In the 2016 primary election, the Iowa Democratic Party has an unusual system in the event of a caucus vote ending in a tie: They flip a coin. There were six caucus locations on the election Monday in which unallocated delegates were assigned by a coin toss (Hillary Clinton vs. Bernie Sanders). Hillary Clinton won all six. Let's conduct a hypothesis on this case.

What are the null and alternative hypotheses? Model? Evidence? P-value? Proper size of the test? Your conclusion?