It is long known that the law of the  maximum of branching Brownian motion is related to solutions of a nonlinear PDE, the F-KPP equation. It was established in the 70’s that there is a restricted class of nonlinear PDE’s that can be obtained from such models. Following initial work by Etheridge, Freeman and Pennington and follow up by O’Dowd and by An, Henderson and Ryzhik, this class was enlarged considerably by considering ``voting’’ mechanism built on the branching process. I will discuss discrete analogues of these developments, and show how they can be used to prove a restricted notion of tightness. In particular, I will explain why the minimum over all binary subtrees of a ternary tree, of the maximum of the associated branching random walk on the binary subtree, is tight.