Market Structure (ICT)
Mitigation Block
Same origin candle, same impulse move, one difference: whether the impulse left a clean fair value gap behind it. That single condition is what separates a mitigation block from an order block.
Run the same setup two ways, side by side
Every run builds two matched zones from the same starting price and roughly the same move size. The left one's impulse leaves a real three-candle imbalance, an order block. The right one's impulse covers similar ground with no imbalance, a mitigation block. Both get retested independently. Run enough pairs and the cumulative hold rates below tell you whether that one condition is actually doing anything in this simulation.
Order Block (leaves a fair value gap)
Origin candle, impulse, follow-through gaps clear, then a retest.
Mitigation Block (no fair value gap)
Same origin and impulse size, but the follow-through overlaps, no gap.
How it works
- The zone itself is defined identically both times. It's the full high-to-low range of the origin candle, the last candle opposing the impulse, in both panels.
- The only engineered difference is the gap. One impulse is sized so the candle two bars later leaves real open space above the origin candle's range, wick to wick, the other is sized to cover similar ground while still overlapping.
- Both zones get an independent, randomized retest. Nothing about whether one panel "should" hold is baked in ahead of time, the retest bars are generated fresh each run for each panel separately.
- A hold means the retest closed back on the origin side of the zone every bar. One close through it, on either panel, counts as a fail for that panel.
- Running just one or two pairs tells you almost nothing. The hold-rate stat only starts meaning something after enough pairs accumulate, which is exactly why the 10-pairs button exists.
Where this breaks
This simulation can't actually build in the reason order blocks are supposed to be stronger
The theory behind an order block outperforming a mitigation block is that a fair value gap is evidence of real, aggressive participation, size that traded through fast enough to leave a visible imbalance, which is what's supposed to make dealers or large participants defend that level again later. This page has no concept of real participation, it only checks one geometric condition, whether three candles happen to leave a gap. Run enough pairs and the two hold rates may end up close to each other, which would not prove the underlying theory wrong, it would only show that this simplified simulation isn't sophisticated enough to capture the mechanism the theory actually depends on. Treat any gap between the two numbers, in either direction, as a property of this toy model, not as proof about how real markets behave.