Automated Team Assembly in Mobile Games: A Data-Driven Evolutionary Approach Using a Deep Learning Surrogate

Many mobile games adopt autobattle systems in which the major consideration of players is how to assemble strong teams. The automated team assembly (ATA) becomes a crucial issue from different standpoints, such as assisting the game designers in performing balance analysis and directing the players to configure teams. Since the ATA is generally a combinatorial optimization problem, this article exploits the evolutionary optimizers. However, unlike the traditional problems that the evaluation functions are explicit, in the ATA, we are unable to evaluate the team strengths straightforwardly. To address this issue, we collect data from the server and build an end-to-end deep learning surrogate for estimating the team strengths. The model has a three-layer architecture of a feature embedding layer, a sequential relation layer, and a regression layer, which is able to characterize the complex dependencies between the sparse input features and the team strengths. The evolutionary algorithms are then guided by the constructed surrogate to seek for the strongest teams. Several adjustments are also made on the evolutionary algorithms to adapt it to the ATA problem with multiple constraints. Simulations on the game
Romance of the Three Kingdoms: Strategy Edition
validate a good performance of the proposed data-driven evolutionary optimizers.