Shiva seems to be a huckster. I don't trust what he says about anything.
Why didn't he show us a normal, unmanipulated county? All he's really proved here is that if you make the voter pool smaller by removing reliable trump votes, trump doesn't do as well.
He actually did show a county that appeared normal for comparison. It was Wayne or Kent county I believe. Watch it again. The difference between the counties is very obvious manipulation was going on with votes.
Kent has the same downtrend. Wayne is nothing like a republican county so using them as a baseline doesn't produce results because there are no (or very few) data points past where the algo would show up.
But they didn't apply this test to it. They showed how a manually counted county didn't have the curve that machine counted counties have. It was one of the first slides.
Shiva seems to be a huckster. I don't trust what he says about anything.
Why didn't he show us a normal, unmanipulated county? All he's really proved here is that if you make the voter pool smaller by removing reliable trump votes, trump doesn't do as well.
He actually did show a county that appeared normal for comparison. It was Wayne or Kent county I believe. Watch it again. The difference between the counties is very obvious manipulation was going on with votes.
Kent has the same downtrend. Wayne is nothing like a republican county so using them as a baseline doesn't produce results because there are no (or very few) data points past where the algo would show up.
They discussed a county with manual ballot counting and highlighted the difference
But they didn't apply this test to it. They showed how a manually counted county didn't have the curve that machine counted counties have. It was one of the first slides.
Okay Eeyore.
I'd love to be wrong here but he gave us no controls to look at. Showing that dems started voting for trump in Wayne County isn't anything new.