Reproduce “Deep Reinforcement Learning in Ice Hockey for Context-Aware Player Evaluation” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Guiliang Liu, Oliver Schulte
GIM estimates how recorded hockey actions change a team’s probability of scoring the next goal and sums those changes for each player. An LSTM incorporates game context and play history. In 2015–16 NHL data, GIM correlated 0.878 with goal totals, versus 0.745 for SI and 0.752 for GIM-T1; its scope is recorded actions, not complete or causal impact.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.