When predictions are wrong, value still comes from getting other dimensions of the scenario correct (Bishop's team was right on regime change, Hormuz chokepoint, Iran attacking oil, etc.) and from learning to reweight factors for next time rather than abandoning the framework