MultiGrid Sports offers research-grade multi-agent gridworld environments based on Gymnasium for reproducible RL experiments. It includes simulations for soccer, basketball, and American football with features like competitive play, solo training, curriculum learning, and reward shaping. Licensed under Apache-2.0, it is intended for machine learning researchers.
In the Data science & ML workbench space, multigrid-sports takes a focused approach. It focuses on creating reproducible multi-agent reinforcement learning environments for sports simulations. multigrid-sports is an open-source project aimed at AI researchers. The project is open source (Apache-2.0). The product ships for the command line, and it can be self-hosted.
Abdulhamid Mousa builds and maintains multigrid-sports, and the product first shipped in 2026. The project is developed in the open on GitHub with 9 commits in the last 90 days. Among its 4 catalogued features are Multi-Agent Environments, Gridworld Simulations, and Reinforcement Learning.
Latest indexed changes and source events
multigrid-sports verified by the PulseGate indexer
Other apps tracked under the same category.
