From statistics: Simpson's Paradox. Maybe a little more obscure than what some others have mentioned, but it is the idea that there can be patterns in data that

From statistics: Simpson's Paradox. Maybe a little more obscure than what some others have mentioned, but it is the idea that there can be patterns in data that appear in separate groups, but then vanish or reverse when those groups are combined. It can be a useful framework. https://en.wikipedia.org/wiki/Simpson's_paradox https://en.wikipedia.org/wiki/Simpson's_paradox