The Lab
Walk-forward validation of the Medallion engine against two baselines, week by week.
The Backtest Desk is a Quant feature
This is the page that tells you whether any of the rest of the site is worth paying for — so it is behind the tier for people who want to audit the model rather than just use it. Every metric below is computed from raw play data, not self-reported.
Unlock with Quant — $79/moA week-by-week table of the model's out-of-sample R², RMSE and rank correlation, each shown against a three-game exponentially weighted average — a baseline that is much harder to beat than most projection sites admit.
Plus an equity curve: the cumulative points the model's top-12 gained over the naive top-12. If that line is flat, the model is not earning its complexity, and we would rather you could see that than take our word for it.