Table 1 Overview of multi-agent domains evaluated in this paper.

From: α-Rank: Multi-Agent Evaluation by Evolution

Domain Results Symmetric? # of Populations # of Strategies
Rock-Paper-Scissors Section 3.2.1 1 \(\mathrm{[3]}\)
Biased Rock-Paper-Scissors Section 3.2.2 1 \(\mathrm{[3]}\)
Battle of the Sexes Section 3.2.3 2 \([2,2]\)
AlphaGo Section 3.4.1 1 \(\mathrm{[7]}\)
AlphaZero Chess Section 3.4.2 1 \(\mathrm{[56]}\)
MuJoCo Soccer Section 3.4.3 1 \(\mathrm{[10]}\)
Kuhn Poker Section 3.4.4 3 \([4,4,4]\)
  Section 3.4.4 4 \([4,4,4,4]\)
Leduc Poker Section 3.4.5 2 \([3,3]\)
  1. These domains are extensive across multiple axes of complexity, and include symmetric and asymmetric games with different numbers of populations and ranges of strategies.