1: \begin{abstract}

2:         Constructing effective algorithms to converge to Nash Equilibrium (NE) is a important problem in algorithmic game theory.

3:         Prior research generally posits that the upper bound on the convergence rate for games is $O\left(T^{-1/2}\right)$.

4:         This paper introduces a novel perspective,

5:         positing that the key to accelerating convergence in game theory is “rationality”.

6:         Based on this concept,

7:         we propose a Dynamic Weighted Fictitious Play (DW-FP) algorithm.

8:         We demonstrate that this algorithm can converge to a NE and exhibits a convergence rate of $O(T^{-1})$ in experimental evaluations.

9:         \keywords{Game Theory  \and Rationality \and Fictitious Play \and Dynamic Weight.}

10:     \end{abstract}

11: