2 comments

[ 6.6 ms ] story [ 18.0 ms ] thread
The article ends with "0.538% batting average — exceptional by all measures."

Ignoring the extra % in there, a .538 batting average is exceptional... if you are playing baseball. In this case, you are not playing baseball. Ogilvy's predictions were barely over 50% correct, which is actually pretty bad if you think about it. The article's conclusion is completely off.

50% isn't impressive in a binary true/false scenario, but I'd say it's impressive when making unbounded predictions.