Correction: This article misunderstands how delegate selection works in Iowa, so Hillary Clinton didn’t win due to coin tosses, although they might have helped her slightly. See my comment below.
The 2016 Democratic Iowa Caucuses were so close that Hillary Clinton only managed to squeak to victory over Bernie Sanders due to lucky coin tosses. Sanders received 697 and Clinton received 701 state delegate equivalents (SDE), so Clinton tecnically won the caususes by 0.29%. However, the Des Moines Register reports that 6 of the state delegate equivalents were decided with coin tosses at precincts in Ames, in Des Moines, another precinct in Des Moines, in Newton, in West Branch and in Davenport. All the coin tosses went in Clinton’s favor. The probability of winning 6 coin tosses in a row is 1 out of 64 (or 1.6%), so Clinton got incredibly lucky.
If we exclude the SDEs determined by coin tosses, Sanders received 697 SDEs and Clinton received 695 SDEs. If the coin tosses had gone 50% to Sanders and 50% to Clinton, then Sanders would have received 700 SDEs and Clinton would have received 698 SDEs. Given these results, responsible news organizations should report that the Democratic Iowa Caucuses were decided by coin tosses and the election could have gone either way.
What is clearly needed is the release of the raw voting totals from each precinct, so that the public can see how the state delegate equivalents were determined. More importantly, releasing the raw data will allow the public to examine which candidate would have won if Iowa had a popular vote, instead of an archaic system of caucuses.
At the end of the day, it looks like Clinton will get 23 Iowa delegates compared to the 21 for Sanders, which is hardly significant out of a total of 3,636 delegates at the convention. Nonetheless, reporting that Sanders tied Clinton, rather than Clinton won could determine whether voters in other states give Sanders more consideration and start regarding him as a viable candidate for the presidency.