Research map / Multi-agent systems and games / Games and equilibria
Diverse Generation for Multi-Agent Sports Games
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use a permutation-equivariant graph variational recurrent model with agent types | Sports trajectory forecasting · Basketball | L2 (Best): 09.02 ± .02 -> 08.26 ± .04 | Helped |
| Use a permutation-equivariant graph variational recurrent model with agent types | Sports trajectory forecasting · Soccer | L2 (Best): 3.17 ± .01 -> 3.07 ± .02 | Helped |
| Add agent type information (ablation) | GVRNN · Basketball | L2 (Best): 08.39 ± .03 -> 08.26 ± .04 | Helped |
| Add interactions between agents using a full graph (ablation) | GVRNNT · Basketball | Cond. L2 (Best): 10.63 ± .04 -> 07.88 ± .03 | Helped |
| Add variational latent states (ablation) | GRNNT-Full · Basketball | L2 (Best): 11.24 ± .03 -> 08.26 ± .04 | Helped |
| Add agent type information (ablation) | GVRNN · Soccer | L2 (Best): 3.25 ± .01 -> 3.07 ± .02 | Helped |
| Add interactions between agents using a full graph (ablation) | GVRNNT · Soccer | Cond. L2 (Best): 7.77 ± .10 -> 7.61 ± .13 | Helped |
| Add variational latent states (ablation) | GRNNT-Full · Soccer | L2 (Best): 3.05 ± .01 -> 3.07 ± .02 | Hurt |
| Use a variational recurrent model instead of a deterministic recurrent model (ablation) | Sports trajectory forecasting · Soccer | L2 (Best): 3.05 ± .01 vs 3.07 ± .02 | Mixed |
| Use a skip connection from previous observations into the graph networks (ablation) | GVRNN · Basketball | performance | Helped |
| Add a GNN for direct interaction between recurrent hidden states (ablation) | GVRNN · Basketball | performance | No clear effect |
| Use permutation-equivariant graph modeling instead of fixed role-based ordering | Multi-agent sports forecasting · Basketball | L2 (Best): VRNN-Template 09.02 ± .02; GVRNNT-Full 08.26 ± .04 | Helped |
Closest papers by meaning
- Semi-Supervised Generative Models for Multiagent Trajectories (NeurIPS 2022 · similarity 0.90)
- Sports-Traj: A Unified Trajectory Generation Model for Multi-Agent Movement in Sports (ICLR 2025 · similarity 0.88)
- Stochastic Prediction of Multi-Agent Interactions from Partial Observations (ICLR 2019 · similarity 0.88)
- Event2Tracking: Reconstructing Multi-Agent Soccer Trajectories Using Long-Term Multimodal Context (AAAI 2025 · similarity 0.88)
- Unified Uncertainty-Aware Diffusion for Multi-Agent Trajectory Modeling (CVPR 2025 · similarity 0.86)
- The CoachAI Badminton Environment: A Novel Reinforcement Learning Environment with Realistic Opponents (Student Abstract) (AAAI 2024 · similarity 0.82)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.