math.round is only used there because the result will be copied to an integer.
The calculation is done with floats though (at least I assume because you didn’t show variable declarations).
If it weren’t for that, the result would always be rounded down. You can simulate that by rounding up if the decimal part is greater than 0.5.