Development of solving algorithms of Minesweeper

Abstract

The aim of the research is to study the automatic solving algorithms of Minesweeper and to further develop the Single Point Strategy. This procedure is the simplest and fastest solving algorithm, since it makes a decision about the next tile to be uncovered based on the examination of a single tile. During the investigation, we implemented 4 versions of the Single Point Strategy. The basic algorithm stops if it cannot proceed with 100% certainty. The other three algorithms use different guessing strategies when they cannot make a safe decision. The presented results prove that the effectiveness of our best method approaches that of the benchmark CSPS backtracking algorithm at the beginner and intermediate levels.