← Back to NitroPosted 31 Jul 2026 · Commenced 24 Sep 2026
Nitro Trades · video
The Most PERFECT Trade You Will Find Everyday | Candle Range Theory (CRT) Advanced
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).
The Most PERFECT Trade You Will Find Everyday | Candle Range Theory (CRT) Advanced
Key frames 34 flagged
Source: all 34 WATCH_ALONG.md Timeline frames (deduped key frames).
00:10 · Talking head intro, captions burned in.
On screen Talking head intro, captions burned in.
Said "what a perfect CRT will look like and how you can find that exact setup"
Rule illustrated Premise: an 'advanced' CRT filter set.
00:34 · Whiteboard 'CRT ADVANCED GUIDE' with 6 bullets: Bias & Trend / Session Times / Price Action / Timeframes / News/Days of the week / Break vs Retest. 3-candle diagram: blue C1, red C2 wick above the C1 high, red C3 breaking down to the C1 low; red lines at the C1 high/low.
On screen Whiteboard 'CRT ADVANCED GUIDE' with 6 bullets: Bias & Trend / Session Times / Price Action / Timeframes / News/Days of the week / Break vs Retest. 3-candle diagram: blue C1, red C2 wick above the C1 high, red C3 breaking down to the C1 low; red lines at the C1 high/low.
Said "today I want to break down that exact perfect setup. Knowing when to skip, knowing when to size heavier or size smaller"
Rule illustrated The checklist itself (the backbone of the card).
01:35 · Arrows: price sweeps the C1 high; the C2 body closes back inside.
On screen Arrows: price sweeps the C1 high; the C2 body closes back inside.
Said "price comes and breaks that high and gets an immediate reversal back down… closes back underneath the high"
Rule illustrated CRT validity: sweep + body close back inside the range.
02:28 · Arrow from C2 down to the circled C1 low; a mark at the C2 low (entry).
On screen Arrow from C2 down to the circled C1 low; a mark at the C2 low (entry).
Said "enter on the break of this middle candle and target a move down to this low"
Rule illustrated Entry = break of the middle candle; target = the opposite C1 extreme.
03:23 · Right side: 1H HH/HL zig-zag sketch sweeping a high.
On screen Right side: 1H HH/HL zig-zag sketch sweeping a high.
Said "This is the 1-hour time frame… price comes and sweeps it"
Rule illustrated Bias & Trend: HTF sweep context.
04:10 · Sketch finished: a sweep of the high, then an arrow down (massive rejection).
On screen Sketch finished: a sweep of the high, then an arrow down (massive rejection).
Said "if we just swept a key level, then I could definitely flip my bias. Even though we're not making lower highs"
Rule illustrated Bias-flip override (counter-trend allowed after an HTF sweep/OB).
06:12 · Handwritten: 'A – 7–10pm', 'NY – 9:30–12/1pm', 'L – 4am–9am'.
On screen Handwritten: 'A – 7–10pm', 'NY – 9:30–12/1pm', 'L – 4am–9am'.
Said "New York… 9:30 until about 12:00 to 1:00 p.m. Eastern… Asian… 7:00 to 10:00 p.m.… London… 4:00 a.m… until about 9:00 a.m."
Rule illustrated Session windows (ET). NY is best; Asia 8pm is his most-traded time.
07:55 · '15m / 30m / 1H' written top right; chop candles sketched (big bodies, wicks on both sides).
On screen '15m / 30m / 1H' written top right; chop candles sketched (big bodies, wicks on both sides).
Said "massive wick to both sides… This is not going to get followed" / "15-minute, 30-minute, and 1-hour for the CRT itself"
Rule illustrated Price-action (1m/2m) chop filter + CRT TF set.
10:01 · A text box being typed: 'Monday – / Tuesday –'.
On screen A text box being typed: 'Monday – / Tuesday –'.
Said "if you trade on a no news Monday, you're asking to lose money"
Rule illustrated News/day filter: skip no-news Mondays; red-folder days are best.
11:30 · Full board: scribbles on C3 showing a break and a retest of the C2 low.
On screen Full board: scribbles on C3 showing a break and a retest of the C2 low.
Said "If it's a lot choppier day, I'll wait for a retest. If I see it's a lot cleaner, I'll just go for a break."
Rule illustrated Entry mode: break vs retest.
11:54 · TradingView replay, MNQ1! 1h (CME), 17–24 Jul 2026. Horizontal line ≈28,410 (17–18 Jul low); volume profile.
On screen TradingView replay, MNQ1! 1h (CME), 17–24 Jul 2026. Horizontal line ≈28,410 (17–18 Jul low); volume profile.
Said "the first thing is I'm marking out my key levels… any point where price had a strong reversal"
Rule illustrated Key-level definition.
12:25 · The level ≈28,410 circled in red with a second line ≈28,655; the last 1h candle wicks to ≈28,254 and closes ≈28,449 (header O28,323.25 H28,461.75 L28,254.25 C28,449.00).
On screen The level ≈28,410 circled in red with a second line ≈28,655; the last 1h candle wicks to ≈28,254 and closes ≈28,449 (header O28,323.25 H28,461.75 L28,254.25 C28,449.00).
Said "we just came and took out a very key hourly level… insta candle reversal back up"
Rule illustrated Bias flip to bullish on the 1H key-level sweep.
12:51 · 1m chart, Fri 24 Jul 2026, 09:00–13:00 ET; 15m/30m/1H candle overlay on the right; grey range box ≈28,345–28,452; blue box above.
On screen 1m chart, Fri 24 Jul 2026, 09:00–13:00 ET; 15m/30m/1H candle overlay on the right; grey range box ≈28,345–28,452; blue box above.
Said "you have a nice little 30-minute CRT… hourly and all these different CRTs that are forming"
Rule illustrated CRT forms on 30m and 1H (context 1H, execution 1m).
13:15 · 🟢 A+ LONG #1 — A+ LONG #1: position tool, entry ≈28,450; 'Stop: 104.00 (0.366%) 416'; 'Target: 172.50 (0.606%) 690'; 'Risk/reward ratio: 1.66'.
On screen 🟢 A+ LONG #1 — A+ LONG #1: position tool, entry ≈28,450; 'Stop: 104.00 (0.366%) 416'; 'Target: 172.50 (0.606%) 690'; 'Risk/reward ratio: 1.66'.
Said "when it breaks that middle candle, this is an understandable entry. I can see why I would play against the trend"
Rule illustrated Entry on break of middle candle; stop/target geometry.
13:36 · 🟢 A+ LONG #1 — Replay advanced: price ≈28,512 at ~11:45 ET, trending up inside the blue target box.
On screen 🟢 A+ LONG #1 — Replay advanced: price ≈28,512 at ~11:45 ET, trending up inside the blue target box.
Said "You place your stop and TP, and you let it do its thing."
Rule illustrated Set-and-forget management.
13:46 · 🟢 A+ LONG #1 — Price ≈28,596–28,618 around 12:05 ET, at the top of the target box.
On screen 🟢 A+ LONG #1 — Price ≈28,596–28,618 around 12:05 ET, at the top of the target box.
Said "the hourly CRT, the 30-minute… both got followed through because it took out a liquidity key level"
Rule illustrated Outcome: target reached.
14:10 · Back on the whiteboard checklist to recap.
On screen Back on the whiteboard checklist to recap.
Said "Number one… bias and trend… Session times… around 10:58, which is totally fine"
Rule illustrated Checklist walk-through applied to Ex1.
14:52 · 🟢 A+ LONG #1 — Ex1 re-shown: full position tool over 08:30–14:45 ET; entry break at ~10:55–11:00.
On screen 🟢 A+ LONG #1 — Ex1 re-shown: full position tool over 08:30–14:45 ET; entry break at ~10:55–11:00.
Said "You can either play the breakout, or… first initial failure, and break a second time, or break out and pull back to that midpoint"
Rule illustrated Three entry options (break / 2nd break / pullback to 'midpoint' [AMBIG]).
15:21 · 🟢 A+ LONG #1 — Ex1: price at the top of the target box ≈28,620 around 12:05 ET.
On screen 🟢 A+ LONG #1 — Ex1: price at the top of the target box ≈28,620 around 12:05 ET.
Said "instead of like… a 65% win rate… go for 75, 80"
Rule illustrated Win-rate claim (unverified).
15:27 · MNQ1! 1h, ~30 Jun–13 Jul 2026: grey 1H order block ≈30,000–30,290.
On screen MNQ1! 1h, ~30 Jun–13 Jul 2026: grey 1H order block ≈30,000–30,290.
Said "this time with a hourly order block"
Rule illustrated HTF OB context.
15:41 · Rejections at the OB circled in red (4 taps); crosshair Thu 09 Jul '26 20:00.
On screen Rejections at the OB circled in red (4 taps); crosshair Thu 09 Jul '26 20:00.
Said "First time tapped, rejected. Second time tapped, rejected. Third time came close… Fourth time again"
Rule illustrated Don't long into a repeatedly rejecting 1H OB.
15:51 · Zoomed: 4 circled taps and a red arrow projecting down from the OB.
On screen Zoomed: 4 circled taps and a red arrow projecting down from the OB.
Said "why would I long in this range? This is why the higher time frames are so important"
Rule illustrated HTF OB overrides LTF higher lows.
15:56 · 1m, Mon 13 Jul 2026, ~08:45–11:00 ET; the OB zone (red outline) projected overhead at ≈30,100.
On screen 1m, Mon 13 Jul 2026, ~08:45–11:00 ET; the OB zone (red outline) projected overhead at ≈30,100.
Said "I want to drop down to a lower time frame"
Rule illustrated Top-down: 1H → 1m.
16:07 · 1m + 15m/30m/1H overlay around 11:00 ET, price ≈29,696.
On screen 1m + 15m/30m/1H overlay around 11:00 ET, price ≈29,696.
Said "You do have bullish CRT… But… the best one… is going to be the bear CRT because we're rejecting off that range"
Rule illustrated Choosing the direction against the LTF trend.
16:30 · 1m ~12:00 ET, price ≈29,606; a red mark over the 30m candle; the 1H candle is visible.
On screen 1m ~12:00 ET, price ≈29,606; a red mark over the 30m candle; the 1H candle is visible.
Said "There was a nice 30-minute one… you also have nothing on 15. You have a nice hourly right here."
Rule illustrated CRT TF check (30m/1H yes, 15m no).
16:40 · 🔴 A+ SHORT #2 — A+ SHORT #2: position tool 'Stop: 68.75 (0.232%) 275', 'Target: 102.75 (0.347%) 411', 'Risk/reward ratio: 1.49', entry 29,596.
On screen 🔴 A+ SHORT #2 — A+ SHORT #2: position tool 'Stop: 68.75 (0.232%) 275', 'Target: 102.75 (0.347%) 411', 'Risk/reward ratio: 1.49', entry 29,596.
Said "You have TP at the lows, stops right here, and we're going against the trend"
Rule illustrated Stop/target shown (stop rule never stated).
16:47 · 🔴 A+ SHORT #2 — Live orders: sell 5 @ 29,596.00; SL 29,661.50 (−655 USD); TP 29,493.25 (+1,027.50 USD); toast 'Take Profit order placed… Buy 5 at 29,493.25'; tan zone ≈29,596–29,664.
On screen 🔴 A+ SHORT #2 — Live orders: sell 5 @ 29,596.00; SL 29,661.50 (−655 USD); TP 29,493.25 (+1,027.50 USD); toast 'Take Profit order placed… Buy 5 at 29,493.25'; tan zone ≈29,596–29,664.
Said "what would you do from here? You would let your trade ride"
Rule illustrated Real order bracket (qty 5 MNQ).
17:00 · 🔴 A+ SHORT #2 — SL now 29,665.50 (−695 USD); open P&L +297.50 USD; price ≈29,566.
On screen 🔴 A+ SHORT #2 — SL now 29,665.50 (−695 USD); open P&L +297.50 USD; price ≈29,566.
Said "The chop is not really crazy. Nothing wild is going on here."
Rule illustrated Stop moved 4 pts wider (screen only, not mentioned).
17:11 · 🔴 A+ SHORT #2 — Toast 'Stop Loss order cancelled… Buy 5 at 29,665.50' → TP filled (OCO); price ≈29,499 at ~13:50 ET.
On screen 🔴 A+ SHORT #2 — Toast 'Stop Loss order cancelled… Buy 5 at 29,665.50' → TP filled (OCO); price ≈29,499 at ~13:50 ET.
Said "let it do its thing until you get full TP or full stop loss"
Rule illustrated Outcome: TP hit (+$1,027.50 shown).
17:17 · 🔴 A+ SHORT #2 — Mon 13 Jul 14:03 ET, price ≈29,489, below the TP.
On screen 🔴 A+ SHORT #2 — Mon 13 Jul 14:03 ET, price ≈29,489, below the TP.
Said "A lot of people might try a long CRT here and get kind of cooked"
Rule illustrated HTF level > LTF trend.
17:43 · ⛔ NO-TRADE — NO-TRADE: 1m, Tue 21 Jul 2026, 09:15–11:05 ET; large alternating candles, 29,060–29,200.
On screen ⛔ NO-TRADE — NO-TRADE: 1m, Tue 21 Jul 2026, 09:15–11:05 ET; large alternating candles, 29,060–29,200.
Said "massive green candle… followed by red… Both sides are just wicking"
Rule illustrated Chop filter, live.
18:02 · ⛔ NO-TRADE — Red lines under the lows marking failed breakdowns; red scribbles on the wicks.
On screen ⛔ NO-TRADE — Red lines under the lows marking failed breakdowns; red scribbles on the wicks.
Said "Look at all these breakdowns fail. Breakdown fail. Breakdown fail."
Rule illustrated Skip choppy days.
18:16 · ⛔ NO-TRADE — Same chart, annotations complete.
On screen ⛔ NO-TRADE — Same chart, annotations complete.
Said "There is just no point in me playing that setup."
Rule illustrated Skip.
19:00 · Outro over the same chart.
On screen Outro over the same chart.
Said "study this one really well"
Rule illustrated —
A+ example cards (as shown)
| 🟢 A+ LONG #1 | 🔴 A+ SHORT #2 | |
|---|---|---|
| Video time | 11:53–15:21 | 15:20–17:20 |
| Instrument | MNQ1! (CME micro Nasdaq) | MNQ1! |
| Date / weekday | Fri 24 Jul 2026 | Mon 13 Jul 2026 (news status not stated) |
| HTF context | 1H key level ≈28,410 swept and reclaimed (plus a 2nd level); 1H trend was lower highs | 1H order block ≈30,000–30,290 rejected 4×; LTF made higher lows |
| Trend relation | Counter-trend (bias flipped by the sweep) | Counter-trend (bias from the OB) |
| CRT TF | 30m and 1H formed; took the 'better' one (unspecified) | 30m and 1H ("nothing on 15") |
| Exec TF | 1m | 1m |
| Entry mode | Break of the middle candle (his preference) | Break (the order sits at the bottom of the tan zone) |
| Entry time | ~10:58 ET (said) | ~12:00 ET (screen) |
| Entry | ≈28,450 | 29,596.00 (sell 5) |
| Stop | −104.00 pts → ≈28,346 (bottom of grey zone) | 29,661.50 → moved to 29,665.50 (top of tan zone), −655 → −695 USD |
| Target | +172.50 pts → ≈28,622 | 29,493.25 (+1,027.50 USD); tool 102.75 pts |
| R:R shown | 1.66 | 1.49 (tool); ≈1.57 → 1.48 (orders) |
| Outcome shown | Target box reached ~12:05 ET | TP filled (SL-cancel toast) ~13:50 ET |
| Frames | 014, 015, 016, 018, 019 | 026, 027, 028, 029, 030 |
🟢 A+ LONG #1 · frames
🔴 A+ SHORT #2 · frames
Example 3 — NO-TRADE: chop / volatility (17:40–18:13)
- Chart: MNQ1! 1m, Tue 21 Jul 2026, ~09:15–11:05 ET, price 29,060–29,200 (frames 031–033).
- He draws red lines under several lows and red scribbles over wicks. Said: "massive green, massive red… Both sides are just wicking… Your CRT is not going to get followed through. Look at all these breakdowns fail. Breakdown fail… There is just no point in me playing that setup." He calls reading this "intuitiveness".
- Rule: skip days where the 1m alternates big candles with two-sided wicks and repeated failed breaks.
Summary bullets 49
Sources: ../DEEP_SUMMARY.md (DS §) and the original English captions ../sources/QCt7_Cq9pQY.en-orig.timed.txt (cap [mm:ss] = caption chunk start; chunks run about 17 s, so a quote can sit a few seconds after the chunk stamp). Video times are mm:ss. Chart/clock times are ET. Instrument shown: MNQ1!. Tags: [SAID] spoken · [SCREEN] on screen only · [AMBIG] unclear or conflicting.
1. Main logic (CRT)
- 01:15 [SAID+SCREEN] Mark the most recent 15m/30m/1H candle's high and low. "go on the 15-minute, the 30-minute, or the 1-hour time frame and all you're doing is marking out the most recent candle's high and the most recent candle's low" (cap [01:10]; DS §1)
- 01:27 [SAID] Those levels are liquidity. The trade picks orders up at one end and delivers them to the other. "pick up orders and deliver them to the opposite end of that candle" (DS §1)
- 01:40 [SAID+SCREEN f003] Setup: the next candle (the middle candle, C2) breaks C1's high (or low), then reverses. "price comes and breaks that high and gets an immediate reversal back down… it closes back underneath the high after breaking it" (DS §1)
- 02:14 [SAID] Validity: the body closes back inside. "The candle must close, the body of the candle must close back inside the range" (cap [02:01]→[02:19])
- 02:48 [SAID] Default: trade with the trend. "I always say take the CRTs that are with the trend, and that's the simple way to do it." (cap [02:36])
- 03:29 [SAID] Counter-trend override: a sweep of an HTF key level, or an order block that's tapped and rejects, lets him flip bias without a structure shift. "if we just swept a key level, then I could definitely flip my bias. Even though we're not making lower highs… I can still flip if we are taking out order blocks or key liquidity levels." (cap [03:29])
- 03:55 [SAID] "we got a sweep or we tapped an order block and we rejected, and that is why I played a bear CRT against the trend" (cap [03:48])
- 11:58 [SAID+SCREEN f011] A key level is "any point where price had a strong reversal" (DS §2.1)
- 17:28 [SAID] Best single setup: "the best setup is still going to be the first trade of the New York session with the trend." (cap [17:22])
2. Entry rules
- 02:25 [SAID+SCREEN f004] "you're going to enter on the break of this middle candle and target a move down to this low" (cap [02:19])
- 10:35 [SAID] Break mode: "enter on the break of this middle candle, which is giving you a guaranteed entry, but you can get faked out" (cap [10:35])
- 10:50 [SAID] Retest mode: "comes back up and retests your level and then gets another rejection. Way safer, way higher win rate… but you will probably miss a lot of setups" (cap [10:53])
- 11:33 [SAID] How he picks the mode: "If it's a lot choppier day, I'll wait for a retest. If I see it's a lot cleaner, I'll just go for a break." (cap [11:28])
- 14:33 [SAID+SCREEN f018] Three entry options: (1) the breakout; (2) "first initial failure, and break a second time"; (3) "pull back to that midpoint, and that's where you enter" (cap [14:45]). [AMBIG] "midpoint" is never defined.
- 11:11 [AMBIG] The win-rate sentence contradicts itself: retest is called "lower win rate" once, and "higher win rate" everywhere else (DS §2.6).
- Examples [SCREEN] Ex1 long (Fri 24 Jul 2026) entry ≈28,450 at ~10:58 ET ("It's around 10:58", cap [13:50]). Ex2 short (Mon 13 Jul 2026): sell 5 MNQ @ 29,596.00 (DS §4).
3. Stop logic
- 16:40 [SAID, not defined] "You have TP at the lows, stops right here" (cap [16:32]). He never states a stop rule (DS §3, §6).
- Ex1 [SCREEN f014] Position tool: "Stop: 104.00 (0.366%) 416", so the stop is ≈28,346, the bottom of the grey range box (≈28,345–28,452) (DS §4; CARD_GAP §3.3).
- Ex2 [SCREEN f026–028] SL 29,661.50, then moved to 29,665.50 (4 pts wider, not mentioned). That's just above the top of the tan zone (≈29,596–29,664). Entry is at the zone's bottom (DS §4).
- CARD_GAP §3.3: "Shown stops sat at range-box edges… not at the HTF sweep extreme."
- 13:33 / 16:52 [SAID] Management is set-and-forget. "You place your stop and TP, and you let it do its thing." (cap [13:33]) / "until you get full TP or full stop loss" (cap [16:49])
4. Target logic
- 02:19 [SAID] "price wants to make a move all the way back down to the opposite end of that candle" (cap [02:19])
- 02:25 [SAID] Target = C1's opposite extreme: "target a move down to this low" (cap [02:19])
- 16:40 [SAID] "You have TP at the lows" (cap [16:32])
- Examples [SCREEN] Ex1 target +172.50 pts (≈28,622), R:R 1.66. Ex2 TP 29,493.25, R:R 1.49 (tool) / ≈1.57 (orders). No R:R rule is stated (DS §3).
5. Risk filters
- 05:25 [SAID] No CRTs outside the sessions: "a lot of people are just taking CRTs in the middle of like no man's land, right? There's no volume" (cap [05:17])
- 06:30 [SAID] Check the 1m/2m: "the 1-minute time frame, the two, and all these lower time frames tell us the real story" (cap [06:25])
- 06:42 [SAID+SCREEN f008] Skip chop, e.g. "massive wick to both sides… flip… and then another massive wick right here, flip" (cap [06:59]). "there might be the best-looking setups on the CRT ever. But if your price action sucks…" (DS §2.3)
- 17:43 [SAID+SCREEN f031–033] No-trade example (Tue 21 Jul 2026): "massive green, massive red… Both sides are just wicking… Look at all these breakdowns fail" (DS §4 Ex3)
- 08:47 [SAID] Skip no-news Mondays: "if you trade on a no news Monday, you're asking to lose money" (cap [08:50])
- 10:01 [SAID] Red folder: "whichever day has red folder news is the day you will have the best CRTs… If it has no red folder news, maybe best to just skip it all together." (cap [10:01]). Calendar: forexfactory.com (09:55).
- 08:11 [SAID] Avoid too much range: "I don't want to go for anything with too much range… I'll go for the tightest range possible." (cap [08:11])
- 17:32 [SAID] Max trades: "You want to go for one CRT or maximum two if you really want to." (cap [17:22])
- 00:52 [SAID] Sizing is mentioned ("size heavier or size smaller") but no rule is given (cap [00:52]).
6. Timeframe(s)
- 07:47 [SAID+SCREEN] "15-minute, 30-minute, and 1-hour for the CRT itself. Some people like the 4-hour." (DS §2.4)
- 08:00 [SAID] Selection: "I'll find one setup, whichever one forms first." (cap [07:53])
- 08:05 [SAID] If several form together: "I'll go for the lowest one because the 15-minute usually has the easiest follow-through" (cap [07:53]→[08:11]). "whichever one has the tighter range, if all three of them have it, is the one you will trade" (cap [08:30])
- Context TF [SCREEN] 1H, for key levels and order blocks (f011, f020–022). Execution/PA TF [SCREEN] 1m chart with a 15m/30m/1H candle overlay (f013). PA check on 1m/2m (06:30).
-
Sessions (ET) [SAID+SCREEN f007]:
- NY 09:30 to "12:00 to 1:00 p.m.", "I usually cut it off around like 1:00" (cap [04:05], [05:01])
- Asia 19:00–22:00, "8:00 p.m. Eastern time is where I personally trade the most" (cap [04:23]–[04:41])
- London 04:00–09:00, "a lot slower" (cap [05:17])
7. Exceptions / ambiguities
- Counter-trend is allowed after an HTF key-level sweep or a 1H OB rejection (03:29). Both A+ examples are counter-trend: "we're going against the trend… This is against what the basic idea of CRT is. But, how we find the perfect setups." (cap [16:32]–[16:49])
- 14:18 [AMBIG] Ex1 TF pick: "I personally prefer playing the one that has a little bit better setup". That's discretionary, and it conflicts with "lowest/tightest" (DS §2.4).
- [AMBIG] Ex1 is a day "there wasn't news" yet it's judged fine. That conflicts with 10:01 (DS §4).
- [AMBIG] Ex2 is on a Monday (13 Jul 2026). News status isn't shown (DS §4).
- [AMBIG] "Tuesday –" is written on the board but never explained (DS §2.5).
- 01:44 [AMBIG] "sell-side" is said for a high sweep. Probably a slip of the tongue (DS §1).
- Not in the video: an exact stop rule, sizing math, an OB drawing definition, the "midpoint" definition, partials/trailing (he says full TP or full SL), and instruments other than MNQ (DS §6).
Pipeline outcome
The Most PERFECT Trade You Will Find Everyday / Candle Range Theory (CRT) Advanced
Video QCt7_Cq9pQY · updated 2026-10-01
1Strategy write-up
- Main logic
- CRT marks most recent 15m/30m/1H range; C2 sweeps and body closes inside; target is opposite end, normally with trend, with a key-level/OB sweep override.
- Entry
- Break, retest, second break or midpoint-pullback modes; clean day favors break, choppy day retest (10:35–14:45).
- Stop
- No stop rule is stated; screenshots place stops near range-box edges, not sweep extreme (16:40, examples).
- Target
- Opposite C1 extreme (02:19, 16:40).
- Risk filters
- Avoid no-man’s-land, 1m/2m chop, too much range, no-news Monday; red-folder preference; max 1–2 CRTs.
- Timeframe
- 1m execution/1m–2m PA check; CRT 15m/30m/1H; first formed, lowest/tightest if simultaneous; NY 09:30–12/13, Asia 19–22, London 04–09 ET.
- Exceptions
- Counter-trend override after HTF key-level sweep/OB rejection; exact stop/midpoint/sizing/news data not defined.
2Already in All-in-One
- Existing CRT runner/sweep structure, target, trend proxy, day cap, and FVG/OB/swing primitives are cited in the existing AIO_DIFF.md.
3New features
| Feature | Notes | Status |
|---|---|---|
| crt_multi_tf | Clock-anchored multi-TF selection. | not yet coded |
| entry_middle_break | Executable first break at C2 opposite extreme. | not yet coded |
| entry_mode | Break/retest/second-break/midpoint modes. | not yet coded |
| sessions_multi | NY/London/Asia windows. | not yet coded |
| htf_bias_override | Key-level/OB counter-trend override. | not yet coded |
| skip_monday | Monday filter. | not yet coded |
| red_folder_filter | Red-folder/no-news gate; no data on box. | not yet coded |
| ltf_chop_filter | 1m/2m chop veto and mode selection. | not yet coded |
| range_cap | Avoid too-large C1 ranges. | not yet coded |
| first_ny_with_trend | First NY trend trade selector. | not yet coded |
| stop_modes | Box-edge/other stop modes; video stop unspecified. | not yet coded |
| mnq_economics | MNQ economics. | not yet coded |
| target_front_run | Optional target offset. | not yet coded |
Playbook summary what this video teaches
Limited data in this video's pack: no stop rule is stated (screens show stops at range-box edges), 'midpoint' is never defined, and no sizing rule is given.
- Setup: mark the most recent 15m/30m/1H candle's high/low; the next (middle) candle sweeps one side and its body closes back inside; target the opposite end.
- Bias: with the trend by default; counter-trend allowed after an HTF key-level sweep or 1H order-block rejection (both A+ examples were counter-trend).
- Timeframe: CRT 15m/30m/1H, first formed, lowest/tightest if simultaneous; 1m execution, 1m/2m price-action check. NY 09:30 to 12:00-13:00, Asia 19:00-22:00, London 04:00-09:00 ET.
- Entry: break of the middle candle, retest, second break, or pullback to a midpoint. Clean day favours the break; choppy day the retest.
- Stop/target: TP at C1's opposite extreme; stop not stated (screens put it near the range-box edge, not the sweep extreme). Set-and-forget.
- Filters: no CRTs in no-man's-land, skip 1m/2m chop, too much range and no-news Mondays; prefer red-folder days; max 1-2 CRTs.
- Best single setup: the first NY trade with the trend.
Source: /workspace/nitro_deep_card_test_2026-10-01/aio/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 stop rule is stated (screens show stops at range-box edges), 'midpoint' is never defined, and no sizing rule is given.
(a) Parameters and suggested defaults
| Parameter | Suggested default | Source |
|---|---|---|
crt_tf | 15m / 30m / 1H; first formed, tightest if tie | video (current code: 5m) |
bias | trend; override on HTF key-level / OB sweep | video (current code: 15m EMA20/50, no override) |
entry_mode | break / retest / second break / midpoint | video (current code: rebreak after C2-body midpoint pullback) |
exec_tf | 1m | video (current code: 5m) |
target | C1 opposite extreme | video = current code: c1_opposite |
stop | not stated in video | current code: beyond sweep + min distance |
sessions | NY 09:30-13:00, Asia 19:00-22:00, London 04:00-09:00 ET | video (current code: NY 09:30-12:00) |
max_crt_per_day | 1, max 2 | video (current code: 2) |
range_cap | not specified in video, needs testing | video says avoid too much range; no number |
(b) Detection and entry logic
- Scan 15m/30m/1H for C2 sweep + body close inside; take the first formed, tightest range on a tie.
- Direction with trend unless an HTF key level / OB was just swept and rejected (override flag).
- Choose entry mode from 1m/2m chop: clean = break of the middle candle, choppy = retest or second break.
- TP = C1 opposite; stop mode is an open parameter (box edge per screens).
- Vetoes: outside sessions, no-news Monday, no red-folder news, oversized range.
(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_multi_tf | Clock-anchored multi-TF selection. | 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_middle_break | Executable first break at C2 opposite extreme. | not covered: runner skips first breaks without a pullback (b_pack_no_pullback); NITRO_CRT_B_FIRST_BREAK is a contract only | not yet coded |
entry_mode | Break/retest/second-break/midpoint modes. | 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/London/Asia windows. | not covered: runner is NY 09:30-12:00 only | not yet coded |
htf_bias_override | Key-level/OB counter-trend override. | not covered: counter-bias CRTs are always skipped | not yet coded |
skip_monday | Monday filter. | not covered | not yet coded |
red_folder_filter | Red-folder/no-news gate; no data on box. | not covered: no calendar data on box | not yet coded |
ltf_chop_filter | 1m/2m chop veto and mode selection. | not covered | not yet coded |
range_cap | Avoid too-large C1 ranges. | not covered: sweep_core atr0.5 is a minimum-range filter, not a cap | not yet coded |
first_ny_with_trend | First NY trend trade selector. | partly covered by run_nitro_crt_a_rebreak_bt.py (09:30-12:00 window + 15m EMA bias, but it takes any qualifying CRT up to 2/day, not only the first); still marked not yet coded pending test | not yet coded |
stop_modes | Box-edge/other stop modes; video stop unspecified. | partly covered: runner has one structural stop mode (sweep extreme + min distance); box-edge / other modes not coded; still marked not yet coded pending test | not yet coded |
mnq_economics | MNQ economics. | partly covered: runner has NQ specs only (NQ_POINT_VALUE=20, NQ_COMMISSION_PER_SIDE=2.50); no MNQ/MES/MGC; still marked not yet coded pending test | not yet coded |
target_front_run | Optional target offset. | not covered | 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