Evidence-led research, transparently documented.
Six principles govern every publication we release, from data integrity and statistical validation to limitations and reproducible code.
Every conclusion must be supported by evidence.
Our methodology is designed to make every published strategy easier to understand, reproduce, challenge and evaluate independently.
Documented Assumptions
Every backtest states its data window, transaction cost model, slippage assumptions and any survivorship or lookahead controls.
Reproducible Code
Complete Python and Pine Script source is included with every strategy so results can be independently reproduced.
Standardised Metrics
CAGR, Sharpe, Sortino, maximum drawdown, profit factor and win rate are reported consistently across all publications.
Robustness Discussion
Where practical we discuss walk-forward performance, parameter sensitivity and regime behaviour.
Honest Limitations
Every dossier includes a limitations section explaining conditions under which the strategy is not expected to perform.
No Guarantees
Historical and hypothetical results are always distinguished from live outcomes. Past performance is not a reliable indicator of future results.