Solving the (n² - 1)-Puzzle with 8/3 n³ Expected Moves (open access)

Solving the (n² - 1)-Puzzle with 8/3 n³ Expected Moves

In this article, it is shown that the greedy algorithm for the (n² - 1)-puzzle makes 8/3 + O(n²) expected moves. This analysis is verified experimentally on 10,000 random instances each of the (n² - 1)-puzzle for 4 ≤ n ≤ 200.
Date: July 10, 2015
Creator: Parberry, Ian
System: The UNT Digital Library