In a previous post about causation and correlation, Tom Wc addressed a related fallacy, the use of anecdotes as evidence.
For example, star wide receiver Chris Carter recalled Randy Moss' greatest performance in which Moss ate nothing all day until before the game, he chugged a box of candy, Hot Tamales, before dominating offensively.
Athletes looking for maximum performance would be well served NOT to look to Moss' "training methods." Moss's nutritional program is an exception, not a rule.
Likewise, a college student who gets drunk the night before a calculus exam and gets 100% correct is not proving that alcohol enhances test performance.
We should throw anecdotes out the window and rely on statistical evidence.