Let’s give an illustration of this where say Clinton is ahead of Sanders in every district by just a smidgen. How do the rules amplify this split. The districts with an even number of delegates erase the difference in giving an even split, but the odd number of delegate districts lean to Clinton by one delegate, like 5 gets split 3 for Clinton and 2 for Sanders.
In the California Primary, of the 53 districts, there is one 4, 22 6s, and 3 8s, for a total of 26 even splits. The odd number districts are 17 5s, 9 7s, and one 9 for a total of 27 districts that will give Clinton a total of 27 extra delegates over Sanders, where no excess should really be deserved. Thus of the 317 district delegates, 145 would go to Sanders, and 172 would go to Clinton. The 158 other pledged delegates would split evenly if the vote difference was only a smidgen. So of the 475 pledged delegates, 224 would go to Sanders and 251 would go to Clinton. Of course if Sanders led Clinton everywhere by a smidgen, the numbers would be reversed. Of the 475 delegates, the lead by 27 is 27/475 = 5.7%.
A similar analysis of the New York Democratic primary’s 27 districts has for odd numbers of delegates: 4 5s and 5 7s. For even numbers of delegates it has 18 6s. This leaves an excess of only 9 delegates from districts with 5 or 7 delegates, which out of 247 pledged delegates gives an smaller excess of 9/247 = 3.6%, but still significant if there were no differences in the district votes.
In order to spend campaign time and money most effectively, it makes no sense to invest in districts with 6 delegates which will end up with 3-3 delegate splits, unless the vote split is greater than 16.6%, as shown in the previous article.
It makes a lot of sense in districts with an odd number of delegates to fight hard, since any lead in votes yields an extra delegate over your opponent.
Since the average number of delegates per district is 6, the Democratic Party was effective in taking districts with more Democrats and giving them 7 delegates, making them more sensitive deciders of a states winner. However, the party erred by taking districts with fewer Democrats and giving them 5 delegates, since they are now as effective in picking the state leader as the districts with more delegates. They could have given districts with medium and larger number of Democratic voters 5 and 7 delegates, respectively, and districts with a low number of Democrats 4 delegates.
