In this paper, we re-examine various previous experimental studies of the Centipede Game in the literature. These experiments found that players rarely follow the subgame-perfect equilibrium strategies of the game, and various modifications to the game were proposed to explain the outcomes of the experiments. We here offer yet another modification. Players have a choice of whether or not to believe that their opponents use subgame-perfect equilibrium strategies. We define a `behavioral equilibrium' for this game. This equilibrium concept can reproduce the outcomes of those experiments.