The pattern of increased candidate percentages in large precincts does not occur in the hand-counted elections that we analyzed. Columbia County, New York — a hand-counted county — offers a good example.
In this graph of the 2016 Democratic presidential primary, there is a large degree of fluctuation both up and down on the left side of the graph. Then, by about 1,300 votes, the graph settles into a fairly even straight line, with small, random fluctuations. It maintains a basically flat line through all of the largest precincts. This is a very normal-looking CVT (cumulative vote tally) graph.
In the section of our paper, titled, A Normal Graph we detail Columbia County's protocols:
Columbia County Democratic Election Commissioner Virginia Martin says they look at every single ballot in competitive races. In an interview with the Register-Star, she describes their process: “When a voter scans a ballot, it drops down into a ballot bag at the bottom of the optical scanning machine. At the end of the day, two inspectors, one Republican and one Democrat, open the machine, pull out the bag and zip it shut.” Then they proceed to count 100% of the ballots by hand. In an interview on public radio, she said that Columbia County’s vote-counting process is focused on security, accuracy, and transparency: “We are very, very careful about the chain of custody … Everything gets hand counted, and the results that we certify are based on that hand count.” She concluded by pointing out that the hand count is open to the public.
2016 Wisconsin Republican presidential primary
In the 2016 Wisconsin Republican primary the hand-counted precincts and the machine counted precincts have very different statistical patterns. Trump received 17% less support in the largest machine-counted precincts than he did in the hand-counted precincts, while Cruz' percentage increases significantly in the largest machine-counted precincts. The pattern of increased candidate percentages in large precincts does not occur in the hand-counted precincts.
The story of how Phil Evans developed this statistical method is detailed in the section of our paper called What is that Pattern? Our expectations of the statistical patterns are determined by the Law of Large Numbers, and the methodology of the study is explained in that same section of the paper.