How Elimination Works
Our AI runs 8 independent probability simulation batches, each seeded from the most recent actual draw result. Numbers that appear in none or very few of these batches are flagged as "losing numbers" — candidates for elimination. Numbers appearing across many batches form the recommended pool.
The Elimination Threshold
Approximately 10–15 numbers are eliminated each draw, reducing the field from 50 to approximately 35–40 numbers. This pool is then validated against historical balance patterns (odd/even, high/low) before being published.
Elimination vs Prediction
Elimination is a fundamentally different strategy from prediction. Prediction tries to identify specific winning numbers — which is impossible. Elimination removes the weakest candidates, leaving a stronger-calibrated pool to choose from. It is the same logic used in medicine (eliminate unlikely diagnoses) and engineering (eliminate failure modes).
After Elimination
Once you have the reduced pool, Pro subscribers receive AI-generated lines that cover every pool number in a balanced, weighted line set. Free users receive the elimination list and can construct their own lines from the remaining numbers.
Frequently Asked Questions
What if an eliminated number wins?
It will happen occasionally — no elimination system is perfect. In backtesting, our pool missed all 5 winning mains 0% of the time across 100 tests. But individual eliminated numbers do occasionally appear.
Is the elimination list different each draw?
Yes — it is fully recalculated before every draw based on the latest seed. Tuesday and Friday each get their own independent analysis.
Can I customise which numbers are eliminated?
Pro users can see the full pool and losing list, and the Regenerate function allows fresh line sets. Manual number selection from the pool is always available.