The paper studies a specific adaptive learning rule when each player faces a unidimensional strategy set. The rule states that a player keeps on incrementing her strategy in the same direction if her utility increased and reverses direction if it decreased. The paper concentrates on games on the square [0,1]×[0,1]\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$[0,1]\times [0,1]$$\end{document} as mixed extensions of 2×2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$2\times 2$$\end{document} games. We study in general the (...) behavior of the system in the interior as well as on the borders of the strategy space. We then describe the system asymptotic behavior for symmetric, zero-sum, and twin games. Original patterns emerge. For instance, for the “prisoner’s dilemma” with symmetric initial conditions, the system goes directly to the symmetric Pareto optimum. For “matching pennies,” the system follows slowly expanding cycles around the mixed strategy equilibrium. (shrink)