2 ms·
"You just look at the probably of all events that can happen not just one outcome." You're probably committing the Gaussian fallacy, which is to assume that al
by jerf 10y ago
"You just look at the probably of all events that can happen not just one outcome."
You're probably committing the Gaussian fallacy, which is to assume that all events are always drawn from a Gaussian distribution. Drawing a series of events from a Gaussian distribution centered around zero will tend to at least stick close to zero, sure, but there's no reason to assume that fraud distribution will be Gaussian. Especially in the face of hostile and intelligent entities who know you're making that assumption and will abuse you with it. Very little math or math intuition is structured for dealing with intelligent hostile exploitation. (Game theory, of course. And a lot of math involves assuming the worst of your counterexamples, but that doesn't really come from an intelligence with goals, even if we sometimes are tempted to think otherwise.)