Sora TODA

Related Paper: https://ieeexplore.ieee.org/document/10572104

In this study, I presented a research paper demonstrating how quantum annealing and Ising machines can be effectively applied to the problem of team assignment in multiplayer games. The results demonstrated that this approach enables both rapid and well-balanced team composition.

Multiplayer games involve players from around the world competing in real time by forming teams over the internet. In such games, the way players are assigned to teams significantly affects both fairness and player enjoyment. An effective team assignment system must rapidly take into account player attributes—such as preferred roles and skill levels—while minimizing imbalances in team strength.

In existing systems, team assignments can sometimes be unbalanced, which may result in player dissatisfaction. I hypothesized that this challenge could be addressed through combinatorial optimization. Quantum annealing and Ising machines, in particular, excel at rapidly identifying optimal combinations that satisfy complex constraints. This makes them highly suitable for addressing team assignment problems in online games.

I formulated the team assignment as an optimization problem. The objective was to maximize player-role preferences while minimizing variance in team strength. As a result, I was able to form well-balanced teams, notably, in a very short time.

Although I was a beginner in Python, I found the Fixstars Amplify library to be highly intuitive. Both objective functions and constraints could be implemented with ease. In particular, the helper functions for defining constraints proved to be extremely useful. While tuning the weights of constraints and objectives required some trial and error, recent tutorials now provide more practical guidance and tips^*1, making the process even more accessible.

I am currently conducting research on applications of quantum computing in the context of the quantum internet—specifically, on methods for establishing connections between arbitrary endpoints. Should I encounter another combinatorial optimization problem in the future, I would strongly consider using Fixstars Amplify again.

*All information in this article is based on the interview conducted in April 2025.