>>Here's what NPR had to say on the subject:
>>http://www.npr.org/2016/02/02/465268206/coin-toss-fact-check-no-coin-flips-did-not-win-iowa-for-hillary-clinton
>>
>>and the WaPo article was updated to say that, in fact, the distribution of all coin tosses was closer to 50/50.
>>
>>Which begs the question: Do you really think that coin tosses should play a part in the election process?
>
>Fair question. What should happen when an election ends in an exact tie? That rarely happens in elections involving thousands, ten of thousands, or more voters, but it certainly does come up for local elections. I think a lot of places use the equivalent of a coin toss. (http://fivethirtyeight.com/datalab/the-2014-elections-that-ended-in-a-tie/ bears that out.)
>
>In this case, it's important to note that the coin toss was for allocation of the last delegate of a group of delegates. So in each case, one delegate (to the county convention, if I understand what I read correctly) was determined by a coin toss. All told, the Dem caucuses chose 11,000 delegates and change and about a dozen were selected by coin toss.
>
>In all seriousness, what would you suggest when there's an exact tie?

I suppose you could just do what Bush did - have your brother recount the votes until you get the result you want..hahahaha