[ransomnote]

 

/Rookie question sub-routine/

Wouldn't you have to use decimals if you were going to be shifting a finite number of votes and don't want a rounding error?

I think the plot of the movie Office Space is that they found that financial rounding of currency meant that rounding resulted in unclaimed stash of pennies adding up to tens of thousands of dollars.

/End of Rookie sub-routine/

~~~~~~~~~~~~~~~~~~~

/function call/

I was thinking percentages. If they didn't know how many votes to steal they didn't want to hard wire a finite steal number. So use a percentage but avoid rounding error by using a floating point.

/end function call/

[grey_whiskers]

Note the New York Slimes edison results don't have integer votes for each candidate, but an overall total and a percentage with three significant figures.

[numberonepal]

So use a percentage but avoid rounding error by using a floating point.

Exactly correct. Votes are INTEGERS. You can get percentages from integers, but you can't get (all) integers from percentages. The system seems to have functionality to cap a percentage for a certain candidate.

[darbymcgill]

/function call/

I was thinking percentages. If they didn't know how many votes to steal they didn't want to hard wire a finite steal number. So use a percentage but avoid rounding error by using a floating point.

/end function call/

It's been my experience, if it runs enough cycles, you'll see a rounding error eventually. Even with floating point..