The downward slope is expected, not a horizontal line. In his own words he said that in the 45% straight ticket R precinct, he expects 45% Trump votes from the non-straight ticket voters. Instead there was 25%.
But the other 55% voters are not R voters. That's his claim and his claim is wrong.
Also ridiculously easy to disprove, just hand count one of the lower right precincts. Just 2,000 votes to count at the allegedly affected precinct.
It is plotted as the difference between the percentage of Trump votes for that precinct which has the straight line ballot percentage subtracted from it, which is then shown as a deviation up or down from the straight party line votes.
You may take issue with the assumption that the percentages of the straight line ballots for both Democrat and Republican votes in a precinct may be used as a proxy for the real percentage of Democrat and Republican voters in a precinct, but for the purposes of this, I believe it is a worthy assumption.
It is the UNIFORM downward slope through all of the counties looked at EXCEPT Wayne county (which is largely Leftist) that is damning.
THAT slope, reproducible in each of the three largest precincts, is NOT a natural pattern. It is not only unnatural, but with that uniformity could likely be plotted to show the actual formula used in the algorithm.
I don’t disagree that it may be “ridiculously easy to disprove”. Sure. Then why hasn’t that been done and released? Or is just the reluctance to do just that a telling sign of how the rule of law is being steamrolled by people who have a vested interest in doing so?