The Most Important Thing: Uncommon Sense for the Thoughtful Investor (Columbia Business School Publishing)

by Howard Marks

It’s impossible to assess the accuracy of probability estimates other than 0 and 100 except over a very large number of trials.