It's like sudoku but rather than solving the puzzle by placing in numbers, the player is given the fully solved puzzle and has to *prove* the solution. Players select one cell to prove from the completed puzzle, then select other cells as evidence to prove their initial selection. The program knows once enough evidence is found and the the player is prompted to lock in that answer and move on to another cell. The goal is to complete the puzzle using the least amount of evidence as possible, and see if you can finish using less evidence than the logical computer program does.