The Curious Case of Kory for County Clerk

This is another important post on the 2022 CO R primaries and the prerequisites for this post are that you go read both the “How the Machines Mauled MAGA” and “Pasco County, FL” posts first to fully make sense of it. Or not.

We open with the headline chart for Kory Nelson (R-MAGA) for County Clerk in the 2022 Douglas County, CO R primary:

He lost of course and notice all those red boxed surges in forgetting to vote which almost always seem to correlate to lower vote accumulations for Nelson (red) but not Davis (blue). And a trend toward the end to more and more voter forgetfulness together with apparent downward pressure on Nelson (and more places we could put less glaring red boxes). The general level of missing votes in this race runs at the 10-15%+ level which is a big red flag indicated with high magenta dots across the full length of the chart as opposed to just the odd few (dots appear at over 6% missing). I think we can already make a good guess who was openly talking about election integrity regarding how they would be running the Clerk’s office now, can’t we?

Which you can compare to this chart for Hanks (of multiple charts including Lopez and Peters showing the same thing) in the Garfield MAGA races:

People keep forgetting who they were going to vote for in temporal surges when it comes to MAGA candidates for some reason in Garfield … and now it has spread to Douglas too!

And remember the R overlaid on D primary vote progress overview chart for Garfield?:

We noted in Garfield that the down ballot races (mid height line group with heavy lines above) also curiously took on the same “all of a sudden forgetting to vote” vote progress temporal curve pattern as the MAGA races! This makes sense when you go back and look at the start of the Pasco post which explains that the way votes are selected by the algos is that entire ballots up and down the races are dropped.1 And so even the highly unlikely to be MAGA dogcatcher (in Garfield we used the example of the coroner) can take it in the shorts re their total vote count!

Now remember that other than the (non-trivial) new discovery of variable ratio MAGA temporal forgetting to vote, Garfield WAS NOT a fixed ratio data leak like Pasco where entire ballots are blank except for odd end of ballot patterns or like Clark/ Las Vegas where numbers are missing (gaps) from the otherwise correct (though inflated) CVR number sequence. But it turns out that this damage to the lower uncontested races becomes another tell-tale signature of an N:1 leak!

In a smaller county like Garfield, the waveform was of a significant enough magnitude (due to my sampling resolution bandwidth and the performance and settings of the algos) that the “forgetting to vote” stood out like a sore thumb. They were really hoping we wouldn’t look at the smaller counties to see how glaring it was there. Because in much larger Douglas we see the below which people might have an easier time believing is just noise regarding the line group pointed at by the arrow:

Except WOW: Are all those down ballot unopposed races pointed to by the arrow ever having everyone forget to vote for them in a crazily varying but completely stacked on top of each other consistent pattern — we don’t even need to run a Pearson’s test on that correlation to know it’s brutal. And when you look down at the MAGA races (next group down) there looks to be a lot of though not perfect visual correlation in the waveforms with those down ballot unopposed races (pointed to by the arrow) except the Peters race (blue, which was a 3 way — I need to write about why they like N-way and ranked choice sometime) stands out a bit less.

What we’re really being asked to swallow here is that the “noise” of R unopposed race missing votes really is just noise and we shouldn’t be surprised that R’s (unlike Dems!) just very heavily skip over uncontested down ballot races and only vote in the contested up ballot races! Oh, err, except for including skipping down and voting heavily in Kory Nelson’s down ballot race for County Clerk too!2 See the heavy solid lines up high for total of all candidates votes in R contested races that are at the uncontested Dem levels of total voting percentages (dashed lines are Dems and they are ALL unopposed if you didn't know — but not forgetting to vote like Rs!) as proof that only when R’s have a choice they vote just like Dems!

Most of you who were paying attention are hopping up and down shouting epithets at this point LOL. You know what’s going on. But how can we highlight it?

Back to signal processing: We are being asked to believe that the uncontested missing votes are just noise and carry no signal. Those dumb Rs. So let’s test for signal using Pearson’s correlation test: If those missing ballots are just random then they should not correlate to anything we care about like MAGA vs RINO in the major races, right? Then why don’t we check if there is a correlation between missing votes for down ballot uncontested (just have to pick any uncontested candidate out of the arrowed group as a baseline since they are all incredibly the same) and missing votes for up ballot candidates when MAGA vs RINO.

Here’s the correlation between Dave Gill for County Treasurer (unknown inclinations but part of the arrowed group) not getting a vote and Ron Hanks (MAGA) vs Joe O’Dea (RINO) not getting a vote on the same ballot:

Umm. I think we can see both visually and in the Pearson test values that the missing vote “noise” we’re supposed to ignore is impressively correlated with Hanks-R/MAGA. When there is a missing vote for Gill it is highly correlated that there will be a missing vote for (MAGA) Hanks. Full stop.

Remember that what we see for ballot records in these down ballot races were “let through” by the machines in the process of removing over +/-10,000 of them (depending on how many legit undervotes there should be as more closely illustrated by the Dem down ballot races) for every down ballot race in Douglas.

And yes, we see this across all MAGA vs RINO in the upper races in the CO primaries. This post would be interminable if I posted every test I have run so far.

But let’s not forget about the Kory Nelson (of title fame) vs Sheri Davis race (of course Davis was the winner and is a current employee of the clerk’s office in case you needed more RINO credentials):

Oh. My. Full. Stop.

Looks to me like the machines thought Nelson was MAGA and if you look at his campaign messaging you can confirm it.

That will teach Kory to try to get elected to replace Skynet when Skynet runs the elections!3

And do we see this anywhere else in CO in local races as opposed to the upper ballot races? Why yes, yes we do. We see it in the Larimer Sheriff’s race where Jeff Fisher was the un-mistakeably Constitutional candidate in messaging as well as this chart but (surprise, surprise, surprise) lost:

There is a signal heavily targeting MAGA in this “forgot to vote” “noise” in the uncontested R primary races. And this is NOT based on whether the voters themselves perceived candidates as MAGA or not; it is based on what the machines were programmed to treat them as. And we don’t see ANY of this in D races (this test across D races basically jumps around trying to correlate actual random noise) including the fact that D voters only forget to vote in a reasonable and low incremental progression as they near the end of the ballot completely in line with what election process subject matter experts would expect.

Also, based on this analysis there are various takes I am working on to reconstruct what the results would actually be without the algos which all result in the losing MAGA candidates in CO being at least in much closer striking distance if not winning. And of course the MAGA candidates almost always lose in CO except in the deepest of red counties and often even there as you would expect based on what you know now about the machines.

If my analysis is not correct, then we are talking about being well on our way toward life of the universe4 correlated statistically minuscule probabilities that I’m wrong -- p-values can shift by 20 orders never mind the visuals LOL -- although I welcome reproduction of these results of course!

And how I wish someone would prove me wrong because I would much rather live in that world…