Regarding case choice, Gary King, Robert Keohane, and Sidney Verba alert from "picking within the dependent variable". They argue by way of example that researchers are unable to make legitimate causal inferences about war outbreaks by only looking at cases in which war did materialize