Live Experiment
Combinatorial vs Empirical
Two models score the same random numbers, then we check both against real Powerball draws. Pure math vs historical patterns — which matches more?
Combinatorial
Pure math. Every criterion derived from combinatorial probability — no historical patterns, no mean-reversion, no gambler's fallacy.
Empirical
Historical patterns. Structural base plus co-occurrence pairs, hot/cold numbers, drought bonuses, and Powerball weighting from v3/v4 era.
How This Works
Same Random Pool
50,000 random number combinations are generated once per draw. Both models score the exact same candidates — the only difference is how they rank them.
Top 10 Per Model
Each model selects its top 10 highest-scoring picks from the shared pool. These picks are stored before the draw happens — no hindsight.
Real Draw Comparison
After each draw, we count how many white balls and Powerballs each model's picks matched. Prize tiers are calculated using official Powerball rules.