not really related, but i wrote a freecells solver some time ago. it either solves the board in a couple hundred milliseconds or runs out of memory in a couple

not really related, but i wrote a freecells solver some time ago. it either solves the board in a couple hundred milliseconds or runs out of memory in a couple of minutes for unsolvable boards. one thing i've been interested in was seeing how my solvers results (i.e. # of steps, visited boards, time to solve) correlated with perceived difficulty for a human player - i didn't go that far though.