Deal survival model

A deal wins only if it survives every approver in turn. If the tests are separate, multiply. Five approvers each at 0.8 give 0.8 to the power 5, which is 0.328: about one in three, not 80 in 100.

Add a sixth approver at 0.5 and the deal halves again. The blocker was always there. Learning about them lowered your forecast to meet the true chance, which is why a good map makes forecasts fall, and why that is cheaper to learn in week three than in week twelve. A seat is dangerous because its approval chance is low, not because the person is powerful.