Om! Nom! Nash!
summary
An AI solver for Om Nom Nom claims optimal play by combining Nash equilibrium reasoning with tree search implemented in Python; the raw branching is 36 then 25 then … totaling ~600k states, computing equilibria takes minutes, so memoization cuts distinct states to ≈14,000 and clever canonicalization using board symmetries shrinks solve time to a few seconds. The solver works for guaranteed-best strategies but misses high-variance “Hail Mary” plays where a risky seven beats a steady four.
AI-written summary · about
