Stanford Graduate School of Business (GSB)

Ep73 “The Dangers of Group Think on Decision Making” With Adi Sunderam

The podcast episode “The Dangers of Group Think on Decision Making” features Jules Van Binsburgen, Jonathan Burke, and Harvard professor Adi Sunderam discussing how people update beliefs and the pitfalls when they restrict the set of models they consider. They explain Bayesian (Beijian) updating, illustrate with the Monty Hall problem, and argue that many real‑world errors stem not from ignoring new data but from a “dogmatic prior” that discards plausible explanations. The paper they discuss proposes a framework where decision‑makers have a limited “model set” and fit new evidence only to those, leading to increasingly convoluted stories when the true model is excluded. Examples include the early COVID‑19 lab‑origin debate, inflation‑hawk/dove Twitter fights, and stock‑market narratives on StockTwits. A striking quote: “If the truth is impossible, you are forced to give higher probability to more unlikely explanations.” Experiments cited show that giving people a concrete model (e.g., a technical‑analysis pattern) sways beliefs far more than raw forecasts. The insight suggests that organizations should surface alternative models, encourage meta‑questions about what evidence would change minds, and guard against echo chambers that cement dogmatic priors. By expanding the considered model space, firms can avoid costly mis‑interpretations and improve strategic decisions.