On learning and the evolutionarily stable strategy