You paid a fortune for a non-refundable ticket, the show is awful within twenty minutes, and you sit through three more hours anyway because you already paid. This episode explains why. It starts with the economic baseline: the distinction between sunk and prospective costs, the bygones principle, and a worked factory example where 30 million dollars already spent should be ignored and the only question is whether spending 70 million more is worth the asset you get back.

The hosts then show how that logic collapses in practice, from the Concorde supersonic jet that Britain and France kept funding after it was clearly a commercial disaster, to nuclear plants utilities refused to abandon in the 1970s and 80s, to research showing people stay in failing relationships because of time already invested. They distinguish genuine sunk cost errors from incentive problems, cover plan continuation bias in the Torrey Canyon oil spill and NASA aviation accident data, and explain the psychology underneath: the 1968 racetrack study where placing a bet inflated confidence, Staw’s R&D simulation on personal responsibility, Kahneman and Tversky’s framing effects, and neuroeconomic evidence that mice and rats fall for the same trap.

• Why finishing the tasting menu when you are full means suffering twice
• How cognitive dissonance makes pilots invent optimistic estimates mid-flight
• Why a manager protecting his bonus is rational for himself but bad for the company
• The evolutionary case that stubborn persistence once kept our ancestors alive
• How stress dials down the prefrontal cortex and makes you double down