Sequential games like tic-tac-toe and the centipede game are different from simultaneous games because players can see their opponent's moves. In these games, players can use a strategy called backward induction to find the best moves and predict the outcome. By looking at the last player's best decision and working backwards, players can determine the optimal strategy for winning these games. This method can be used for any finite, two-person sequential game of perfect information.