You know, the one about the probability of two people in a group having the same birthday. The probability exceeds 50% when there are 23 people in the group.
An explanation:
[And yet, it's not true, because in the RealWorld births are not evenly distributed.]
The limited variation in birthdate frequencies has a negligible effect on the overall result.
See also ProgrammerMathSkills and InevitableIllusions.