Every result comes from tested, deterministic functions and seeded synthetic data. Nothing is connected to live prices, staking or real money.
Turn a price and a probability estimate into break-even, expected ROI and a sensitivity band — and see when an assumed edge survives uncertainty.
Related: Implied probability and bookmaker margin · Expected value under uncertainty · Probability uncertainty and robust decisions
Open lab 1Enter home, draw and away odds and compare proportional, additive, power and Shin fair probabilities side by side.
Related: Implied probability and bookmaker margin · Removing the margin: four de-vig methods · Building a consensus fair price
Open lab 2Simulate thousands of synthetic bankroll paths under flat, fixed-fraction, Kelly and uncertainty-aware staking, including what happens when your probability is wrong.
Related: Kelly and fractional Kelly · Drawdowns and risk of ruin · Probability uncertainty and robust decisions
Open lab 3Score synthetic forecasts with Brier and log loss, read a reliability diagram, split by date and watch a leaky model collapse.
Related: Calibration, scoring rules and model drift · Backtesting without fooling yourself · Audit a machine-learning sports forecast
Open lab 4Audit a deliberately flawed synthetic backtest: find the bad price, look-ahead quotes, duplicate rows and unexecutable stakes.
Related: Anatomy of a false-positive backtest · Backtesting without fooling yourself · Write a reproducible research report
Open lab 5