Pattern Memory Size
Number of past rounds (20-200) AI analyzes to detect patterns
Definition
Pattern Memory Size determines how far back in history the AI looks when analyzing patterns and making predictions. Larger memory (100-200 rounds) provides more statistical reliability and smoother trend detection but responds slower to recent changes. Smaller memory (20-50 rounds) is more reactive to current conditions but may be influenced by short-term noise. This is like choosing between long-term memory (big picture perspective) vs short-term memory (recent events focus). The optimal size balances stability with responsiveness.
Example
Memory = 30 rounds: AI analyzes only last 30 multipliers. Very responsive to current 'hot' or 'cold' streaks, but may overreact to random fluctuations. Memory = 100 rounds: AI considers last 100 multipliers. More stable, identifies genuine long-term patterns, but slower to detect sudden shifts. With 100 rounds, a 5-round anomaly is just 5% of data vs 16% with 30-round memory.
💡Strategy Tip
Recommended settings: Short-term reactive (30-50 rounds) - good for volatile sessions, Medium-term balanced (50-80 rounds) - best for most players, Long-term stable (100-150 rounds) - best for conservative strategies seeking reliable patterns. Start with 50 and increase if you notice too many false signals.