k / P1FastLucidEval25k appear in the strategy lists — P1FastCore is its
engine and never shows up anywhere. That is normal.| Setting | Value |
|---|---|
| Strategy | P1FastLucidEval25k, all parameters at defaults (LiveMode = False) |
| Data series | 1 Minute — anything else and the strategy disables itself (see Control Center → Log) |
| Trading hours | CME US Index Futures ETH (24h). An RTH template silently breaks the overnight levels — #1 mistake. |
| Slippage / Commission | 0 / off (compare against the commission-free numbers below) |
| Order fill resolution | Standard |
| Instrument + range | Tradovate data: front contract only, from when it became front month (MES Sep26 → 2026-06-15 onward; earlier bars are thin back-month garbage). Imported engine data: plain MES (or a far contract as container), any range 2019 → Jul 2026. |
| Window | Trades | Expected net |
|---|---|---|
| 2023-08-01 → 2026-07-01 | ~222 | ≈ $38,200 (at $0.74/side: ≈ $30.8k–32.1k) |
| 2026-01-01 → 2026-07-01 | ~37 | ≈ $13,000 |
| 2026-06-15 → 2026-07-08 | 7 | ≈ +$718 (Jun 17 ×2 losses, Jun 25 BE, Jun 29 loss, Jun 30 + Jul 1 wins, Jul 8 BE) |
Sanity fingerprints of a correct run: largest loss ≈ −$800 (the 1R cap) · a batch of $0 “even” trades (breakeven exits) · wins ≈ $1,300–1,600 (2R) · entries only 09:36–16:00 ET · flat by 15:56 · max one long + one short per day.
trading/nt8/incoming/) for the trade-by-trade diff against the
verified book.trading/nt8/NT8-GUIDE.md on
the Mac. Backtest numbers here use an optimistic fill model — ceilings, not live expectations. Sim ≠ live.