Reach Probability (Reach %)

Reach Probability answers one question for every marker on the chart: what are the odds price gets to this level by my horizon? Open it with the 🎯 Reach % button in the nav bar (next to the view toggles), set your options, and press Compute. Each GEX level then shows a heat-colored % chip — green/warm for likely, cool for long-shot.

Under the hood it runs the same Monte Carlo engine as the Forecast chart: 20,000 price paths seeded with the current ATM implied volatility and bent by the live GEX surface. One simulation is run per (horizon, GEX impact) pair and the whole distribution is cached, so reading the odds for every marker is instant.

The three controls:

• Horizon — how far ahead. The deadline the odds are measured to (8h, 1d, 3d, 7d, 14d, 30d, this Friday, or end-of-month). The line under the buttons spells out the exact UTC target time and which expiration's IV is used.
• GEX Impact — how strongly gamma pins or amplifies moves. Light ≈ plain lognormal odds (weak gamma pull). Normal is the balanced default. Strong pulls paths hard toward the walls, tightening the distribution.
• Direction — which side to measure. auto ↑/↓ reads up to levels above spot and down to levels below spot. ≥ forces "ends at or above"; ≤ forces "ends at or below."

Reading the chips:

• ↑ — odds price is at or above the level by your horizon.
• ↓ — odds price is at or below it.
• <1% / >99% — the level is beyond the simulation's resolution. It means "vanishingly unlikely / near-certain," not a hard 0% or 100%.

How to use it: a nearby wall with a high Reach % is a realistic magnet for the session; a far wall at <5% is unlikely to trade before your horizon. Pair it with Flip Point regime context — strong-impact odds tighten fast once price is in positive-gamma territory.

Same limits as Monte Carlo: inputs are a snapshot now — the model doesn't anticipate new positioning or vol regime changes, the cone widens with time (short horizons are far more reliable), and tail odds are model-dependent. Treat the numbers as well-calibrated odds, not promises.