Building on a recently developed methodology for sensitivity analysis that parametrizes omitted variable bias in terms of partial R2 measures, I propose a simple statistic to capture the severity of omitted variable bias in any observational study: the probability of omitted variable bias...