Language models manage professional sports teams across multiple seasons
GM League Arena: An Interactive Long-Horizon Task Benchmark Based on Professional Sports League Management
Computation and Language
Summary
Long-term decision-making during ongoing competition is challenging because success in the short term does not always lead to winning over time. The authors created GM Bench, a test environment where language models act as general managers of sports teams, balancing money, player needs, and competition over several seasons. They included simulations of four different sports with realistic player data and financial rules. The paper evaluates how well different language models perform in managing teams for sustained success, with some models ranking higher while maintaining strong finances.
What this means in practice
- •For game developers: Create sports management games with AI-driven team management that simulates long-term strategic decisions over multiple seasons.
- •For sports analytics teams: Test AI models’ ability to plan roster and financial moves under realistic competitive and market constraints for professional sports.
Authors
Junkai Zhou, Ziming Pan, Zhaoyi Zhang
Abstract
Long-horizon decision-making under sustained competition requires large language models (LLMs) to balance immediate investment with future resource needs. However, local success in multi-step interactions can not adequately capture cumulative decision effects or demonstrate long-term goal achievement under resource constraints and sustained competition. To address this gap, we introduce GM Bench, a benchmark for managing professional sports teams across seasons under resource constraints and competition among teams. We build the environment simulators for four sports and add financial constraints, player demands, and market mechanisms to establish a unified task across sports. Diverse rosters of comparable strength are constructed from real-world player data, and each LLM is evaluated as a team general manager to assess its long-term management capabilities. We evaluate competitive performance, cash reserves, and cross-season improvement across four sports. Over five NBA seasons, Claude Opus 5.5 ranks first among 29 LLMs, while GPT-6 Astra and GPT-6.1 Sol combine strong performance with larger cash reserves.