Backtesting
A hunch is cheap. Run your strategy against seasons of real historical odds — walk-forward validated, the vig included, the drawdowns shown — and see whether it actually held up, or just felt like it did.
How it runs
Pick a model, a market, and a stretch of seasons. Blitzen replays every game in that window, places the bet the model would have placed at the real odds, and scores it against what actually happened.
Choose a model and market, a span of seasons, and your filters — an edge threshold, confidence tier, even the juice you actually pay.
A fresh model is trained per season on prior years only, then every game in the test season is bet at its real closing odds.
Each bet is scored win/loss/push against the real result — and against the close — rolling up into ROI, units, drawdown, and CLV.
No leakage
The fastest way to fake a great backtest is to let the model peek at the answer. We don't. Blitzen trains a fresh model for each season using only the seasons before it, then predicts that held-out season cold — every graded game is one the model never saw.
The window expands as it walks forward in time, exactly the way you'd actually bet: with the past, not the future. No leakage, no cherry-picked sample.
By confidence tier
A positive season that still doesn't beat the close — exactly the kind of result we show rather than hide. Strong bets run by default; this sample opts leans back in too. Toss-ups and avoids stay filtered — their win rates sit near a coin flip and drown the signal.
The whole story
A backtest that only shows green is a sales pitch, not a tool. Yours reports the full ledger: record, ROI, net units, the worst drawdown, the longest losing streak — at flat 1-unit stakes and real closing prices, so the vig is already working against you.
Break it down by market, by season, and by confidence tier. Then judge it yourself, with the losses in plain sight.
Graded against the close
Every backtest also reports closing-line value — did your bets get a better number than the market's sharpest closing price? Beat the close consistently and you're genuinely ahead of the market; win a coin-flip streak and you're not. See how we measure it →
Your controls
A backtest is only as honest as its assumptions, so the assumptions are yours to set.
Any span from 2022 through the current season — one year or all of them.
Only count a bet when the model disagrees with the market by at least your minimum edge.
Strong-only by default; opt into leans and toss-ups to see the full distribution.
Set the spread/totals price you actually pay; moneylines use the real odds per game.
Backtest any strategy against seasons of real odds. Free to start — bring your own read.