Many papers have been published for solving paradox of voter turnout. In this paper, we adopt the strategy method to mimic Levine and Palfrey (2007)' voting games. We find the underdog effect and competition effect supported by our data, but we cannot replicate the Levine and Palfrey (2007)’ result for neither the strategy method, nor direct response method. We also find evidences indicates that subjects don't use fixed cut-off strategies. Finally, out data shows that voters are highly responsive to historical pivotal event, which the most important implication of rational choice model.