Spanning trees and the complexity of flood-filling games

TL;DR

This paper establishes a relationship between flood-filling game complexity and spanning trees, providing polynomial-time algorithms for specific flood-filling problems on graphs.

Contribution

It proves that flood-filling complexity on a graph equals the minimum over all spanning trees, enabling efficient algorithms for certain flood-filling scenarios.

Findings

Minimum flood-filling moves equal the minimum over all spanning trees.

Polynomial-time algorithms for flood-filling with polynomial subgraph counts.

Polynomial-time computation of connecting subsets in coloured graphs.

Abstract

We consider problems related to the combinatorial game (Free-)Flood-It, in which players aim to make a coloured graph monochromatic with the minimum possible number of flooding operations. We show that the minimum number of moves required to flood any given graph G is equal to the minimum, taken over all spanning trees T of G, of the number of moves required to flood T. This result is then applied to give two polynomial-time algorithms for flood-filling problems. Firstly, we can compute in polynomial time the minimum number of moves required to flood a graph with only a polynomial number of connected subgraphs. Secondly, given any coloured connected graph and a subset of the vertices of bounded size, the number of moves required to connect this subset can be computed in polynomial time.