Skip to content
Nocterm Docs

How detection works

A detector runs on every closed bar, across eight timeframes, looking for the geometry of known setups. When it finds one it draws the pattern and pins down three levels: where you’d enter, where you’d take profit, and where you’d be wrong. An odds engine then grades that geometry against the asset’s own volatility - so you see not just what fired but how good the shape is.

Where: Setups panelTier: Trader
📷
Screenshot needed
id: setups-how-detection-works

A setup on the chart with its drawn geometry - entry, target, and stop - alongside its graded odds

suggested 1280x720

Overview

Section titled “Overview”

Detection is two stages. First the detector reads price action bar-by-bar and recognises a setup’s shape, then defines the trade around it: a trigger to enter, a target to aim for, and a stop that invalidates the idea. Second, the odds engine takes that fixed geometry and asks a simple question - given how this asset actually moves, how likely is target-before-stop?

It answers with a barrier-touch Monte-Carlo: it simulates many price paths calibrated to the asset’s recent volatility and counts how often the target is hit before the stop. From that it reports the probability of success, the reward-to-risk ratio, the expected value, and a rough time-to-resolve.

The grade is about the shape, not the future. The engine measures whether the geometry is favourable relative to noise - it does not predict that price will go up or down. The directional call stays yours.

How to use

Section titled “How to use”

  1. Open the Setups panel and watch the feed. Every closed bar can produce new detections, across all eight timeframes from 1m up to 1d.

  2. Read the quality metrics on each card - win %, EV (in R), and confluence - to triage the shape before you even look at the chart.

    📷
    Screenshot needed
    id: setups-how-detection-works-metrics

    A setup card showing win percent, EV in R, and a confluence count

    suggested 640x480

  3. Open Why it fired to see the trigger, target, and stop the detector defined, plus the graded odds behind the numbers.

What the odds engine reports

Section titled “What the odds engine reports”
MetricWhat it means
Win %Probability the target is touched before the stop, by simulation
R:RReward-to-risk - target distance over stop distance
EV (in R)Expected value per unit of risk, win % and R:R folded together
Time-to-resolveRoughly how long until the setup hits its target or its stop
Section titled “Related”