Bayes Theorem and the Case of the Busted Alibi

  • July 25, 2019, 4:30 pm US/Central
  • User's Center
  • Steve Kent
The oldest cold case murder in US history ever to be solved occurred right here in Northern Illinois, in the city of Sycamore.  The crime involved the kidnapping of a young girl in 1957.  The story made national news.  The FBI entered the case in force and conducted an exhaustive investigation.  However, no suspect was ever charged. The case went cold for over 50 years.  It was revived by the Illinois State Police (ISP) in 2008 when, acting on a tip, a new suspect, Jack McCullough, was fingered.  After three years of investigation, he was arrested, tried, convicted, and sentenced to life in prison.  Finally, justice had been served.
 Or had it?  He claimed he had an alibi.  The ISP countered that it had been busted.  Who was right?  This talk will conduct an in-depth re-examination of McCullough’s alleged alibi from 1957, this time using Bayesian statistical methods, familiar to us but not previously considered by the FBI or the ISP.  Does McCullough’s alibi “hold water?” Or did the ISP nab “the right guy?”