top of page
Glossary
Definitions used across the framework — consistent & measurable.
This glossary is the reference point for every dataset and episode. Definitions are standardised so outputs remain comparable across time and conditions. If a term appears on the Series or Public Data pages, it should be defined here.
-
If a word is unclear, search this page before asking in Discord.
-
Definitions here are written to be measurable, not poetic.
How To Use The Glossary
Core Terms
Term | Definition |
|---|---|
Condition / Variable | A feature used to segment events into comparable groups (e.g., context, location, or state), allowing probabilities and performance to be measured conditionally rather than averaged. |
Event | A precisely defined occurrence in price data that triggers measurement (the “thing that happens”), specified by objective rules so it can be identified consistently. |
Expectancy (R-Multiple) | The average risk-adjusted return per opportunity, expressed in R-multiples (realised risk-to-reward ratios), combining both win rate and the distribution of wins/losses into a single measurable edge. |
Market Regime/Market Structure | A classification of market conditions that groups periods where price behaviour is meaningfully similar, so results can be compared within similar environments. |
Objective / Target | A predefined level or condition that defines “success” for an outcome, used as the reference point for measurement rather than a discretionary profit-taking decision. |
Outcome | The measurable result observed after an event, defined in advance by explicit criteria. |
General Terms (A -> Z)
Term | Definition |
|---|---|
Binary Outcome | A yes/no definition of success for a hypothesis (e.g., “target reached” vs “not reached”), set before testing. |
Conditional Probability | The probability of an outcome given specific conditions are present. |
Expectancy | The average R-multiple per opportunity over many trades; positive expectancy implies a statistical edge. |
Feature | A measurable attribute engineered from raw price data, used to describe an event or its context so outcomes can be compared across similar situations. |
Hypothesis | A testable claim linking an event + conditions to an expected outcome, written so it can be proven wrong. |
Liquidity | The availability of resting orders; practically, areas where price can move quickly due to clustered orders. |
Overfitting | When rules match past data extremely well but fail forward because they captured noise, not a stable relationship. |
R-Multiple | Profit or loss expressed in units of risk (R); +1R means you made the same amount you risked. |
Sample Size | The number of observations used to calculate a metric; larger n (sample size) reduces noise and improves reliability. |
Signal | A rule that triggers an action (or flags an opportunity) based on predefined conditions; a signal is not “edge” until validated. |
Stop Loss (SL) | The predefined invalidation level where a trade is exited to cap loss. |
Subset | A filtered portion of data defined by conditions. |
Take Profit (TP) | The predefined target where profit is realised; usually tied to the hypothesis outcome definition. |
bottom of page