The science of ecology is increasingly used to support decisions in nature conservation and in land management. As ecologists, we are confident that ecology is a robust and useful science. But how true is this presumption? In this talk, I revisit three central concepts in ecology, in order to find out how robust they are. The theories to be inspected are the competitive exclusion principle (Carmel et al. 2017, Oikos), optimal foraging (Carmel and Ben Haim 2005, American Naturalist), and the intermediate disturbance hypothesis (Mackey and Currie 2001, Ecology). Surprisingly, in all three cases, the evidence against the theory far exceeds the evidence supporting it. At the same time, all three theories are doing well; they continue to be featured in text books, and are passed on to the new generation of ecologists as general laws. This situation may be understood in light of Kuhn's notion of paradigms (Kuhn, 1962). In this context, we will look at the fate of findings that contradict a paradigm. Two possible ways to mitigate this undesirable situation concern our own view of ecology, and the way we teach our students, respectively; they will be humbly proposed. For an encore, I will present a case study of world grasslands. We conducted a meta-analysis of the impact of grazing on species richness (Gao and Carmel, in review). We tested the well-known notion that old-world grasslands are long-adapted to livestock grazing, as oppose to new-world grasslands, which are much less resilient. Similar to paradigms (although smaller in scope), this notion gained wide acceptance but was rarely tested. We found that globally, grazing had a net positive effect on species richness; there was no difference between old-world and new-world grasslands.