Methodology
How PatternX works
Short answer
PatternX detects 20 chart patterns with fixed rules, scores each setup with a transparent 13-factor PatternX Score, optionally adds an 8-factor AI context evaluation (Pro and Elite), and checks every order against your risk limits — with a stop placed at the broker on live trades.
PatternX is built deterministic-first. Twenty fixed, rule-based detectors find chart patterns; a transparent 13-factor score rates each setup; your own settings and a risk engine decide what — if anything — is traded. The optional PatternX AI layer only evaluates setups the engine has already found: it can veto a setup or re-weight its confidence, and it never invents one.
These pages describe what the product actually does, with the real thresholds, so that you — or anyone citing PatternX — can check the claims. They also list what PatternX can't do: start with limitations if you want the caveats first. Proprietary source code and secrets are not published; the rules and numbers that decide your trades are.
- MethodologyPatternX Classic vs PatternX AI: how each worksHow PatternX Classic and PatternX AI work: the same 20 rule-based detectors and 13-factor score, plus the 8-factor context layer only AI mode adds.
- MethodologyHow PatternX detects chart patternsHow PatternX's 20 rule-based detectors read candles, find swing points, confirm breakouts, apply market filters and set targets and stops, step by step.
- MethodologyThe PatternX Score: how setups are scoredHow the PatternX Score rates each detected setup from 0 to 100 with 13 fixed weighted factors, how it becomes confidence, and why it is not a probability.
- MethodologyPatternX AI: the context layer explainedHow PatternX AI rates the context of setups the engine already found: 8 fixed weighted factors, the veto rule, confidence re-weighting and its limits.
- MethodologyPatternX risk controls: what is checked before every orderEvery check PatternX runs before an order (kill switches, quotas, PDT guard, loss limits, Elite halts, sizing, margin) and how live positions stay protected.
- MethodologyMarket data at PatternX: sources, freshness and limitsWhere PatternX gets its data (venue candles, order books, funding, broker feeds, IBKR, Alpaca and Polygon for Momentum Lab), how fresh it is and its limits.
- MethodologyMomentum Lab methodology: strategies, states and protectionHow Momentum Lab detects premarket momentum, moves stocks through states, gates them by market cap, sizes, exits and protects positions, and what a trail cannot do.
- MethodologyHow PatternX backtests work: assumptions and limitsWhat PatternX backtests replay, the costs and defaults they assume, the statistics they report, Momentum Lab replay limits and how to read results honestly.
- MethodologyWhat PatternX can't do: limitations and honest caveatsThe honest limits of PatternX: forming bars, fixed windows and priors, scan cadence, simplified backtests, software-managed stops, and venue and data risks.
Educational content, not financial advice. Chart patterns describe probabilities, not certainties; trading — especially with leverage or outside regular hours — can lose money, including more than you expect. See the risk disclaimer.