Cross-game strategy scoreboard
| Strategy | Scored | Hits | Chance predicts | Hit rate | vs chance (z) | Best single |
|---|
Reading this: z measures how far a strategy sits from pure chance, in standard deviations. Values bouncing around 0 mean it performs exactly as randomness predicts — the expected result. Cell shading grows with |z| in either direction; darker just means “further from chance so far,” not “winning.” It would take a sustained |z| > 3 over many draws to suggest anything real, and no strategy is expected to get there.
Per-game leaderboards
Each game has its own exact null expectation (hypergeometric for jackpot games, binomial for digit games) — arms are never pooled across games here.
Today's picks
Experiment outputs, not advice — every arm is expected to score at chance. Once results land, actual numbers and match counts appear beside each pick.
Exploit watch
| Game | Combinations | Full-buyout cost | Break-even jackpot | Roll-down defect |
|---|
No tracked game has a capped-jackpot roll-down (the WinFall defect); unclaimed jackpots roll over, keeping per-ticket EV below cost. Digit-game payouts are fixed-odds — a structural house edge no monitor can wait out. “No opportunity” is the expected, and real, finding.