Predicting Future Lifespan: The Lindy Effect, Gott's Predictions and Caves' Corrections, and Confidence Intervals

• Activity Number: 33
• Type: Contributed
• Date/Time: Sunday, August 3, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Colman Humphrey*+
• Companies: Wharton School
• Keywords: Gott ; Caves ; Lindy ; Jeffreys ; Copernican ; Prediction

Abstract

Old things live long, young things die young: a view Albert Goldman first acknowledged about comedians' career spans - the Lindy Effect. Richard Gott wrote a paper on predicting future life spans [Nature 363, 315 (1993)], inspired by the Berlin Wall. He formed confidence intervals based on a form of the Copernican principle, which states that we are (unless otherwise informed) not in a special place or time. Later, Bayes' techniques were used to form the same intervals, using a natural Jeffreys prior. Carleton Caves has produced a well-thought out critique of Gott's methodology [arXiv:0806.3538 [astro-ph]]. We intend to meld the critique together with its original paper, to see from what point of view are the predictions Bayesian in nature, and from what point of view are they not. Also we discuss how these topics pinpoint the philosophical differences between confidence intervals and credible intervals.