Trenchlight

Lesson 1 · Orientation

Start here

Every fifteen minutes (and, in the short version, every five) the market asks one question: will Bitcoin finish this round above the price it started at? This playbook is about answering that question with a process instead of a hunch — and about paying a fair price for the answer.

The BTC Prediction Machine tab — the playbook turned into instruments. Every lesson ends with where the idea lives on this screen.

How to use it

Read the lessons in order. Each one ends with an In the terminal box that shows where the idea lives in the BTC Prediction Machine tab, so theory and screen stay connected.
Tick the checkboxes as you go; your progress and checklists are saved in this browser.
Keep the last lesson, the cheat sheet, open while you trade. It is the whole method on one page.

The setup

One chart: BTC/USDT · Binance · 1-minute candles. Five indicators: session VWAP, EMA 9 and EMA 21, pivot highs and lows over the last ten bars, RSI 14, MACD 12 / 26 / 9. The terminal draws all of them and adds the order-flow and probability layers that a plain chart cannot show.

The one idea behind everything In a binary round the price of a share is the crowd's probability. You only make money when your own probability is higher than the price you pay — and when you pay a price that leaves room for being wrong. Most losses here are not wrong calls; they are right calls bought at 85¢.

What to expect

The edge is small and comes from four boring things: entering in the right price zone, entering at the right minute, sizing so that no single round matters, and taking profit on a ladder instead of hoping for 100¢. Direction calls are the easy part; the terminal will hand you those. Discipline is the hard part, and no software supplies it.

Three stages, one routine.

The rules below follow the widely used 15-minute BTC playbook (the PolySigma course), retold in our own words, checked against how the markets actually resolve today, and extended w ith our own probability model and accuracy journal. Education only — nothing here is financial advice, and a binary share can go to zero.

In the terminal The tab next to this one. Top-right switch 15M / 5M picks the round type; the left column is the market and your paper account, the centre is the chart, the right column is the decision stack: FORECAST → SIGNAL → CONFLUENCE → ENTRY CHECK.

Lesson 2 · Market structure

How a round works

A new 15-minute round opens at :00, :15, :30 and :45 (times are shown in New York time, the way the markets are named); 5-minute rounds open every five minutes. Both run around the clock, seven days a week.

Price to beat and the result

At the open the round records the price to beat — Bitcoin's reference price at that second. At the close the same reference is read again as a 60-second time-weighted average (Chainlink BTC/USD), which is why the last minute cannot be gamed by a single print. Average at or above the price to beat → UP; below → DOWN. The difference between the current reference price and the price to beat is the gap, the single most important number of the round.

A round on the clock: watch early, act in the window, manage late, hold through the TWAP.

Two shares, one dollar

Each round has an UP share (YES) and a DOWN share (NO), priced from 0¢ to 100¢. The winner pays $1, the loser pays nothing, and the two prices add up to about 100¢ (the 1–2¢ that is missing is the spread). Read a price as a probability: UP at 64¢ means the crowd thinks UP has a 64% chance.

A share is a probability with a $1 payout: the price you pay is the win rate you need.

Time left	Phase	What it means for you
15:00 → 10:00	Early	Little information, wide spreads, coin-flip prices. Watch, do not buy.
10:00 → 5:00	Best window	Enough candles to judge momentum while the favourite still costs 55–70¢.
5:00 → 1:00	Late	The favourite is already 80–95¢: tiny payoff, full risk. Manage exits, do not open.
1:00 → 0:00	Resolution	Only the TWAP matters now. Nothing to do but hold or sell.

Favourite–longshot bias Late in a round crowds overpay for the obvious side. The zone that has historically paid the buyer is 55–72¢. Above 72¢ the break-even win rate you need is higher than what any signal can honestly promise. Never buy above 72¢.

Reading the order book

Micro-price = (bid × ask size + ask × bid size) ÷ (bid size + ask size). Above the mid-price → buyers are pressing; below → sellers are.
Thin book: fewer than roughly 1,500–2,000 shares across the top five levels means your own order moves the price. Skip.
Spread gate: do not enter with a spread wider than 3¢. Spreads widen in the first 60 seconds of a round, when Bitcoin moves more than $100 in a minute, and at night and on weekends.
Latency edge: Binance prints move one to three seconds before the round's prices react. During $200-per-minute moves the market price is stale — that is where a fast feed earns its keep.

The order book: spread, micro-price and depth tell you whether an entry is fair right now.

The market card: price to beat, live reference with its delta, sparkline, YES / NO prices, gap and the countdown.

In the terminal The market card shows the price to beat, the live reference price with its delta, the gap, YES / NO prices and the countdown; the price strip under the chart repeats YES / NO with the spread. The FLOW section has OB WALL for book thickness and the depth ratio.

I can find the price to beat, the gap and the time left in under two seconds. I know the four phases and that I only open in the 10:00 → 5:00 window.

Lesson 3 · Market math

The math

A round is a bet with a known payout, so it can be priced. Do the arithmetic once, properly, and the rules of the rest of the course stop looking arbitrary.

Break-even and expected value

Buy at p cents and you need to win more than p% of the time just to stay flat: 60¢ needs a 60% win rate, 80¢ needs 80%. Expected value per share is EV = WR × profit − (1 − WR) × loss. With the exit ladder of Lesson 6 the average sale lands near 94¢, and with a −25% stop the loss is a quarter of the entry price.

Entry	Profit if right (to ~94¢)	Loss if wrong (−25%)	EV at 65% win rate
55¢	+39¢	−14¢	+20.5¢ per share
65¢	+29¢	−16¢	+13.2¢
72¢	+22¢	−18¢	+8.0¢
85¢	+9¢	−21¢	−1.6¢ — negative even when you are usually right

Break-even win rate = entry price. Above 72¢ no honest signal gets there.

Expected value at a 65% hit rate: positive under 72¢, negative at 85¢.

That last row is the whole argument for the price zone: at 85¢ a 65% hit rate loses money.

The gates

Entry price 50–72¢; the best fills are 50–58¢.
Spread at most 3¢.
No new entries in the last 1–2 minutes.
Daily stop at −15% of balance; after three losses in a row, stop for the day.
Flat regime: if Bitcoin moved less than $30–40 in the last fifteen minutes there is nothing to ride. Do not trade until it wakes up.

The dearer the share, the smaller the win and the larger the loss.

Size while the account is small

With a balance under about $75 the playbook allows ~20% per round so that a winning streak actually compounds; once the account has grown, the standing size becomes 12% (Lesson 7 refines this).

Best hours, in one line London morning (2:00–7:00 ET) and the London–New York overlap (7:00–10:00 ET) have the deepest books and the tightest spreads. Fridays and Sundays are thin.

ENTRY CHECK grades the four gates for the selected round: price zone, time window, ATR regime, signal.

In the terminal ENTRY CHECK grades the price zone (50–72¢), the time window and the ATR zone for the selected round; the SIGNAL card prints edge = confidence − price. The paper-trading card lets you run the whole ladder with virtual money before risking any.

I can compute the break-even win rate for any price in my head. My daily stop and my streak rule are written down before the session starts.

Lesson 4 · Indicators

The five indicators

None of these predicts the next candle. Together they tell you whether the current move has momentum, structure and room — the three things a 15-minute continuation trade needs.

Each indicator answers a different question; none of them decides alone.

RSI 14

Above 70 the move is stretched: wait for RSI to drop back under 70 and for MACD to agree, then lean DOWN. Below 30, mirror it: a climb back above 30 plus MACD → UP. Between 45 and 55 RSI says nothing — treat it as a skip zone. Price making a new high while RSI does not (divergence) is weakness, not strength.

MACD 12 / 26 / 9

Above zero the regime is bullish, below zero bearish. The MACD / signal cross is the main trigger; a growing histogram means the move is accelerating, a histogram shrinking toward zero is an early warning. The best entries come from a cross on the 1-minute or 5-minute chart in the first 3–7 minutes of a round. When the 15-minute ATR is under $30, MACD crosses are noise.

Session VWAP

Price above VWAP → UP bias, below → DOWN bias. VWAP is also a magnet: the further price is stretched from it, the more likely a snap back. Three setups: Trend (break through VWAP with MACD green and RSI above 50), Fade (far above VWAP with RSI above 70 → buy DOWN), Bounce (price touches VWAP from above and turns → buy UP). Most reliable around the New York open.

EMA 9 / 21

EMA 9 crossing above EMA 21 → UP, crossing below → DOWN. The distance between them is the trend's health: widening means strength, narrowing means fading, parallel and close means chop. A cross inside the first 3–5 minutes of a round is an early entry.

Pivot highs and lows (10 bars)

The highest high and lowest low of the last ten candles are the local resistance and support. Break above the high → UP; break below the low → DOWN; a bounce off the low → UP; a rejection at the high → DOWN. Price wandering between the two levels is chop — no entry.

Combination rule Enter only when at least 3 of the 5 point the same way. Size by agreement: 5 / 5 → 15% of balance · 4 / 5 → 12% · 3 / 5 → 6–8% or skip · fewer → skip.

Direction from the gap, permission from the agreement count.

Indicator	UP checklist	DOWN checklist
RSI 14	Came back up through 30 (or rising through 50)	Came back down through 70 (or falling through 50)
MACD	Histogram green and growing, or a cross up	Histogram red and growing, or a cross down
VWAP	Price above VWAP	Price below VWAP
EMA 9 / 21	EMA 9 above EMA 21, gap widening	EMA 9 below EMA 21, gap widening
Pivots HL10	Break above the 10-bar high	Break below the 10-bar low

The chart with VWAP, EMA 9/21, MACD and RSI (with its moving average) — the five indicators on one screen.

In the terminal The chart toggles VWAP, EMA 9/21, HL10, MACD and RSI (with its moving average). CORE SCORE x/10 weighs the four heavy hitters — gap, order-book imbalance, RSI, MACD — and CONFLUENCE x/16 counts every factor that agrees with the signal's direction.

I can name the UP and DOWN reading of each indicator without looking. I will not enter with fewer than three of five in agreement.

Lesson 5 · Trend

Candles and trend

Indicators lag; candles do not. The shape of the last three bars tells you whether a move has conviction, and the clock tells you whether there is still time to be paid for it.

Heikin Ashi as a trend lens

Heikin Ashi candles average out the noise: two or more consecutive candles of the same colour with no wick on the opposite side are a strong confirmation that the move is still alive. The terminal counts the run for you.

Read the body and the wicks before you read any indicator.

The three-candle filter

Three consecutive candles of the same colour, each closing beyond the previous one, with bodies larger than 40% of their range and wicks smaller than half the body. Simple, and historically right about 65% of the time on its own — which is exactly why it is a filter, not a system.

Body above 60% of the range: a strong candle. Below 30%: indecision; below 20%: a doji, skip.
Small candles of equal size in a row: low ATR, nothing to trade.
Shrinking bodies inside a move: momentum fading — start walking down the exit ladder.
Trade continuation, not reversal. If the momentum has not shown up in the first 5–8 minutes of the round, it is not coming.

The three-candle filter: momentum must already be visible in the closes.

Four timeframes, one decision

Timeframe	Role
1m	Tactics: the entry candle and the indicator crosses.
5m	Confirmation that the 1-minute move is not a blip.
15m	Structure: the previous 15-minute candle sets the bias for this round.
4h	Macro direction. Fighting it is allowed but costs size.

Three of four timeframes must agree; the price strip shows all four.

Enter when at least three of the four agree with your side.

The clock Best entries fall on minutes 5–10 of the round. The last five minutes are risk without payoff, and the last two minutes are for exits only. ATR under $50 → do not trade; ATR above $80 → ideal.

The price strip: YES / NO with the spread, the gap, the 4H · 1H · 15M · 5M trend badges and the time left.

In the terminal The price strip shows trend arrows for 4H · 1H · 15M · 5M; the chart has a Heikin Ashi toggle; the 3-candle factor and MOMENTUM panel feed the signal; ENTRY CHECK's time gate turns green in the right minutes.

Lesson 6 · Exits

The exit ladder

A share that will pay $1 at the close is worth less than $1 while the round is still open, because the round can still turn. Selling in steps converts most of that value to cash before the last minute decides everything.

Price reached	Sell	Formula (shares)
90¢	25% of the position	floor(shares × 0.25)
93¢	33% of what is left	floor(rest × 0.33)
95¢	50% of what is left	floor(rest × 0.50)
97¢	everything	rest

Each step is roughly a quarter of the original position, so the average sale lands near 94¢. The point is not the last three cents; it is that a late reversal can no longer take the whole trade.

The ladder turns a 62% winner into money before the last minute can take it back.

Worked example

$100 into UP at 58¢ → 172 shares. At 90¢ sell 43 (38.70). At 93¢ sell 42 of the remaining 129 (39.06). At 95¢ sell 43 of the remaining 87 (40.85). At 97¢ sell the last 44 (42.68). Proceeds $161.29 on a $99.76 cost: +62%, average exit 93.8¢.

Same trade, two endings.

The stop A hard stop at −25% of the entry price is not optional: bought at 58¢, you are out at 43–44¢. Once the first rung (90¢) has filled, move the stop to break-even. With under two minutes left and 97¢ not reached, sell everything at market — the resolution is not worth the last few cents.

In the terminal The paper-trading card shows your open position with live P&L and a SELL button; the position note reminds you that an unsold share auto-settles at 1.00 / 0.00. Run the ladder there until the four rungs are muscle memory.

I know my four ladder prices and my stop before I click buy. I never hold an unsold position into the last minute hoping for 100¢.

Lesson 7 · Sizing

Position sizing

Size is where most accounts are actually lost. The rule set below keeps any single round small enough to be survivable and lets the size grow only when the evidence and the account both allow it.

Base and Kelly

The standing size is 12% of balance. It comes from the Kelly formula f* = (b·p − q) ÷ b (b = net odds, p = win rate, q = 1 − p) applied at about 60% of full Kelly, because your win-rate estimate is never as good as you think. The practical range is 8–20%.

Condition	Adjustment
Entry price 55–60¢	+2%
All four timeframes agree	+2%
London–NY overlap or NY open	+1%
Two losses in a row	−3%
Drawdown deeper than 10%	−3%
Three or more wins in a row (do not get cocky)	−1%
Spread wider than 2¢	−1%
Asian session	−2%
Abnormally high ATR	−2%
Five losses in one day	stop until tomorrow

Size = 12% base plus the adjustments, inside 8–20%.

Scale with the account, shrink with the drawdown

Balance	Base size	Drawdown from peak	Multiplier
$100	20%	0–5%	× 1.0
$300	18%	5–10%	× 0.8
$500	15%	10–15%	× 0.6
$1,000+	12%	15–20%	× 0.4
		over 20%	stop, audit

Steady size compounds; one oversized loss flattens the curve for weeks (log scale).

Signal strength sets the final number inside the range: weak 8–10%, normal 12–15%, strong 16–20%. One position per round, never two sides, never two rounds at once while you are learning.

Example: balance $500 → base 15% = $75. Price 57¢ (+2%), all timeframes agree (+2%), but it is the Asian session (−2%) → 17% = $85.

FORECAST with the CALLS box underneath: budget, sizing mode, price cap, and the call for the current round — the sizing arithmetic done for you.

In the terminal The paper card's slider sets the stake; the SIGNAL card's strength badge (POSSIBLE → SIGNAL ✓ → STRONG 🔥) is the "weak / normal / strong" input to this table. The session badge top-left tells you which session adjustment applies.

Lesson 8 · Advanced

Advanced signals

The five indicators say which way. This layer says whether the conditions allow a trade at all, and it outranks them.

ATR — the master switch

ATR(14) on the 15-minute chart measures how much Bitcoin typically moves per round. It decides both whether you trade and how much.

15-min ATR	Regime	Size multiplier	Which signals work
under $30	Quiet	no trade	none — everything is noise
$30–60	Moderate	× 0.8	EMA and pivots
$60–100	Golden	× 1.0 (base 12%)	all five
$100–120	High	× 0.9	RSI, MACD, EMA, pivots
over $120	Extreme	× 0.6	pivots only

ATR shift: when the last four candles are clearly wider than the four before, the regime just changed — skip the first candle of the new regime rather than trading the transition.

ATR(14) on the 15-minute chart decides whether to trade at all and how much.

Bollinger Bands (20, 2) on 5 minutes

Bands pinching together is a squeeze: energy building, direction unknown — wait. A close outside the upper band → UP, outside the lower band → DOWN.

A squeeze warns; the close outside the band confirms.

StochRSI (14, 14, 3, 3) on 1 minute

Above 80 the move is overheated: do not open UP. Below 20: do not open DOWN. A cross down through 80 is bearish, a cross up through 20 is bullish. In the score it is a warning, not a veto: −1 when it is overheated against your side, +1 when it just crossed in your favour.

The spread gate

under 1¢ clean · 1–2¢ acceptable · 2–3¢ only with a strong signal (more than half the factors) · over 3¢ entry blocked.

Never enter with a spread above 3¢.

Hierarchy ATR → RSI → MACD → EMA 9/21 → pivots → StochRSI (warning only). If RSI and MACD disagree, skip. You need at least two confirmations among RSI, MACD and EMA for any entry.

The FLOW cards: CVD, gap momentum, Coinbase premium, MARKET REGIME (the ATR zone), VPIN, large trades, order book, volume delta, StochRSI.

In the terminal FLOW card MARKET REGIME prints the ATR zone (OPTIMAL — golden zone, all signals reliable / TREND — follow momentum / QUIET), STOCHRSI shows k / d and the state (TURNING UP, OVERBOUGHT, IN RANGE…), OB WALL reads the depth, and the strip shows the spread. ENTRY CHECK's ATR gate is this table.

I check the ATR regime before I look at any other indicator. Spread over 3¢ or RSI/MACD conflict = no trade, no exceptions.

Lesson 9 · Execution

Psychology and the journal

Every rule so far can be executed by a calm person in ten seconds. The job of this lesson is to make sure the person at the keyboard is calm — and to notice, from the numbers, when they are not.

The streak protocol

Losses in a row	Action
2	Pause 30–60 minutes. Walk away from the screen.
3	Stop for the day.
4+	No trading until you have audited the journal and can name what went wrong.

Think of it as emotional capital: tilt spends it faster than money, and a tilted trader breaks the price zone first, then the sizing, then the stop. The rules exist so that decisions are made before the round opens, when you are still the person who wrote them.

The streak protocol: decided before the round opens, executed without discussion.

Journal every round you touch

Date and time (ET), direction, entry price, ATR at entry, RSI, MACD state, EMA 9/21 state, the reason in one sentence, exit price(s), P&L, and a single honest column: rule violated? A journal with that column is the only cure for revenge trading, because revenge trades always violate a rule.

Before the session

Am I inside a good session (04:00–12:00 ET)?
What is the ATR regime right now?
Balance versus its high-water mark — is the drawdown multiplier in effect?
Daily stop written down (−15%).
Any FOMC / CPI / NFP / PCE release in the next two hours?

Five minutes before the first round.

FORECAST ACCURACY: the terminal grading its own calls — hit rate, calibration by confidence, accuracy by time left, and the call history with its balance.

In the terminal The paper card's History is your practice journal. FORECAST ACCURACY (bottom-left) is the terminal's own journal: it stamps its forecast at fixed checkpoints of every round and marks it ✓ / ✗ against the official result — hit rate, calibration by confidence, accuracy by time left, and what following every "IF YOU BUY" would have returned. Read it before you trust any single call.

Lesson 10 · Sessions

Sessions and timing

Bitcoin trades all day, but its order books do not. The same signal is worth more at 08:30 ET than at 18:30 ET because the people on the other side are different.

ET	Session	Verdict
20:00–05:00	Night / Asia	✅ tradable, smaller size (−2%)
06:00–10:00	London + New York	🏆 best: tightest spreads, deepest books
10:00–15:00	NY midday	⚠️ volume fades after 11:00; be selective
15:00–16:00	NY close	✅ a second burst of activity
16:00–20:00	Pacific	❌ thin, wide spreads — avoid

The day in Eastern Time: London and the overlap pay, nights and weekends bite.

Key moments: 04:00 ET London opens, 09:30 ET the NYSE opens, and 08:00–12:00 ET is the daily peak.

When to stand aside

FOMC: two hours before and two hours after.
CPI, NFP, PCE (08:30 ET): from 30 minutes before to 2 hours after.
Sunday 21:00–23:00 ET (the weekly open).
ATR under $30 for four hours or more: wait until it has been above $40 for two hours.
The first 5–10 minutes after any headline — whipsaw territory.

When to stand aside — six situations where the best trade is none.

Macro background

S&P 500 down 2% or more → be careful with UP. Dollar index up 1% or more → prefer DOWN. Fear & Greed under 20 → the tape is unstable; size down or stay out.

ATR by session Dead under $30 → no trade · Quiet $30–60 → 7–9% · Golden $60–100 → 12% · High $100–120 → 7–9% and demand 3 of 5 · Extreme over $120 → 7% or skip.

The session card: which session it is, the last ten 15-minute and 5-minute results and the majority lean.

In the terminal The session badge (top-left) names the session and its quality; MACRO NEWS lists today's high-impact releases with a rule reminder and the daily brief; the FORECAST card's SESSION row compares the volatility right now with what is typical for this hour (CALM / NORMAL / HOT / EXTREME) and widens its noise estimate around scheduled releases.

Lesson 11 · Risk

Capital protection

With a 12% position and a −25% stop, a losing round costs 3% of the balance. That number is small on purpose: it means a bad afternoon is recoverable and a bad week is survivable.

Event	Response
2 losses in a row	30-minute pause
5 losses in a day	Stop for the day
−20% in a week	Stop for the week
−35% from the all-time high	Full audit before the next trade

Protection tightens as the drawdown grows.

High-water-mark floor: if the balance falls more than 10% below its peak, stop for the day. Reserve: keep the account split 70 / 30 — only 70% is ever at risk in the market; the 30% is the rebuild fund and never migrates.

The high-water-mark floor follows the peak and only rises.

The money you trade is not all the money you have.

The order of failure Accounts do not die from one bad round. They die from the price zone being ignored after a loss, the size being doubled after two, and the stop being "given a little room" after three. The cascade above interrupts that sequence mechanically.

In the terminal The paper balance and its P&L line make the drawdown visible at all times; Reset lets you restart a practice block cleanly. The accuracy journal shows whether the calls you are following deserve the size you are giving them.

Lesson 12 · Practice

Worked examples

Numbers are invented but typical. Judge each round with the gates in order: regime, session, price, time, signal.

A · Clean continuation

09:47 ET, 15-minute round, 8 minutes left. ATR $78 (golden). MACD crossed up on the 1-minute at minute 4, RSI 58 and rising, price above VWAP, EMA 9 above EMA 21 and widening, and the last candle closed above the 10-bar high: 5 of 5. UP costs 61¢, spread 1¢. Gap +$45 and growing. Everything agrees → buy UP at 61¢ with 12% (+2% for a full agreement). Ladder from 90¢. Stop 46¢.

B · Right direction, wrong price

Same signals, but you looked at minute 9 and UP is already 77¢. Out of zone — skip, however good it looks. At 77¢ you need a 77% win rate just to break even; no continuation setup delivers that.

C · The fade

14:21 ET, 9 minutes left. Bitcoin has run $180 in twenty minutes and sits far above VWAP; RSI printed 76 and has turned down; the MACD histogram is shrinking; StochRSI crossed down through 80. Trend is still UP on the 4-hour, so the trade is a fade against the macro: allowed, but size down. DOWN costs 56¢, spread 1¢ → buy DOWN at 56¢ with 8%. Stop 42¢.

D · The trap

08:31 ET, one minute after a CPI release. ATR just jumped to $140, the spread is 4¢, the last three candles have wicks longer than their bodies. Three gates fail at once (extreme regime, news window, spread). Skip, and do not look for the "obvious" trade the next round either — the first 5–10 minutes after a release are whipsaw.

Four rounds through the same checklist.

Two patterns worth remembering Rounds that end at 09:00–09:15 ET have leaned UP in our sample (64%, 16 of 25 — a small sample, treat it as a hint, not a law) — the pre-market bid. And when Bitcoin has already moved $200 or more in the last thirty minutes, the momentum you would be buying is usually spent.

Entry checklist

Time: 5–13 minutes after the round opened (5-minute rounds: 0.5–4 minutes left). Price: 50–72¢, ideally 50–58¢; spread 3¢ or less. Regime: ATR $60–100 golden (or the multiplier applied); no release inside the window. Signal: RSI and MACD agree; at least 3 of 5 indicators; 1H and 4H trend not both against me. Size: 5–15% after adjustments; stop −25% placed; ladder 90 / 93 / 95 / 97 planned.

In the terminal These five lines are ENTRY CHECK (price, time, ATR, signal) plus the SIGNAL card's reasons list. When ENTRY CHECK shows 4 / 4 and the FORECAST card's IF YOU BUY row shows a positive EV, you are looking at example A.

Lesson 13 · The tool

Reading the terminal

The BTC Prediction Machine tab is the playbook turned into instruments. Read this once with the tab open next to it.

The whole tab at a glance: market and paper account on the left, chart and news in the centre, the decision stack on the right.

Left column — the market and you

Session badge with the last ten 15-minute and 5-minute results (green / red dots) and the majority lean.
Market card: price to beat, live reference price with delta, sparkline, YES / NO best prices, gap, countdown. The 15M / 5M switch at the top changes which round everything refers to.
Paper trading: a $1,000 practice account — BUY YES / BUY NO, SELL, History, Reset. The ladder and the stop are practised here.
FLOW cards: CVD divergence, gap momentum, Coinbase premium, market regime (ATR zone), VPIN, large trades, order book, volume delta, StochRSI, walls, open-interest delta. Each card is one factor of the signal.
FORECAST ACCURACY: the terminal grading its own forecasts (Lesson 9).

Centre — the chart and the news

1m / 5m / 15m / 1h candles with VWAP, EMA 9/21, Heikin Ashi, HL10 pivots, MACD, RSI + its average, the price-to-beat line and a 15-minute grid. The strip below shows YES / NO, spread, price, gap, trend arrows for 4H · 1H · 15M · 5M and the countdown. Under it, the macro card: headlines, today's high-impact events, a daily brief.

Chart, indicator toggles and the price strip.

Right column — the decision stack

Panel	What it answers
FORECAST — win odds	The probability that each side wins this round: distance to the price to beat versus the noise left in the remaining minutes, tilted by the 16 factors, the slot's history and the news. Levels: HIGH ≥ 85% · MEDIUM ≥ 68% · LOW ≥ 56% · TOSS-UP. The MARKET row compares it with the crowd's price; IF YOU BUY prices the better side as a bet (win chance, payout, EV).
SIGNAL	Direction and strength: SCANNING → POSSIBLE → SIGNAL ✓ → STRONG 🔥, or GAP LOCK 🔒 when the gap is too large for the time left to reverse. Shows early lean, gap, points, entry side and price, edge, and the reasons.
CONFLUENCE x/16 · CORE SCORE x/10	How many factors agree, and the weighted score of the four heavy ones (gap · order-book imbalance · RSI · MACD).
HISTORY	How this time slot resolved on each weekday over the past year (‹ › moves through slots).
MOMENTUM · REVERSAL	Delta, book and trend bars; the reversal detector with its RSI / MACD / StochRSI evidence.
ENTRY CHECK	The four gates of Lesson 12 — price zone, time window, ATR zone, signal — plus the probability that the current gap holds.

FORECAST — win odds, with the CALLS box.

SIGNAL: direction, strength and the reasons list.

HISTORY: how this time slot resolved on each weekday.

One round, start to finish

Minutes 0–4: watch. Note the regime (MARKET REGIME card), the session, the events.
Minutes 5–10: read FORECAST and SIGNAL. If they agree and ENTRY CHECK is 4 / 4, check the price is inside 50–72¢ and the spread under 3¢.
Size from Lesson 7, buy the side, note the stop.
Walk the ladder from 90¢; sell everything at market with two minutes left if 97¢ has not printed.
After the close, look at FORECAST ACCURACY: was the call right, and was the confidence honest?

What the terminal is not It is an estimate, printed with its own error rate next to it. A HIGH forecast wins about as often as it claims, which means it also loses about as often as it claims. Position size and the stop are what make that arithmetic survivable.

Lesson 14 · 5-minute markets

Five-minute rounds

A 5-minute round has five 1-minute candles. RSI and MACD barely move in that time, so the tools change: structure, the shape of the path to the target, and the order book.

Three structural signals

MSB — market structure break: a candle closes beyond the last swing high (→ UP) or swing low (→ DOWN). Wicks do not count.
FVG — fair value gap: a three-candle sequence where the first candle's high is below the third candle's low (or the mirror). The gap marks an imbalance that price tends to revisit — or to run away from when the trend is strong.
CISD — change in state of delivery: after a run of 2–4 candles of one colour, price closes back through the open of the first candle of that run. The run is over.

Structure first on the 1-minute chart.

Read them in that order: MSB first (is there a break?), FVG second (is there fuel?), CISD last (did the move just fail?).

The path to the target

Look at what lies between price and the price to beat. A clean path (few wicks, no consolidation) lets the gap close or extend quickly; a dirty path (long wicks, sideways clusters) means the move will stall. On the depth map, liquidity walls, absorption (big orders eating a move without price budging) and icebergs (orders refilling at one price) are the visible version of the same thing.

Levels that hold versus levels that give way.

Bias and volatility Take the bias from the 15-minute and 1-hour charts; if they disagree, skip the round. Moderate volatility is the window — dead tape gives no move, extreme tape gives no reliable structure. Entries only with 0.5–4 minutes left, never in the final thirty seconds.

In the terminal Flip the switch to 5M: the market card, signal, forecast, paper trading, history and the accuracy journal all switch to the 5-minute round. Forecast checkpoints for 5M are 4 · 3 · 2 · 1 · 0.5 minutes left. The OB WALL card reports walls and thin books, and the strip's 15M / 5M arrows are the bias check.

Lesson 15 · Reference

Cheat sheet

Keep this open while trading. Every number below is explained in the lessons.

Gates

Price 50–72¢ (best 50–58¢)
Spread ≤ 3¢ (≤ 2¢ for full size)
15m round: enter with 10 → 5 min left; 5m round: 4 → 0.5 min
No entries in the last 1–2 min
ATR 15m ≥ $60 golden; < $30 no trade
Flat tape (< $30–40 in 15 min) → wait

Signal

≥ 3 of 5 indicators agree
RSI and MACD must agree; ≥ 2 of RSI / MACD / EMA
≥ 3 of 4 timeframes (1m · 5m · 15m · 4h)
StochRSI > 80 → no UP; < 20 → no DOWN
Three-candle filter for continuation
Hierarchy: ATR → RSI → MACD → EMA → pivots → StochRSI

Size

Base 12% (20% under $100, 18% at $300, 15% at $500)
5/5 → 15% · 4/5 → 12% · 3/5 → 6–8%
+2% price 55–60¢ · +2% all TFs · +1% overlap
−3% two losses · −3% DD > 10% · −2% Asia · −2% extreme ATR · −1% spread > 2¢
ATR × 0.8 (30–60) · × 0.9 (100–120) · × 0.6 (> 120)
One position per round

Exits

Ladder 90¢ 25% · 93¢ 33% · 95¢ 50% · 97¢ all
Stop −25% of entry; break-even after 90¢
< 2 min and no 97¢ → sell all at market
Shrinking bodies → start selling

Sessions (ET)

🏆 06:00–10:00 London + NY
✅ 20:00–05:00 Asia (−2%) · 15:00–16:00 NY close
⚠️ 10:00–15:00 midday
❌ 16:00–20:00 Pacific · Sunday 21:00–23:00
FOMC ±2 h · CPI / NFP / PCE −30 min → +2 h
5–10 min after any headline: stand aside

Protection

Daily stop −15% · 3 losses in a row → day over
2 losses → 30–60 min pause · 5 losses → day over
−20% week → week over · −35% from ATH → audit
HWM floor: −10% from peak → stop for the day
70 / 30 reserve — the 30% never trades
Journal every round, with the "rule violated?" column

The cheat sheet prints on one sheet; the rest of the course stays on screen.

Setting	Default	What it does
DRY_RUN	1	1 is test mode: the bot prints what it would buy and sends nothing. 0 sends real orders. The start script sets it: option 1 is the test, option 2 is real money.
MAX_STAKE	5	The most one call may spend, in dollars (USDC). The bot takes the call’s stake and cuts it to this limit, but never goes below the exchange minimum of 5 shares, counted at its price cap (about $2.65 for a call at 51¢). The start script asks for it in real mode.
SLIP_LOW	0.09	For calls priced under 20¢: pay up to 9¢ more than the call price. Cheap sides move in big steps and still leave room to pay up.
SLIP_MID	0.05	For calls at 20–39¢: up to 5¢ more.
SLIP_HIGH	0.02	For calls at 40¢ and above: up to 2¢ more. Here every extra cent eats most of the edge.
RETRY_MS	3000	If nothing is on offer at the cap, keep trying for this many milliseconds: the price sometimes comes back.
RETRY_GAP	250	The pause between tries, in milliseconds.