Edgeful Playbook

01Summary

Learning path

A suggested study sequence built from the cited lessons. The drills and readiness criteria are editorial exercises, not an edgeful course or a profitability test.

02How to read edgeful data

Rules all the hosts state or share. The speaker is named where it matters.

• A frequency is not a win rate. "IB high breaks first 75% of the time" describes past days. The trade still needs an entry, a stop and a target that fit that frequency. André: "None of what I'm saying is my discretion… it's all stats", but he still chooses the R:R. 20260929 01:09:01 20260929 01:09:51
• The session is a filter. A report only counts what happens inside its session. An IB break or gap fill after the close doesn't count, and moving the close (4:00 vs 4:15) changes the previous close and so the gap. Match TradingView's session to edgeful's: TradingView's futures session runs to 6:15 p.m., edgeful's New York session ends at 4:00 p.m. 20250227 06:17 20250227 09:41 20260520c 55:25
• Check the sample size. Dan always opens "view details". Zinc dismissed an 83% gold stat with only 6 days behind it; readings like 33%, 66% or 100% usually mean three or fewer data points, so open the table. 20261003 18:29 20260915 29:10 20260723a 01:58:23
• Check the regime. André compares 1 year, 6 months, 3 months and 1 month: IB held about 80% at every lookback (edge intact), while gap fill fell from 70–74% to about 50% (out of favour). He uses at most a 2-year lookback; 5 years hides the current regime. 20260728b 01:25:00 20260429b 17:55
• Filter by weekday. "Pick and choose your weekday data." The same IB report can give about 70% on a Friday and about 50% on a Wednesday for gold. André: the by-size and by-weekday subreports "can have a drastic impact on the data". 20260917b 01:40:32 20260918b 25:42 20250227 13:42
• …but don't over-refine. Zinc widened a query resting on 4 data points to a year, and uses weekday filters for IB direction, not IB extensions ("over refining"). 20260904b 01:37:22 20260917b 01:50:55
• A coin flip means no trade. 46/46 or 50/58 splits are skipped; Dan wants roughly 65–80% before acting. 20260930b 08:04 20260923b 01:41:42 20260918b 08:41 20261003 15:42
• Very high numbers revert, and stacking can mislead. High stats revert to their average, often with losing streaks. Zinc took a 100% stat and lost. On 22 July a stack of strong bearish stats all failed, and he concluded that "using a bunch of strong confluences is actually a negative". 20260915 20:25 20260918d 01:01:03 20260723b 49:20
• Lookback changes the number. James's NQ first-hour continuation reads 75/77% year-to-date but 70–71/73–75% over 6 months. Quote the setting with the number. 20260904a 06:24 20260923a 07:16
• Read the ETF, trade the future (Zinc). He reads SPY/QQQ reports because the futures reports carry extra "phantom" data points and the ETFs reach levels first. The exception is the Globex IB, which uses the MNQ report. 20260527a 40:33 20260514 01:29:43 20260527a 01:38:10
• Event days override the data. On FOMC days Zinc expects levels to break 3–5 times, so inside-day and similar stats don't apply. 20260917b 36:41
• Ask about failures. André asks edgeful AI to list every day a setup failed and "find commonalities". One result: an ES IB under 20 points led to a double break 43% of the time. 20260225a 54:30 20260225a 51:40
• Verify it yourself. James tells viewers to check every stat on edgeful rather than copy his. 20260921a 57:18 20260911b 01:40:39

03The reports they use

What each report measures, mostly in André's masterclass definitions, with readings the hosts quoted. Readings are snapshots; check today's numbers yourself.

The daily bias template that powers the screener combines four reports: opening candle continuation, IB, IB by rejection and previous day's range. In André's worked example all four were bullish: he waited for the IB to form at 10:30, bought the pullback with a stop below the lows, and made the IB high his first target. "If the mega-caps are all bullish… over the course of the next hundred trades, you're going to be on the wrong side of that data" if you short. 20260225a 19:48 20260225a 23:13 20260225a 25:14

04James: daily bias from levels

The bias stack

1. Previous day's high or low decides the side. "Above PDH, look only for long continuation; below PDL, only shorts." He uses extended-hours levels including wicks; Zinc uses regular-hours levels. 20250312 06:37 20260915 01:44:22
2. The opening candle confirms it. The first-hour version is where the report "gets really good". A PDL break plus a bearish first hour is his A+ short. His simplest summary: trade in the direction of the first 30–60 minutes. 20260918c 04:36 20260910a 02:16 20260505 07:42
3. The screener breaks ties. His "JB daily screen" runs opening candle continuation, IB, IB by rejection and previous day's range across the 3 US indexes and the Mag 7, read after 30 minutes and again after an hour. All bullish means longs on pullbacks; all bearish, shorts; mixed, maybe no trade. He wants 75–80% one way, not 50/50, and says he lost money mainly on the days he went against it. 20260922c 08:43 20260730a 01:20:42 20260505 19:10
4. Swings use the week. Previous-week high/low and the opening week range give the swing bias, filtered by the 5-day SMA; the EMAs only describe the trend. 20261002a 08:28 20260905 21:33

When signals disagree he trades less or takes profit early. He cut an NQ long at about +37 points because PDL had been taken while the opening candle was bullish. He also stayed flat on a long he liked once the ORB low broke and the opening candle turned bearish: the data overrode his read. 20260910a 02:52 20260929 50:00

Entries

• Break and retest of the opening-range high or low in the bias direction (his usual entry, so an unbroken 15-minute ORB worries him). 20260909a 07:28 20260723a 01:14:14
• Double break back inside. On a long-bias day, if the opening-range low breaks, he goes long once price is back above the New York session VWAP, targeting the opening-range high. Reviewing a YM loss, he said this was the right second entry, not a pullback in the middle of the range. 20260909a 09:51 20260911a 02:42
• VWAP pullbacks. VWAP is his entry tool: the New York session VWAP, the daily VWAP, and a VWAP anchored at 6 p.m. ET. His "bread and butter" is a wick into VWAP and taking the bounce. For longs he buys whichever index is above its New York VWAP. 20260921a 57:18 20260929 07:52 20260731b 49:54 20260527b 01:15:54
• After a big trend day, don't buy the breakout. Wait for the pullback into the level that broke, or to VWAP or the 25% IB retracement. At all-time highs, "buy pullbacks". 20260922c 05:52 20260923b 57:29 20261003 32:44
• Wait for the open to settle. After a losing ES long on 27 July he said his best window is after 10:00 ET and he should have waited for the first 15 minutes. He stays out while price is still inside the opening range. 20260728b 01:00:24 20260924 01:11:53
• He won't short into a level price bounced above at the open, and he doesn't enter with limit orders except for overnight swing fills. 20260910b 01:17:36 20260912 28:06

Stops, targets and management

• The stop goes where the trade is invalidated. He prefers a wide stop for about 1R over a tight stop for 2–3R: 20-point NQ stops are something "I sucked at". CFDs let him fix the dollar risk at any stop width. 20260903a 01:15:40 20260728b 26:13
• Scale out heavily at the first level, then trail. Examples: 75% off at the previous week's high on an NQ swing; 75% off at +33 points, then the rest at the gap fill for +67, on an NQ short. 20261002a 02:38 20260918b 30:25
• Targets are levels: opening-range high/low, premarket high, PDH/PDL, gap fill, daily VWAP. 20260910b 01:23:39 20260922c 05:52
• He cuts when price reclaims the session VWAP against him, or when nothing happens soon after the IB. His own statistic: day trades held longer than 30 minutes usually end red. 20260723a 01:12:22 20260723a 01:18:48

Swings

The opening week range drives his swings. On NQ, once the Monday–Tuesday high broke, the other side had only an 11% history of breaking. He bought the pullback to the weekly VWAP with a stop about 300 points away, took 75% off at the previous week's high and trailed the rest (claimed +500 points). He also front-runs the week's high once the Monday–Tuesday range is set and price is above the weekly VWAP. A TSLA swing came from the streaks report: three large red daily candles, more than the average streak, so he bought the bounce and took 80% off at the opening-week high. 20261002a 03:39 20261002a 02:38 20260723b 52:25 20261003 11:17

Risk and routine

• About $500 risk on day trades and $1,000+ on swings, kept small because copy-trading multiplies it across 5–6 accounts. He keeps his average losing day close to his average winning day. 20260914 31:52 20260507a 01:16:29
• Daily loss limits with an automatic lockout, used for four years (his "maximum pain threshold"). At most two trades a day. He never moves the stop, an old habit he broke. 20260905 29:24 20260921a 01:28:30 20260909a 06:04 20260923a 15:54
• "Day trading is not everyday trading": "cash as a position" when he doesn't see his edge; caution on Mondays (his worst P&L day, though Mondays close green about 65–69% on NQ); "big, small or not at all" before FOMC. 20260723a 47:46 20260914 18:06 20260728b 04:35 20260917b 50:31
• Focus on 2–3 tickers and 2–3 reports; no setup, no trade. From May he traded only ES/NQ. He prefers the lagging index over one at highs. 20250312 01:16 20260505 09:00 20260922a 34:53
• Routine: 8:00 a.m. ET prep marking levels, wait for the opening range and VWAP, then check the screener and stocks. He journals in Tradezella (700+ entries). 20260914 01:29:37 20261003 13:15
• Claimed stats for 2026 (his journal, unverified): 84.35% win rate, 0.43 average R, profit factor 2.33. That combination means many small wins and fewer, larger losses. 20260527b 01:16:30

05Mr Zinc: cherry-picked initial balance

Pick the day, not just the setup

• He filters IB by rejection by weekday and trades only strong combinations. Wednesday has been the best NQ day and Tuesday the worst (about 50/50, and Tuesday has the highest IB double-break rate: 28% vs 3% on Wednesday). On Thursday he trades only if the IB low forms first (about 81% to take the high). 20260911b 01:09:02 20260519 11:39 20260730a 56:43 20260918d 18:55
• Crude on Wednesdays is his main income ("90/90 stats"). CL IB by rejection reads about 88–90% on Wednesdays (20 wins / 3 losses over 6 months) against roughly 53/33 on Thursdays. Commodity IBs follow the pit open: crude 9:00–10:00, metals 8:20–9:20. 20260918d 59:34 20261001a 06:45 20260911b 01:11:14 20260603a 01:11:00
• Gold: only on specific days. He dropped the gold ORB double break because every variant needed a huge stop or a late entry, and he doesn't trade gold in the New York session. 20260922a 01:19:30 20260903a 01:12 20260506 10:56
• SPY and QQQ must agree. When their IBs diverge, both play out correctly only about 17% of the time. He either skips, or waits for one to break and trades the other. He doesn't add a trade on an index that has already reached its level. "SPY tells the truth, QQQ lies." 20260911b 01:09:13 20260506 01:30:30 20260519 01:57:01 20260429a 01:10:51
• Daily bias comes from inside bars (the main one), outside days and gap fill by size. An inside day means targeting the nearer of PDH/PDL; an open above PDH is an outside day that tends to come back. The IB overrides that bias once it forms. 20260514 01:10:42 20260519 15:18 20260722 19:10 20260429a 23:35

The IB trade, and how it changed

• May–August version: wait for the IB, put a limit at the IB midpoint (IB50), target the IB extreme, stop at the opposite extreme (about 1:1), no second target. Cancel if the target trades first. He prefers the 50 to the 25 ("Dan's trade") because 25 entries get stopped on the pullback to the 50. 20260514 10:38 20260506 01:48:45 20260430 01:13:09
• Timing rules: front-run the IB close by at most about 7 minutes (no entry before about 10:23 in New York); in the Globex and Asia sessions front-running is fine. No entry after 11:00, which in his optimizer backtest raised the profit factor from 2.3 to 2.47. Never the first 15 minutes (60% of his losses came from them). 20260506 01:24:09 20260514 06:00 20260514 01:48:30 20260904b 18:25
• One IB a day. Once the target is hit he's done: no runners, no retracement re-entries, because no statistic says the IB re-breaks after a retrace. 20260513 16:33 20260519 50:15 20260514 01:52:27
• IB ending zone: a close high in the range after the high formed first flips the odds, so a close in the 50–75% zone means no trade. In May–June, Wednesdays repeatedly closed near the top of the IB and broke at 10:31. 20260722 01:05:26 20260527b 09:03 20260520c 01:02:54
• Choppy-market version (September): same bias, but orders 10–20 points before the IB (or 5-minute ORB) high/low with more size, targeting the level or the 0.2 extension, accepting slightly negative R:R at the same dollar risk. 20260915 15:01 20260930b 01:16:10 20261001a 19:51
• Extensions: "Wait for the IB to break. Play the extension." The 0.2 extension is usually reached (80–90%). "Anything past 0.2 is just lucky." 20260930b 01:51:49 20260930b 08:04 20260915 01:14:51
• "There's no forever model." The IB data stays the same while the entry model shifts with roughly two-week sector rotations. He calls "edge decay" claims a failure to adapt. 20260904b 13:15 20260930b 01:28:21
• Other sessions: Globex IB 6–7 p.m. ET and Asia IB 8–9 p.m., from the MNQ report with the session set to daily and the next weekday's stats. He never carries a Globex trade into the Asia IB. He also tested a 2–4 p.m. "PMIB". 20260527a 01:38:17 20260514 08:44 20260731b 01:29:47

Gaps, news and context

• Gap fill by spike gives a level: the average run away from the gap before the fill. He buys the bounce there. Gap reports are context, not an entry model. 20260923b 08:36 20260904b 16:34
• The 8:30 news candle is the "data wick": whichever extreme is hit first, the other usually gets swept. He avoids no-news days, doesn't trade before NFP, and does "terribly" on FOMC and pre-NFP days. 20260911b 03:43 20260910b 05:42 20260507b 04:05 20260917b 13:52
• FOMC flip (his concept; the 80–83% figure is from an outside source, not edgeful): never trade the 2:00 p.m. release. Take the opposite of the day's close direction overnight, from the 6 p.m. Globex open, targeting the 2:00 p.m. high/low. He trades it with QQQ/SPY options or a few MNQ on a personal account, with no stop, and shows no backtest. 20260917b 07:24 20260731b 09:11 20260918d 33:37

Risk

• Size from the drawdown. About 12% of the account's drawdown per trade ($120 on a 25K, $240 on a 50K); contracts = target risk ÷ stop in points × $ per point. MES carries about half of MNQ's risk. NY ranges pushed him from 25K to 50K accounts. 20260507b 01:28:45 20260527a 30:36 20260730a 30:02
• No trailing, no partials. Moving a stop to breakeven or trailing it is a new trade risking open profit; if you feel the urge, close instead. "See money, take money." No daily profit target. 20260923b 01:33:24 20260507b 01:50:07 20260909b 01:55:37 20260924 16:32
• Payouts: take the maximum as soon as you qualify (he blew five accounts chasing more); avoid intraday-trailing-drawdown accounts; "prop firms aren't the end goal". His personal account grew from $2K by compounding, and its margin limit caps a tilt day. 20260909b 48:24 20260507a 01:23:11 20260723a 01:29:28
• Profit factor is the only stat he ranks strategies by; expected value matters most, but prop traders rarely have enough trades to measure it. His losses come from psychology ("boredom trading"), not the data. 20260514 18:53 20260918d 01:03:34 20260722 01:47:03

06Dan: VWAP scalps with data filters

• Three-way agreement: he trades only when the edgeful stat, the screener and the side of VWAP all agree. Bias order: VWAP side, then screener, then how the IB formed. 20260905 07:22 20260909b 01:20:09
• His A+ setup: price on the right side of both the daily and New York VWAPs, no candle closes through them, VWAP sloping his way, and a limit order at IB25. A flat VWAP with closes on both sides means no trade. 20260801 53:14 20260728a 01:25:59 20260513 01:05:28
• Skip list: VWAP beyond the IB50 (his invalidation would be too far away); an IB too wide for his risk (a 646-point IB on one micro is more than half a 25K drawdown); the previous session's high/low just taken; price pinched between the VWAPs; chop through VWAP, which he calls the most important. 20260801 56:28 20260801 57:43
• The screener keeps him out. "The screener doesn't give you an entry… but helps you stay out of crappy longs." 20260905 01:00:25 20260905 59:40
• Thresholds: IB setups only at 70–80% (his losses come at 55–60%); gap-fill trades only at 65%+. Go with whichever ticker has the strongest data. 20260918b 08:41 20261003 15:42 20261002b 01:53:30
• Execution: a limit at IB25 placed about 10 minutes before the IB closes. The IB75 is only an emergency stop: he exits manually on a candle close back through VWAP, "the only… major decision". If the IB high breaks without filling his order, he cancels and walks away. 20260527a 01:15:24 20260527a 01:25:24 20260527a 01:17:02
• Targets short of the level: about 0.3–0.5R, because price keeps turning just before the IB target; the IB high is "a draw, not his target". Winners finish in 1–3 minutes. Negative R:R means you must "be the best loser"; newer traders should use 1:1. 20260909b 01:04:16 20260912 01:03:13 20261003 08:50 20260527a 01:02:37 20260728a 01:59:52
• Discipline: "That's my level… that's my edge. I won't deviate." One trade a day to avoid tilt, no entries after 11:00, and about 2–3 trades a week; more trades has meant less money. 20260918b 01:02:29 20260519 02:09:28 20260516a 01:22:02 20261003 38:56
• Sizing and prop: a fixed risk of about $150 aiming at $150 base-hit days for Topstep's payout rule, micros sized from the IB's width, copied across many funded accounts. He judges results by the quarter. March came out barely green from overtrading, and April made double with half the trading days. 20260801 01:26:41 20260527a 01:13:06 20260513 01:23:12 20260728a 01:50:37

07André: conditional probabilities

• Chain conditions into targets. NQ gap fill: the 25% level is hit 87% of the time, and given that, the 50% level 80%. So the first trade targets 25%, with a second trade or runner to 50%. For gaps over 1% the full fill after a half fill drops to 36%. 20260929 04:09 20260918b 31:04 20260723b 59:57
• Use the spike for the stop. Gap fill by spike gives the average adverse run before the fill; the stop goes beyond it. 20260929 06:35
• Cross-market questions. When NQ took its prior-session low, ES also broke its low about 70% of the time. Together with the inside-bar figures, that made the ES PDL his target. He worked through a short with a stop above, about 1.8R, and the level was hit. 20260929 55:07 20260929 01:09:51 20260929 01:15:34
• Report levels replace subjective zones. He builds targets from report levels (outside-day magnet, gap fill by size, inside bars, the midnight open for the runner) and enters on pullbacks to the ORB high or midpoint. 20260505 04:09 20260505 20:52 20260505 53:46
• Target a few points before the level. He gave an example of an algo trade that missed its target by one tick and became a full loss. He targets 40–60 NQ points and accepts 1:1 or less; "no more runners". 20260929 01:14:44 20260516a 01:36:06 20260507a 49:47
• Risk: about 2% per trade, measured against the prop account's drawdown, not its size. At 20% risk per trade, five straight losers end the account, and five in a row within 100 trades is likely even at a 60% win rate. "There's no strategy on the planet that outperforms poor risk management." 20260225a 55:55 20260225a 58:50 20260921a 01:29:42
• Options selling with stats: IB low first, green opening candle, IB high and PDH broken, then sell below on a pullback to IB25/IB50. At most 2–3 a week, never on FOMC, CPI or earnings days. He claims an 80–90% win rate with a bad R:R (about 1:10), so a spread never goes to max loss. 20260918b 01:20:58 20261001a 01:37:39 20260730a 01:22:23
• His no-data trades, labelled as such: fading giant one-minute candles ("zero data behind this") and the "10 a.m. wiggle". He sets a stop immediately on fades; one was stopped in about 35 seconds. 20260505 47:11 20260520c 01:05:29 20260520c 01:11:40
• Algos: paper trade first, re-optimize monthly, walk-forward test with a hold-out. The ORB algo is the most consistent (about 1:1). 20260921a 48:49 20260921a 51:13 20260728b 01:31:04

08Where they agree and disagree

Live disagreements worth studying:

• Monday 28 Sept: James's "Mondays close green" stat pointed long; André's prior-day-low read pointed short. André's played out, and James cut his YM long. 20260929 10:18 20260929 55:07
• 22 July: Zinc stacked bearish stats while James traded NQ longs. James was right: the gap filled and ES took PDH. 20260723b 46:40 20260723b 40:33
• 18 Sept: Dan wanted longs on 75%/71% stats while James and André leaned short; NQ went short. 20260918b 01:06:25 20260918b 20:13

09Unknowns and caveats

• Speaker attribution. Captions have no speaker labels; the notes tag uncertain lines "(speaker?)". This guide uses only statements whose speaker is clear, or names the doubt.
• Percentages drift. The same report is quoted with different numbers across days and lookbacks, and captions garble digits. Treat every number as "about".
• Methods change. Zinc's entry moved from the IB midpoint (May–August) to orders before the extreme (September). Dates matter when you copy a rule.
• Contradictions inside one trader:
  • Zinc says he doesn't use VWAP, but agreed to "IB and New York VWAP" in another stream. 20260917b 03:27 20260915 23:41
  • Zinc says never to trail, yet trails scalps aggressively and often exits early. 20260909b 01:49:26 20260909b 01:05:20 20260430 01:50:59
  • James's ORB fade, taken against the break, sits next to his IB use, taken with the break. 20261002a 07:53 20240912 00:29
  • André says "no more runners" but later suggests a small runner. 20260507a 49:47 20260507a 01:41:44
• Option rules conflict: André never closes a sold option at 50% profit; Zinc closes at +50% in one day. 20261001a 01:52:55 20261001a 59:52
• No-stop trades. Zinc's FOMC flip and some overnight holds have no stop and no shown backtest. Don't copy these.
• Results are claims. Payouts, profit factors, win rates and optimizer backtests are as stated on stream. The interview title says "$20,000/month", but the captions only show a +$11k and a −$8k month. 20250312 16:13 20250312 23:29
• Coverage. See the source library: every downloaded transcript is listed, and those without notes are marked "Review pending" and not used here.

10Glossary

James: day-trade checklist

Review each condition against your chart and written plan.

Mr Zinc: IB checklist

Review each condition against your chart and written plan.

Dan: scalp checklist

Review each condition against your chart and written plan.

Source library

243 downloaded videos of 1592 in the channel catalogue; 61 reviewed in full; 182 reviews pending. "People" lists who the title names; who said what inside a multi-person stream is decided in the notes, and an attribution caveat flags doubt.