Tradeflight Playbook

Start here

The order the work happens in. Each step has its own tab.

DailyMap

Yesterday's high and low, big wicks, old turning points, news on the calendar.

4H / 1HBias

Up, down or unclear. How the lines are stacked. How far price has already run.

1HLevels

Session highs and lows. Which ones are taken, which are still waiting.

5mEntry

The trigger, the stop, the targets. Only if everything above still agrees.

1mZoom

A magnifying glass for the 5m. Room for a stop, small wicks left to take.

RULE 1

Judge on the close

A wick through a level is a test. A candle body closing through it is a break. Never decide mid-candle.

RULE 2

No destination, no trade

If you can't point at where price is going, there's nothing to aim at and nothing to measure the risk against.

RULE 3

No stop level, no trade

A stop is the price that proves the idea wrong. Not a dollar amount, not a round number.

How the tools get weighted Suggested draft

A starting point only. Carlos's full weighting gets explained over time and replaces this.

Always on

The 14 / 50 / 200 and how price behaves at them
Candle closes (wick vs body)
Structure: trend or range

Sets the context

Higher timeframe direction
Levels: session highs/lows, yesterday's high/low, the 9:30 opening range
Time of day: NY open, 9:50 to 10:30, candle closes, news

Adds weight when present

Sweeps and equal highs/lows
FVG, order blocks
Exhaustion wicks, double tops, engulfing

The weights move with the day. On a trend day the lines carry most of the weight. On a range day the levels and edges do. Between 9:50 and 10:30, expect a grab of the nearby highs or lows before trusting the direction.

How sure are we?

Textbook Standard, widely documented.

Seen live Watched it happen on NQ in our sessions.

Working idea A good guess still being tested.

Tested Checked on NQ history. Numbers in the Evidence tab.

From BDH Taught by D-Money (BDH Trading) in videos you follow.

From TJR From TJR's "Seek and Destroy" idea.

How to read the test numbers

Most numbers answer one question: from the signal candle's close, did price move one "normal candle size" (1 ATR) your way before it moved one against you?
A coin flip is about 49%. 55% or more with a decent sample leans useful. Under 45% means the opposite tends to happen first.
n is how many times it happened. Under 20 is a hint, not proof.
New words are explained in the Definitions tab.

Map: the daily picture

Before looking for a trade, mark where price has reacted before and what's on the news calendar.

Mark these first Textbook

Previous day high and low (PDH / PDL). Watched by nearly everyone.
Long wicks on the daily and 4H. Price went there and got pushed back.
Old turning points (swing highs and lows) and zones that held before.
Round numbers. People pick tidy prices for orders more often than they should.

What NQ did with yesterday's levels Tested

On 90% of days price took out yesterday's high or yesterday's low. Each one on its own was about half the days.
The first time price poked through yesterday's high, the candle usually closed beyond it (61% of the time), so it broke rather than faked.
Going with that break close worked better than fading it. Details in the Levels tab.

News and economic events

You don't need to forecast the news. You need to know when it lands.

Price moves on the surprise Textbook

Every release has a forecast that's already in the price.
The move comes from the gap between the forecast and the real number, not the number itself.
A bad number that was expected to be bad can still rally the market.

The big ones (ET) Textbook

8:30 CPI, PPI, jobs report (NFP), weekly jobless claims, retail sales, GDP.
10:00 ISM, consumer sentiment, some housing data.
14:00 FOMC rate decision, then the press conference at 14:30.
Check the calendar before the session, every day.

How to handle it Seen live

Spreads widen and price can jump past a normal stop in one tick.
Many funded traders sit out the release, then trade the direction once it settles.
Hot inflation usually means lower Nasdaq and lower gold. Tech moves harder on rate news than the S&P.

Event	Nasdaq (NQ)	Gold	Dollar
CPI hotter than expected	Down	Down	Up
CPI cooler than expected	Up	Up	Down
Hawkish Fed (rates higher or longer)	Down	Down	Up
Dovish Fed (cuts or easier)	Up	Up	Down
Strong jobs data	Depends on what the market fears	Down	Up
Geopolitical shock	Down	Up	Up

Tendencies, not rules. What the market already expected decides the reaction.

Bias: which way, and how far already

On the 4H and 1H. Up, down, or honestly unclear.

Structure Textbook

Uptrend: each high is higher, each pullback stops higher.
Downtrend: lower highs and lower lows.
Range: bouncing between the same two levels.
A break of that pattern is information, not an order to trade.

How the lines are stacked Textbook

Red on top, Gold middle, Purple bottom, all pointing up = uptrend.
Purple on top, Gold middle, Red bottom, all pointing down = downtrend.
Tangled and flat = range or changing over.

Different timeframes, different answers Textbook

Up on the 5m and down on the 4H at the same time is normal.
A sharp 5m drop can be an ordinary pullback on the 4H.
The higher timeframe says where price should stop. The lower one says when it has.

Trend day or range day?

Decide this early. It changes which tools get the most weight.

Trend day Textbook From BDH

Looks like: lines spread apart and pointing the same way, price holding one side of the 14, clean drop / pullback / drop (or the mirror), the opening range breaks and stays broken.
Trade it: pullbacks to the 14 or 50 with the trend, stop past the wick, targets at the old highs or lows to the left. Let winners run and trail.
Don't: fade it because it looks stretched.

Range day Textbook From BDH

Looks like: lines flat and tangled, candles overlapping, price crossing the 14 back and forth, the opening range breaks and falls back inside.
Trade it: the edges of the range (sweeps of the high or low that close back inside), targets in the middle or the other edge. Smaller size, quicker exits.
Don't: take crosses. In chop they whipsaw (our tests agree).

How far has it already run?

Measured as distance from the Red 14, in normal candle sizes (ATR).

StretchSeen liveTested

Stretch from the Red 14

What: How far the candle closed from the Red 14, measured in ATR.
Then: A little stretched usually comes back to the line. A lot stretched usually means a strong push that keeps going.

Distance above Red 14 (15m, NQ)	Comes back to it within 10 candles	Going with it
Sitting on it (within about 20 pts)	98% · 851	-
Medium: about 60 to 105 pts	70% · 350	50% · 341
Far: more than about 105 pts	35% · 105	64% · 100
Far below (105+ pts under)	61% · 71	47% · 70

How to read: out of all the times price closed that far above the Red 14, the middle column is how often it came back to touch the line within 10 candles; the last is how often going with the move reached one candle size of profit before one candle size of loss. Point ranges are based on a normal 15m candle of about 42 pts in this period; on fast days they stretch. The Tradeflight MTF Panel shows the live distance. Surprise: a big stretch up was a sign of strength, not a pullback warning. A medium stretch is where the snap back usually happens.

When timeframes disagree

Why more than one trade can be right at the same time, and why the big target sometimes never gets hit.

Nearer lines get respected first Seen live

A 4H long can be right, and still stall at the 1H Gold 50 sitting overhead.
The line closest to price on a bigger timeframe is both a magnet and a wall.
Plan to take some off there, and only expect the bigger target once that line is closed through.

Two trades, two timeframes Seen live

A 5m short can work inside a 1H uptrend: it's the 1H pulling back to its line.
That short has a short life. Its target is the 1H line, not a new low.
Once price reaches the bigger line and the small timeframe flips back, the bigger trade is the one to take.

When to just wait Textbook

Timeframes pointing different ways with price in the middle: no clean trade.
A bigger line right in front of your entry: little room, poor reward.
The wait is usually for price to reach a line where the timeframes agree again.

The higher timeframe candle

What a 1H or 4H candle shape says, and what it doesn't.

Use it as context Seen live

A red 1H or 4H candle after a run of green ones is often the first pullback, not a reversal.
Lower timeframe strength inside that red candle is a repair attempt. Whether it worked shows at the close.
Don't trust a 5m breakout in the last minutes before a 1H or 4H close.

Not a next-candle predictor Tested

Red 1H with a long lower wick: the next hour was green only 43% of the time (n122).
Red 1H closing at its low with a full body: next hour green 56% (n103).
Baseline: 51%. So the candle shape alone doesn't call the next hour. Use it with levels.

Levels: session highs and lows

The places stops pile up, and where price tends to go next.

Asia20:00 to 00:00

London02:00 to 05:00

NY AM09:30 to 11:00

NY PM13:30 to 16:00

New York time. These match the killzones on your chart.

The 9:30 opening range From BDH

Mark the high and low of the first candle(s) after the 9:30 ET open.
A break and close below the range low is a short trigger; above the range high, a long.
Stronger when price is already on that side of all three lines, or pushes through the lines and gets rejected right back (his example: a 9:30 push above all three EMAs, rejected, then shorts on each pullback).
The other side of the range, or its extreme, is a natural target. Working idea until tested.

9:50 to 10:30: seek and destroy From TJR Tested

The idea: between about 9:50 and 10:30 ET, price often reverses or wicks into the liquidity on the left (the pre-open or opening lows or highs), then carries on in the real direction.
On NQ (15m, 60 days): this window set the NY session high or low on 40% of days, about twice as often as 11:00 to 11:45 (22%).
Uptrend days (price above the 15m Purple 200 at 9:30): the window swept the pre-open or opening low on about half the days. After the sweep, price was higher by 12:00 on 59%, and the sweep low was the low of the NY session 47% of the time.
Downtrend days were weaker: the mirror sweep was almost never the session high (1 of 12). Treat that side as unproven.
Use it: don't chase the first move after 9:30. Watch the 9:50 to 10:30 window for a wick through the nearby lows (in an uptrend) that closes back above, ideally at a line, then take the pullback entry.
Small samples: a lean, not a rule. Send more of TJR's context and it gets refined.

Why the levels matter Textbook

Stops sit just above obvious highs and just below obvious lows.
Big orders need a lot of other orders to fill against, so price gets drawn to those clusters.
Untaken highs and lows are targets. Taken ones have done their job.

Sweep or break? Textbook

Sweep: price pokes past the level, then the candle closes back inside. Stops taken, move rejected.
Break: the candle closes beyond the level and holds.
They look the same for a few seconds. The close tells you which.

What NQ actually did

15m data, 62 trading days, 30 Jun to 25 Sep 2026. Tested

90%of days London took the Asia high or the Asia low

0%of days both Asia levels survived untouched by 16:00

71%of days NY took the London high (66% the London low)

90%of days price took yesterday's high or low

First poke through the level	How often it was a sweep (closed back)	Trade that followed	Result
Asia high	34% · 50	Break close, go long	48% · 33
Asia low	51% · 47	Break close, go short	70% · 23
Asia low	-	Sweep, go long	46% · 24
Yesterday's high (PDH)	39% · 31	Break close, go long	61% · 18
Yesterday's high (PDH)	-	Sweep, go short	33% · 12
Yesterday's low (PDL)	37% · 30	Break close, go short	58% · 19

What this suggests Working idea

Expect at least one side of the Asia range to go during London. It almost always does.
On the first touch of a big level, a close beyond it was more often a real break than a fake. Fading it cold was the weaker trade.
Sweeps look best as the second or later test, or at a zone with the higher timeframe behind you. That combination still needs testing.
Samples here are small (under 35), so treat this as a lean.

The EMA retracement entry

How D-Money (BDH Trading) trades the 14 / 50 / 200. Most of it lines up with what our tests found.

The model, step by step From BDH

Check the 1H and 4H first. Where are the lines, where is price? That sets the direction.
Drop to the 5m. Look for a cross in that direction and a push away from the lines.
Wait for the pullback to the lines. Enter as close to the 14 (or 50) as you can.
Trigger: price breaks the previous candle's high (longs) or low (shorts) off the line, or a candle closes back above / below the 14.
Stop: just past the previous candle's wick, or on the other side of the line. A wider option is beyond the 50.
Target: the old highs or lows to the left, or the opening range extreme. Aim for at least 2 to 1, then trail the stop.

How it matches our tests

Pullback to the 14 in a stacked trend: our best tested setup (15m: 63 of 100, 69 with the higher timeframe). Tested
Don't enter on the cross itself: no edge in our data either. He waits for the pullback after it. Tested
Stop at the wick, not the body: matches (wick stops reached 2x 41 of 100 times, body stops 18). Tested
Wick through the 14, then a close back above = long: matches our "the close above Red 14 is the trigger". Seen live
Breaking the previous candle's high or low as the trigger: new, not tested yet. Working idea

Trending markets only From BDH

The lines are respected when the market trends: drop, pullback, drop, pullback (lower highs, lower lows).
In chop and consolidation they aren't. A messy cross inside a range that never follows through is not a signal.

Spread-out lines = strong trend From BDH

The wider apart the 14, 50 and 200, the stronger the trend, and the better the pullback entries.
Fits our test: a big stretch from the Red 14 was usually strength, not a reversal. Tested

A real trend change holds its first retest From BDH

After a solid cross and a push, the first pullback should hold as support (or resistance).
If that first retest breaks, it probably wasn't a trend change. Get out.

The six crosses

Every pair of lines, both directions. The circle marks where they cross.

Close calls

Two lines meet but don't cross, then pull apart again.

Stacks

The order of the three lines, top to bottom, is the trend at a glance.

What price does at the line

This is where the decision gets made.

Patterns

What happens, then what it means.

Taking a trade, step by step

The questions to answer, in order, before clicking. The Live desk has a ticket with the same seven boxes.

Context. What's the bias on the 4H and 1H? How are the lines stacked? How far has price already run from the Red 14?
Why here. What is price reacting to: a line holding, a level swept, a level broken and retested, a zone? "It looks like support" isn't enough. Name it.
Target, and what it's based on. The last swing high or low, an untaken session high or low, yesterday's high or low, a line on a higher timeframe. If there's none, stop here.
Invalidation. Which close, on which timeframe, proves you wrong? A body close back under Gold 50, a close back inside the swept level, a close under the sweep wick.
Stop. Just past the invalidation, usually under the wick, never under a body. See the table below.
Path. What sits between entry and target? Plan to take some off there.
If it fails. What does the chart become? Often the opposite trade is right there.

Entry, path and targets

Decided before the order goes in.

1 · The entry

The trigger is a closed candle at your level, not a touch.
Best tested entry: the first pullback to the Red 14 once the lines are stacked in your direction (15m: 63%, 69% with the higher timeframe).
Don't enter on the cross itself, and don't chase a breakout candle. Wait for the retest.

2 · Check the path

What sits between price and the target? A 1H line price has respected, an old high, a session level, a round number.
Price has to close through it or bounce hard off it first.
Take some off there, and judge the trade at that spot, not at the final target.

3 · Targets

Nearest first: the last lower high (or higher low), then the middle swing, then the session high or low.
Untaken session highs and lows and yesterday's high and low are natural targets.
If price stalls at the first target and closes back under the Red 14, it was just a bounce.

Where the stop goes

The stop sits where the idea is proven wrong. Where exactly depends on the setup.

Stop optionsTested

Four places a stop can go

What: Under the entry candle's wick, under its body, under the Gold 50, or under the last swing low.
Then: Body stops were too tight and got hit by normal wicks. Wick and swing stops survived.

Where the stop went	Reached 1x the stop in profit	Reached 2x the stop in profit	Typical stop size (NQ points)
Under the entry candle's wick	57 of 100	41 of 100	about 22 usually 11 to 33
Under the entry candle's body	23 of 100	18 of 100	often just a few ticks up to about 14
Under the Gold 50	63 of 100	35 of 100	about 82 usually 45 to 112
Under the last swing low	58 of 100	42 of 100	about 60 usually 39 to 96

How to read this table

We found about 100 past trades on NQ 15m where price pulled back to the Red 14 in an uptrend, and replayed each one four times with the stop in a different place.
"Reached 1x the stop in profit" asks: if your stop was, say, 10 away, did price go 10 in your favour before it hit the stop? 57 of 100 = it did 57 times, and got stopped first 43 times.
"2x" is the same question with a target twice as far as the stop.
Typical stop size is the middle trade's stop in NQ points on the 15m, with the usual range under it (the middle half of all trades). For scale, a normal 15m candle over this period was about 42 points.

What it means

Body stops are too tight. On many entries the body's edge was right at the entry price, so the stop sat a few ticks away. Normal wicks took 77 of 100 out before they reached 1x.
Wick and swing-low stops held up best when aiming for 2x: about 41 to 42 of 100 reached it. Winning 41 times at 2x makes 82 units and losing 59 times costs 59, so over many trades that comes out ahead.
The Gold 50 stop survives most often, but it sits far away, so 2x targets are hard to reach.

Pullback to a line

Stop under the wick of the entry candle or the candle before it, or just on the other side of the line.
Not under the body: price wicks through the line all the time.
BDH: if the 14 breaks and price keeps going, you don't want to be in. A wider version sits beyond the 50. From BDH

Sweep of a low

Stop under the sweep wick itself. That low is the level that proves the idea wrong.
Tested: wick stop 47% at 1R vs 35% for a stop just under the swept level.

Break and retest

Stop back on the other side of the level that broke, past the retest candle's wick.
If price closes back through the level, the break failed. Get out.

Eval plan Your rule

For passing prop evals: small, quick, planned trades. Missing a move costs nothing. An unplanned loss is what fails an eval.

Per trade

Target: 7 pts, taken off the first push. No runners, no moving the target.
Stop goes in with the entry every time (bracket or ATM order). Never enter and add the stop later.
Stop about the size of the target: behind a wick, roughly 7 to 10 pts. If the right stop is bigger, pass.

Per day

Up to 3 wins, then done.
Stop after 2 losses.
No trade to win back a loss. The next one has to qualify on its own.

What qualifies

Entry at a line (Red 14 or Gold 50), with a wick into it that closes back.
In the direction of the 15m. Skip when the 15m and 5m disagree and price sits in the middle.
Clean candles around the entry. Dojis, overlapping wicks and back-and-forth bodies mean the market is unsure: lower confidence, smaller or no trade.

FOMO check Your rule

If price already left the line and you're thinking of getting in anyway, that's the chase. The entry is gone; wait for the next retest.
If you're already in the target zone (where the trade would take profit), that side is done. Only the opposite trade can set up there.
Example, 2026-09-25: short taken at 30,856 after the drop, inside the take-profit zone, instead of on the retest at 30,877 to 30,880. Stopped, then price went the planned way.

When price slices through your entry Seen live

How it came in matters more than how many lines are there. Big red bodies driving into the zone with no slowing down usually go through. Smaller candles and wicks as it gets close usually hold.
A level that just flipped is weak on its first retest after a fast move. Price broke up through the cluster only minutes earlier, so buyers there hadn't defended it yet.
After a sharp pop, the pullback often goes to the line that started the move. That's usually the slower line underneath (for example the 1m 50 EMA), not the fast one price just left.
Map the next line down before entering. If the zone fails, that is where the real entry often is.
Resting limit orders only when price is drifting in slowly. On a fast approach, wait for the wick and close back. No wick, no entry: that rule is what keeps you out when it slices through.
Example, 2026-09-25 11:10: long planned at the 30,845 to 30,856 EMA cluster after a pop to 30,871.50. Two big 1m bodies cut straight through to the 1m 50 EMA near 30,807 with no wick back.

When not to trade

Sitting out is the plan working, not failing.

No clear destination.
No clear stop level.
News due in the next few minutes.

Lines tangled and flat, price in the middle of the range.
Candles around the entry are messy: dojis, overlapping wicks.
A 1H or 4H candle about to close.

Price medium-stretched from the Red 14 and you'd be chasing.
The path to target is blocked close to your entry.

Sizing and payout pacing live in the Payout Pace Planner and the Tradeflight Trading Primer, Part 4.

Evidence

What the textbook says, what NQ actually did, and what to do where they disagree.

How the test was run

Data: NQ1! from your TradingView. 15m: 30 Jun to 25 Sep 2026 (5,702 bars). 5m: 30 Aug to 25 Sep (5,284). 1H: 1 Jan 2025 to 25 Sep 2026 (10,227).
Score: from the signal candle's close, did price move one normal candle size (1 ATR) your way before one against within 20 candles? A coin flip is about 49%.
Read it as: 55%+ with a decent sample leans useful. Under 45% means the opposite tends to happen first.
Also tested: session and yesterday's levels, stretch from the Red 14, stop placement and higher timeframe candles. Those results sit in the Map, Bias, Levels and Trade plan tabs.
Limits: rules are coded mechanically, so they are rougher than your eye. The period was mostly an uptrend, which flatters longs. Small samples (n under 20) are hints, not proof.

Textbook facts, checked

The ChatGPT list, each drawn out, with what our data says.

Our setups, backtested

First-move score by timeframe, with sample size.

Setup	15m	5m	1H	Verdict
First tag of Red 14 in a bullish stack (long)	63% · 104 69% with higher TF	49% · 96	54% · 227	Best found
Two red candles, then a bullish engulfing at a level	68% · 37	50% · 24	45% · 55	15m / 5m
Same pattern nowhere near a level	41% · 37	36% · 42	56% · 79	Location matters
Retest of Red 14 after it crosses Gold 50 (with the bigger trend)	57% · 21	42% · 12	40% · 35	Mixed
Enter on the candle where Red 14 crosses Gold 50	42% · 26	56% · 16	44% · 45	No edge
Wick through a rising Gold 50, body closes back above (long)	36% · 133	52% · 112	44% · 269	Slow
Body closes below Gold 50 while Gold is above Purple (short)	49% · 107	46% · 99	43% · 223	No edge
Sweep of the 20-candle low, close back (long)	50% · 278	45% · 224	43% · 467	No edge alone
Squeeze break up, enter on the break candle	31% · 39	30% · 23	49% · 49	Contradicts
Squeeze break up, enter when price comes back to the lines	39% · 36	39% · 23	55% · 42	1H only
Two green, then a bearish engulfing at a level (short)	46% · 37	32% · 22	53% · 58	No edge
Gold 50 dips to Purple 200, no cross, pulls away up (long)	58% · 19	-	62% · 26	Supported
Gold 50 rises to Purple 200, no cross, pulls away down (short)	45% · 20	-	63% · 30	1H
Red 14 touches Gold 50, no cross, pulls away (with trend)	28-33% · 67	-	53-58% · 137	Late on 15m
First close beyond the Asia low (short)	70% · 23	-	-	Break beats fade
First close above yesterday's high (long)	61% · 18	-	-	Break beats fade
Fade the first poke above yesterday's high (short)	33% · 12	-	-	Weak
Stretched 2.5+ ATR above Red 14, go with it (long)	64% · 100	-	-	Strength
Stop under the entry candle's body (pullback long, 2R)	18% · 105	-	-	Too tight
Golden Cross (long)	30% · 10	21% · 14	55% · 31	Late

Where the data disagrees: potential plays

Our idea is outcome A. What the data leans toward is outcome B. The context line says when each one is more likely.

Squeeze break

A · our ideaPrice breaks out of the EMA bunch and expands straight away.

B · data leansOn 5m / 15m the break candle usually gets a 1 ATR pullback first (only 30% went straight).

Play: don't chase the break candle. Wait for the pullback into the bunch and a close back out. A is likelier when a sweep happened first (like last night's 1m) or on the 1H.

Rejection of the Gold 50

A · our ideaWick through, close back, price lifts off the 50.

B · data leansPrice chops around the 50 first. The move up tends to come later (40 candles on: 58% on 5m and 1H).

Play: treat the wick as "the level is alive", not the entry. The trigger is the next body close above Red 14. Leave room under the 50 for the chop.

Loss of the Gold 50 in an uptrend

A · our ideaBody closes below the 50, the retest fails, and you short it.

B · data leansWhile Gold 50 is over Purple 200 this is a coin flip, even on the retest (43% to 51%).

Play: it's a warning, not a short. The short needs a bearish stack, or Gold 50 losing Purple 200, or a lower high under the 50.

Sweeps

A · our ideaSweep of the lows plus close back = long.

B · data leansOn its own, a sweep is about 45% to 50%. It needs company.

Play: only count a sweep when it happens at a zone (OB / FVG / session level) or into an EMA, with the higher timeframe on your side. That combination wasn't coded, so it's still a working idea.

Engulfing candles

A · our ideaTwo candles one colour, then an opposite engulfing = a turn.

B · data saysBullish version at a level: 68% on 15m. Nowhere near a level: 41%. The bearish version showed no edge.

Play: location is what makes it work. Use it on 15m / 5m for longs at an EMA or after a sweep. On the 1H the effect flipped, so don't use it there.

Crosses as entries

A · common beliefEnter when Red 14 crosses Gold 50.

B · data saysNo edge on any timeframe. It's late and it whipsaws.

Play: use crosses to set the stack, then enter on the first pullback to Red 14 once the stack is aligned. That was the strongest setup found.

First touch of a big level

A · our ideaPrice pokes above yesterday's high or the Asia high, sweeps it, and reverses.

B · data leansOn the first touch, a close beyond the level was usually a real break (61 to 66% of first pokes). Fading it scored 33%.

Play: on the first touch, wait for the close. Close beyond = trade the break or the retest. Close back inside = the sweep is on. A is likelier on a second test, at a zone, or against a stretched move.

A big stretch from the Red 14

A · common beliefPrice far from the line has to snap back, so fade it.

B · data leansAt 2.5+ ATR above, only 35% came back within 10 candles, and going with it won 64%.

Play: a huge stretch is momentum. Don't fade it without a level and a rejection candle. The medium stretch (1.5 to 2.5 ATR) is where price usually comes back (70%).