News for hackers. Stuff that matters.
Re:German tank problem by sbirch on Friday February 21, 2014 @01:00PM (#7278292)
My favorite explanation of this (posed instead as the Locomotive problem) is in Allen Downey's "Think Bayes," pp.22

It's online too, and worth reading!

http://www.greenteapress.com/thinkbayes/

Re:German tank problem by jmount on Friday February 21, 2014 @01:58PM (#7278666)
Book seems nice, but its discussion on the German Tank problem sets up code to calculate posteriors from priors that are more detailed than the Bayesian argument in the Wikipedia article. That is useful, but you don't get the problem intuition that the answer is the maximum ID you saw inflated by a factor depending on the expect ID-gap. There are some assumptions in the Wikipedia Bayesian analysis, but they are less determining than the ones in the book.