Two tiers. Free runs the core methods at a small scale, which is enough to see whether a strategy survives contact with statistics. Pro removes the caps and adds the sections that need more machinery.
// billing cycle
// tier 00 · taster
FREE
The core methods, at small scale. No card required.
$0/ forever
→Monte Carlo up to 1,000 iterations. Reshuffle the order of your trades, or resample them with replacement, and read the spread of equity curves that comes back.
→Two hypothesis tests to start. Use a t-test where averages carry the story, and Wilcoxon where the data is too skewed for that.
→Sharpe and Sortino as point estimates. That is the single headline number, with no interval around it.
→The five-number summary of your returns: minimum, lower quartile, median, upper quartile, maximum.
→Single-variable linear regression. Fit a line through one predictor, then read its slope, standard error, and R².
→Regime detection with hidden Markov models. Viterbi decoding then labels each period with the state it most likely belonged to.
→Machine learning over your own features. Random forests and gradient boosting rank what actually predicts an outcome, and clustering finds structure when nothing is labelled.
→Correction for testing many things at once. Bootstrap confidence intervals, White’s Reality Check, and Hansen’s SPA all account for the winners that turn up by chance when you try enough candidates.
→Validation that respects time order. Walk-forward, combinatorial purged cross-validation, and purged k-fold keep training data out of the future, and the probability of backtest overfitting estimates how much your selection process flattered you.
→The rest of the hypothesis-testing suite. One-way ANOVA with Tukey HSD compares several strategies at once, and prediction intervals cover where a single new observation is likely to land.
→Performance metrics reported with intervals rather than as bare numbers.
→Position sizing. Put one trade history through different sizing rules and compare what each does to drawdown.
→Prop-firm simulation for eight funding programs, Topstep, Apex, FTMO and My Funded Futures among them. Evaluation and funded phases both run.
→Strategy export to six platform languages. Pine, ThinkScript, MQL5, NinjaTrader, EasyLanguage and Quantower all generate from one definition.
→Trade log import for common brokers and platforms, straight from their CSV exports.
→A TypeScript SDK, for running the same methods locally instead of in the browser.
→Save as many strategies as you like.
// payments processed by stripe · cancel any time · access runs to the end of the period you paid for
// the tiers above describe the plans · enforcement is not wired up yet, so every method is currently open to everyone