Research
Two kinds of writing live here, and they are kept apart on purpose. Measured research takes a claim traders act on and puts it against real one-minute futures bars. Three studies so far. Rule breakdowns explain how an evaluation rule works in plain English. The second kind carries no market data and never pretends to.
The studies
The method under all of it Where the numbers come from, and what we got wrong. Every figure here is produced by a script you can download. Read it →A claim traders act on, put against real one-minute futures bars. Every figure on these pages was computed, and the code and the sample size are published beside it.
Every cyan number on this page came out of a measurement.
Study 01 measured the range of a candle and said plainly that it is not the chance a stop gets hit. This measures the thing itself: the distance price runs against an entry, and which of a stop and a target is reached first.
56.8%of entries run 10 points against you within five minutesA 10 against 20 race goes to the stop 65.1% of the time · 511 sessions · 197,246 entries Read the study → MeasuredStudy 01 · v1Cutoff 17 Sep 2026The claim: a 10-point stop on NQ is tight. We measured the high-to-low range of every US cash-session 5-minute candle, split it by hour, and published the full distribution.
28.00points, the median high to low range41,796 candles · 548 sessions · full distribution published Read the study → MeasuredStudy 02 · v1Cutoff 17 Sep 2026The claim: the first hour decides the day. It is half right, one edge is usually in early, but both edges almost never are, and the closing hour matters more than the saying suggests.
80.8%of sessions set one edge of the day in the first hourBoth edges in the first hour: 3.5% · every half hour block published Read the study → MeasuredStudy 04 · v1Cutoff 17 Sep 2026Most of the day’s high-low range exists early, which is not the same thing as the day being over. Both were measured separately, because reading the second off the first is how the claim goes wrong.
71.3%of the median session’s final range already existed at 11:00Sessions actually finished by then: 9.0% · every half hour block published Read the study → MeasuredStudy 05 · v1Cutoff 17 Sep 2026A session’s daily range and its open-to-close move get quoted as if they were one number. They are not, and the gap between them is large and ordinary. Session OHLC only, so the figure does not depend on bar size.
47.5%of its range separated open from close, on the median NQ sessionSessions under a quarter of their range: 22.9% · full distribution published Read the study → MeasuredStudy 06 · v1Cutoff 17 Sep 2026Counting streaks without a benchmark is the trap, so this counts them twice: once on the real sessions and once on a fair coin over the same number of sessions. The two are hard to tell apart.
20 vs 16.3runs of five sessions or more, measured against what a coin givesLongest run NQ produced: 8, against a coin’s typical 9 · every length published Read the study →How one prop firm rule actually works, taken apart on its own. These are explanations, not measurements: every figure is either arithmetic shown on the page or a number you type in yourself.
Nothing here is measured, so nothing here is cyan.
They are legal, disclosed and buried. Shown as one trading day where nothing goes wrong and the account still fails, then written again the way a firm writes them when it wants you to pass. Check your own firm against them in ten minutes, free.
Includes an eleven question check on your own contract Read the long read → DrawdownExplained6 min readA trailing drawdown follows your account upward and does not come back down. Worked through on a $50,000 account with $2,000 of room.
$2,000 of room becomes $800, on a day that made money Read the breakdown → Daily limitExplained6 min readWhat a daily loss limit actually caps, the three points it can count down from, and the moment it quietly replaces the stop on your chart.
Same $1,000 limit, three different places it starts from Read the breakdown → ConsistencyExplained5 min readA cap on how much of your total profit may come from your best day, and why the base it is measured against changes the answer completely.
One percentage, three bases, three different ceilings Read the breakdown → Account sizeExplained4 min readThe number on the label is not the money you can lose. The drawdown rule decides how much the account can absorb, and it is a fraction of the headline.
A $50,000 label with $2,000 of loss capacity Read the breakdown →What we publish
A study goes up only if it does at least one of three things. Everything else is somebody else's blog post.
And a study stays up with its limits attached. Every one ends with a section on what the measurement does not prove, because a number without its boundaries is just a more confident guess.
The data
All studies use the same source: trade ticks exported through Quantower for XCME-listed instruments, aggregated into one-minute bars by our own code, then restricted to the US cash session, 09:30 to 16:00 New York, unless a study states otherwise. NQ and ES trade on CME almost around the clock, so that window is a choice rather than the instrument’s own trading hours. The chain, the audit, and where the chain stops.
| Instrument | Bars | Period | Sessions in the window | Complete sessions |
|---|---|---|---|---|
| NQ | 747,455 | 23 Jun 2024 to 17 Sep 2026 | 548 | 511 |
| ES | 716,615 | 14 Jul 2024 to 31 Jul 2026 | 523 | 493 |
Every study is frozen to a declared data cutoff and versioned; figures are revised as a new version, never silently. Rows sharing an identical bar timestamp are collapsed, keeping the first occurrence. Front-month bars as recorded by our data pipeline; contract rolls are not back-adjusted, which does not affect intraday range or intraday timing measurements. Two session counts are given because the studies need different things. A study of candle ranges counts every session in the window, since a five-minute candle does not need a complete day around it. A study of when the session high or low is printed counts only complete sessions, meaning all 390 minutes from 09:30 to 15:59 are present, checked against the expected grid rather than by counting rows.
The risk check takes your account and your stop and puts them next to the measured range of a candle on the instrument you trade. It is the first chapter of a paid kit and it costs nothing.
Research notes
Four short emails to start: the free risk check, the 28-point finding and what it changes about stop placement, one prop firm rule trap with the arithmetic shown, and what the paid playbook adds that the free material does not.
DOUBLE OPT-IN: WE SEND ONE CONFIRMATION EMAIL FIRST. THE TOOL OPENS AFTER YOU CONFIRM.
NO SPAM · UNSUBSCRIBE IN ONE CLICK · WE NEVER SELL OR SHARE YOUR ADDRESS
AFTER THE FIRST FOUR EMAILS WE ONLY WRITE WHEN THERE IS A NEW STUDY, TOOL OR RULE BREAKDOWN.
The free risk check, and one line on what the lab does.
The 28-point finding, and why the median range changes how a stop reads.
One prop firm rule trap, with the numbers worked through.
What the playbook adds that the free material does not.