Denrell is Professor of Behavioural Science at Warwick Business School, previously at Oxford’s Saïd Business School, and his research programme is a sustained argument that organisations learn the wrong lessons from success. Sampling bias, survivorship, and regression to the mean are his standing subjects, applied to strategy and organisation theory rather than to laboratory judgment.
The paper cited here, “Top Performers Are Not the Most Impressive When Extreme Performance Indicates Unreliability” (PNAS, 2012, with Chengwei Liu), is the formal version of an argument this page needs. Where performance is noisy, and startup outcomes are noisy, extreme results are evidence of high variance as much as high ability, so the highest performer may not have the highest expected ability, and someone one rung down may be the better model. The practical instruction that follows is the one the page gives: do not build your standard from the single most celebrated person you can reach. The cited paper is analytical models and simulation rather than fieldwork, and his weight is modelling, not operating experience; he is a researcher whose work has been borrowed into management, not a management thinker by vocation.