Previous interactions can affect unrelated future decisions: In a line at a coffee shop, a stranger pays for the coffee of the man behind her, who then pays for the next stranger’s coffee. He’s had no interaction with other customers, and no reason to do them a favor, but he does it anyway. This is an example of crosstalk, in which previous interactions affect unrelated future decisions. And though this notion might seem natural, it had never before been incorporated into simulations of groups engaging in repeated social dilemmas. A new framework developed by computer scientists at IST Austria and their collaborators at Harvard, Yale and Stanford has changed that, and enables the analysis of the effects of crosstalk between games.