Uchicago math. Can't handle non integral mantissas or when they are greater than 6. At work, I do mostly biology, where everything is 1, 2, or 5 - close enough. One sig fig is par for the course, anything else would be an overoptimistic lie.

I'm only half joking. I'll make a buffer solution, recipe calls for 2.4 grams per liter of sodium phosphate monobasic. Ok, 2 is close enough; I'm making 500 ml, that's half, so I'll round down to 1 gram, etc etc etc.