Combinatorial games pose an extreme challenge to combinatorial optimization. Several combinatorial games have been shown to be PSPACE-hard and many more are believed to be so. In this paper, we present a new approach to analyzing combinatorial games, which differs dramatically from current approaches. Using the combinatorial game Chomp as a model system, we employ ideas from physics and dynamical systems theory to unveil deep connections between such games and nonlinear phenomena commonly seen in nature.
展开▼