ICT setups, explained the way they are traded
The glossary tells you what an order block is. These pages tell you what to do with one: what to see, in what order, where invalidation sits, and when the setup does not count.
Intermediate level
The liquidity sweep, then the reversal
The setup everything else is organised around: price reaches past an obvious level for the stops, fails to hold, and turns back. Here is how to read it, and above all when not to take it.
Intermediate level
The Silver Bullet: one hour, one pattern
A one-hour window, a single permitted pattern, and the day is over if it does not show up. It is the most constrained setup in the method, and that is exactly what makes it useful to a beginner.
Beginner level
The order block, from spotting it to entering
The most-searched term in all of ICT, and the one drawn worst. An order block is not just any candle before a move: three things are required, and most of the drawings you see have only one.
Intermediate level
The fair value gap, and what it becomes when it breaks
An imbalance left by a move too fast for everyone to follow. Price often returns to it — and when it cuts through instead of reacting, the same level starts working the other way round.
Advanced level
The New York open continuation
Does the first 15-minute candle of the US session tell you what follows? Yes — by its size, not its direction. That nuance changes the rule you draw from it entirely.
Why these pages rather than the glossary
The glossary answers "what is a liquidity sweep". These pages answer a different question, the one you ask with a chart open: is this one, do I take it, and what will tell me I was wrong. Two different needs, and mixing them produces either an overlong definition or a vague playbook.
Every page follows the same order, deliberately: what the market is doing, what to see and in what sequence, where invalidation sits, and the cases where the setup does not count. That last part is missing almost everywhere else, and it is the only one that saves you money.
What you will not find here
No win rate, no expectancy, no "high-probability setup". That is not editorial caution: fourteen quality criteria have been measured on this approach — level strength, session, premium/discount, higher-timeframe bias, imbalance grading — across several thousand trades, and none separated good setups from bad. Everything stays around 50%.
What these pages can give you, and what is true: what a setup is made of, in what order, and where the point sits that proves the read was wrong. The rest — position size, exit, repetition — is not in the drawing, and yet that is where the outcome is decided.