Inspired by the successful application of probabilistic methods for selection purposes in computer and communication systems as well as in society at large, I advocate to use probabilistic methods for selecting projects for funding, and also for paper selection at conferences, thereby also building on evidence collected in the field of research policy. I also report on real lotteries recently executed at Tilburg University to allocate research grants, before I close with some reflections on my cooperation with Joost-Pieter Katoen.