Study of valuation function and search strategy in random game

The balance between attack and defense is one of the key problems of computer game system. And Einstein Würfelt Nicht! game injects randomness of the game process with the aid of the dice for choosing, increasing the game tun, but also making valuations harder. For this reason, this paper selects two valuation factors of distance and probability, and designs the valuation function specially for this randomness. Introducing multi-thread technology, a Hybrid Optimization Algorithm based on Monte Carlo algorithm and Expected Search algorithm is proposed. From the perspective of system balance, the game system of Einstein Würfelt Nicht! is reconstructed based on the two aspects of attack and defense. The experimental results show that the improved game system continued to maintain a strong chess force and won the national championship.