1: \begin{abstract}

2: \begin{spacing}{1.2}

3: \noindent Potential based no-regret dynamics are shown to be related to fictitious play. Roughly, these are $\eps$-best reply dynamics where $\eps$ is the maximal regret, which vanishes with time. This allows for alternative and sometimes much shorter proofs of known results on convergence of no-regret dynamics to the set of Nash equilibria. 

4: 

5: \medskip

6: \noindent\emph{Keywords:} Regret minimization, no-regret strategy, fictitious play, best reply

7: dynamics, Nash equilibrium, Hannan set, curb set

8: 

9: \medskip

10: \noindent\emph{JEL classification numbers:} C73, D81, D83 

11: \end{spacing}

12: \end{abstract}

13: