Draw Theory

DNA Fingerprinting

Builds a multi-dimensional fingerprint of every draw. Finds the 15 most similar historical draws and votes on what came next. Weight: 26–28%.

Markov Chain Transitions

Constructs a full conditional probability matrix. Predicts which numbers are most likely to follow the last draw. Weight: 22–25%.

Gap Signature Modelling

Models average gaps between consecutive numbers as recurring templates, then projects forward. Weight: 20–24%.

Cohort Ageing

Tracks frequency across number bands and boosts underrepresented zones. Retuned for each game's pool size. Weight: 15–18%.

Volatility Cycles

Detects SPREAD vs CLUSTER phases and biases selection toward statistical reversion. Weight: 10–12%.

AI Synthesis

Claude AI receives all algorithm outputs, detects patterns humans miss — birthday bias, consecutive droughts, zone anomalies, sum drift — and generates the final intelligent pick with reasoning.