← Back to NitroPosted 23 Feb 2026 · Commenced 1 Oct 2026
Nitro Trades · video
This ONE Candle Will Change Your Trading | Candle Range Theory (CRT)
Code base used · SMT All-in-One + Roman Essentials v0.1
Target stack for Nitro CRT. Versions per /workspace/smt-python/VERSION_LOG.md.
- Code base
- SMT_AllinOne + Roman v0.1: Python power-test surface pinned 2026-09-25;
smt_strategy/allinone/, CLIsmt_strategy/run_smt_allinone_roman.py. - Order manager
- Roman Essentials Host v0.1 (pinned 2026-09-24). Param bag
essentials_params_full.json, version 2026-09-25: 294 params across 11 categories and 37 groups. - Entry plugin
- Nitro CRT plugin v1 (
nitro_crt_v1, 2026-09-24) is entry-only: it sends entry signals, and Roman Essentials/ZZ manages the trades (size, stops, TP ladder, session, max trades, trailing). - Ninja (reference)
- Full SMT Strategy V3 (
AllinOneStrategy) on PC2.
Honesty line: today's Nitro CRT backtest (run_nitro_crt_a_rebreak_bt.py, 10 stocks, clean data: −$86,953, PF 0.90, 370 trades) used a simplified Roman-style order-management stand-in (one stop, one target, 0.5% risk, max 2/day), not the full stack. It has not yet been run through the full All-in-One + Roman order manager. Source: smt-python/results/nitro_crt_weekly_clean_2026-10-01/compare_summary.json (STOCKS_COMBINED).
This ONE Candle Will Change Your Trading | Candle Range Theory (CRT)
Key frames 8 flagged
Source: WATCH_ALONG.md Curated timeline.
01:09 · face cam
on screen face cam
what's said '1, 2, or 4-hour... I personally like the 1-hour'
rule it illustrates TF
02:03 · CRT drawing
on screen CRT drawing
what's said 'I don't want to see a break and hold. I want to see a break and immediately reverse'
rule it illustrates C2
03:51 · drawing
on screen drawing
what's said 'it can literally just give you the market direction'
rule it illustrates CRT as bias
05:57 · MNQ 1H CRT, levels drawn
on screen MNQ 1H CRT, levels drawn
what's said 'breaks above the high and reverses right back down'
rule it illustrates valid CRT
07:15 · 1H: candle that never broke middle-candle low
on screen 1H: candle that never broke middle-candle low
what's said 'Price never came and broke it on the next candle'
rule it illustrates no break, no trade
08:00 · 2m chart, middle-candle low line
on screen 2m chart, middle-candle low line
what's said 'the candle must close, it must break and close'
rule it illustrates close rule
08:25 · Short tool: Stop 17.50 $750 / Target 35.00 $1,500 / RR 2; Tue 03 Feb '26 09:10
on screen Short tool: Stop 17.50 $750 / Target 35.00 $1,500 / RR 2; Tue 03 Feb '26 09:10
what's said 'take my profits at 25922.25 because that is where the CRT low is'
rule it illustrates TP/stop
09:15 · target hit
on screen target hit
what's said 'Absolutely flawless'
rule it illustrates result
Worked examples (1 A+ / 2 total)
| id | kind | t | instrument | date on screen | CRT TF | dir | entry mode | entry | stop | target | RR | outcome | frame |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| vAu3-A1 | a_plus | 05:40–09:20 | MNQ1! | Tue 03 Feb '26, entry 09:10 (axis UTC-5 = ET) | 1H | short | 2m candle close below middle candle low (2 earlier breaks ignored) | ≈25,957.75 | ≈25,975.25 (17.50 pts, $750) | 25,922.25 (said) / 25,922.75 (axis label); 35.00 pts, $1,500 | 2 | TP — 'Price takes it out. Absolutely flawless' | 030_05m55s_10s.png |
| vAu3-N1 | no_trade | 07:06–07:24 | MNQ 1H | not shown | not shown | not shown | not shown | not shown | not shown | not shown | Earlier CRT not traded: next candle never broke the middle candle | 039_07m15s_10s_crt.png |
vAu3-A1 · frames
vAu3-N1 · frames
Summary bullets 8
(en-orig deep-card captions; timestamps are video mm:ss; “not stated” is retained where the card gives no rule.)
Main logic
- [01:09–02:40] Mark 1H/2H/4H wick range; C2 sweeps/reverses and closes inside; C3 seeks opposite end.
- [03:51–04:09] CRT can supply direction for BOS-FVG.
Entry
- [04:25, 06:49–08:17] 2m/3m must break and close beyond C2; earlier breaks that do not hold are ignored.
Stop
- [04:44, 08:35] Drag to 2R; example tool shows $750.
Target
- [05:04, 08:25] Opposite C1 extreme; example $1,500.
Risk filters
- [06:49–07:41] no valid hold/no trade; [09:27] sweet spot 09:30–12:30 ET; no news/day filter stated.
Timeframe
- [01:09] 1H preferred, 2H/4H allowed; [04:25] 2m/3m; NY/Asia/London can work.
Exceptions
- [07:06–07:24] earlier CRT skipped because C2 was not broken; example entered at 09:10 before stated sweet spot; HTF bias/sizing not stated.
Pipeline outcome
This ONE Candle Will Change Your Trading / CRT
Video v-Au32NSfS8 · updated 2026-10-01
1Strategy write-up
- Main logic
- CRT is a secondary HTF-direction/standalone model: mark a 1H (or 2H/4H) range, require C2 sweep and close back inside, then seek delivery to the opposite end. It can also provide direction for BOS-FVG.
- Entry
- 2m/3m candle must break and close beyond C2 extreme; ignore earlier breaks that fail to hold (06:49–08:17).
- Stop
- Drag stop to 2R as a rule of thumb; the example shows 17.5 points/$750, but no structural stop rule is given (04:44, 08:35).
- Target
- C1 opposite extreme; example target 35 points/$1,500 (05:04, 08:25).
- Risk filters
- Require a valid close/hold; no news/day filter; sweet spot 09:30–12:30 ET, though the example entry is 09:10 pre-open.
- Timeframe
- 1H preferred, 2H/4H allowed; 2m/3m execution; NY/Asia/London can work; preferred NY/Asia.
- Exceptions
- Earlier CRT was not traded because next candle did not break C2; example is pre-open; CRT is described as advanced/secondary; no HTF bias rule, sizing or news filter is stated.
2Already in All-in-One
- CRT geometry: run_nitro_crt_a_rebreak_bt.py::detect_and_simulate (sw_lo/sw_hi, inside) and sweep_2026-09-29/sweep_core.py::gen_candidates
- Opposite-C1 target: run_nitro_crt_a_rebreak_bt.py::detect_and_simulate target_mode="c1_opposite"; sweep_core.py::TARGETS
- Set-and-forget exit scan: run_nitro_crt_a_rebreak_bt.py exit loop / allinone_v4/engine.py::_exit_scan (stop/target only when BE disabled)
3New features
| Feature | Notes | Status |
|---|---|---|
| crt_htf_bars_clock | Clock-grid 1H/2H/4H bars. | not yet coded |
| entry_confirm_close | 2m/3m close beyond C2. | not yet coded |
| stop_from_target_rr | 2R target-anchored construction. | not yet coded |
| entry_retest | Ignore weak first breaks / later hold. | not yet coded |
| sessions_multi | NY/Asia/London windows. | not yet coded |
Playbook summary what this video teaches
Limited data in this video's pack: no HTF bias rule, structural stop, sizing or news filter is stated; CRT is presented as a secondary/advanced model.
- Setup: mark a 1H (or 2H/4H) wick range; C2 sweeps and closes back inside; C3 seeks the opposite end. CRT can also give direction for BOS-FVG.
- Entry: a 2m/3m candle must break and close beyond C2's extreme; earlier breaks that don't hold are ignored. An earlier CRT was skipped because C2 wasn't broken.
- Stop/target: TP at C1 opposite (example 35 pts / $1,500); drag stop to 2R as a rule of thumb (example 17.5 pts / $750).
- Timeframe: 1H preferred; 2m/3m execution; sweet spot 09:30-12:30 ET (the example entry was 09:10, pre-open). NY/Asia/London can work.
- No news/day filter stated.
Source: /workspace/nitro-setups/nitro_last15_2026-10-01/v-Au32NSfS8/outcome.json (+ BULLETS.md / DEEP_CARD.md in the same pack)
Code model notes how the ideas map to code
Limited data in this video's pack: no HTF bias rule, structural stop, sizing or news filter is stated; CRT is presented as a secondary/advanced model.
(a) Parameters and suggested defaults
| Parameter | Suggested default | Source |
|---|---|---|
crt_tf | 1H (2H/4H allowed) | video (current code: 5m) |
exec_tf | 2m / 3m | video (current code: 5m) |
entry_confirm | close beyond C2 that holds | video (current code: 5m close beyond C2) |
target | C1 opposite extreme | video = current code: c1_opposite |
stop | 2R from TP | video (current code: structural) |
session | 09:30-12:30 ET sweet spot | video (current code: 09:30-12:00) |
bias | not specified in video, needs testing | not in video (current code: 15m EMA20/50) |
(b) Detection and entry logic
- Detect C2 sweep + close inside on 1H/2H/4H.
- Wait for a 2m/3m close beyond C2's extreme; discard breaks that don't hold.
- TP = C1 opposite; stop = 2R from TP.
- Optionally export CRT direction as a bias input for a BOS-FVG entry.
(c) Already coded in our Nitro CRT implementation
- C1/C2 sweep + close back inside: run_nitro_crt_a_rebreak_bt.py::detect_and_simulate (sw_lo / sw_hi, inside, c2_edge_tol_frac=0.0), on 5m bars only (resample_ohlc '5min').
- Bias: run_nitro_crt_a_rebreak_bt.py::_bias_series bias_mode='ema20_50' on 15m (last completed bar at C2 close); counter-bias skipped.
- Entry: A+ rebreak only. Pullback past the C2 body midpoint ((c2.open+c2.close)/2), then a 5m close beyond the C2 opposite extreme; a first break without pullback is skipped (gate b_pack_no_pullback). 5m close back through the sweep extreme cancels.
- Stop: beyond the C2 sweep extreme, floored by run_nitro_crt_a_rebreak_bt.py::min_stop_distance (stocks max($0.05, 0.05% of price); NQ 5 pts). Structural, not TP-anchored.
- Target: run_nitro_crt_a_rebreak_bt.py::detect_and_simulate target_mode='c1_opposite' (long = C1 high, short = C1 low); skip if R:R < min_rr=0.5 (rr_fallback_R=None).
- Set-and-forget: 1m stop/target scan from the fill, flat at eod_flat 16:00 ET; no BE or trail.
- Session/caps: SESSION_START 09:30 to SESSION_END 12:00 ET, max_per_day=2 (MAX_TRADES_PER_DAY), one position at a time, RISK_PCT=0.005.
- Plugin nitro_crt_v1: python/signal_contract.py::build_signal only formats an entry signal (trigger_price, stop_hint, target_hint); ninja/NitroCRT_Entry_v1.cs is a stub (EmitOnly=true, session + RequireHtfBias knobs, no C1/C2 detector). Roman Essentials/ZZ is the order manager.
- Research only: backtests/sweep_2026-09-29/sweep_core.py grids 15/30/60m range bars, 15m/1h/4h EMA bias, 1R/1.5R/2R targets, 'immediate' (C2-close) entry, maxday 1/2. Not in the runner or plugin; verdict NO PROMOTION (SUMMARY.json).
(d) Still to build (from outcome.json)
| Feature | Note | vs current code | Status |
|---|---|---|---|
crt_htf_bars_clock | Clock-grid 1H/2H/4H bars. | partly covered by sweep_2026-09-29/sweep_core.py::range_bars (research grid: 15/30/60m bars, RTH 60m anchored at 09:30, not CME clock-hour; not in runner/plugin); still marked not yet coded pending test | not yet coded |
entry_confirm_close | 2m/3m close beyond C2. | partly covered by run_nitro_crt_a_rebreak_bt.py (enters on a 5m close beyond the C2 extreme; no 2m/3m bars, no strength test); still marked not yet coded pending test | not yet coded |
stop_from_target_rr | 2R target-anchored construction. | not covered: code stop is structural; sweep_core 1.5R/2R are targets from structural risk (the reverse) | not yet coded |
entry_retest | Ignore weak first breaks / later hold. | partly covered by run_nitro_crt_a_rebreak_bt.py (A+ rebreak: midpoint pullback then 5m close beyond C2; no 1m or break-retest mode); still marked not yet coded pending test | not yet coded |
sessions_multi | NY/Asia/London windows. | not covered: runner is NY 09:30-12:00 only | not yet coded |
Code checked: /workspace/nitro-setups/backtests/run_nitro_crt_a_rebreak_bt.py, /workspace/smt-python/entry_plugins/nitro_crt_v1/ (read only), backtests/sweep_2026-09-29/sweep_core.py. Statuses unchanged from outcome.json.
DSG /dev · Top Traders · training / second-gate only