VERY good summary from Coffee & Covid Jeff Childer's - my favorite email to start each day:
I have seen a lot of chatter about potential cheating in the high profile races, especially in Michigan, Georgia, Arizona and Pennsylvania. There might have been cheating, who knows, it would sure explain bizarre results like Pennsylvanians electing a brain-damaged man to represent them in the Senate without one scrap of proof from a doctor suggesting he can do the job.
Like me, you may have seen alarming looking charts of sudden vote spikes, like this one from the Walker-Warnock race:

That spike in Warnock’s favor, right when Walker was pulling even, sure looks troubling, like it might be a timely booster injection of fraudulent ballots. It’s odd enough that someone on the Walker campaign should demand an explanation from Georgia elections officials, and then they should demand evidence if the explanation doesn’t make sense, or if they don’t get an explanation.
But WE can’t know for sure without more. I thought I’d type up a little primer of some basics about how elections and vote counting work. Some of you may already be well-familiar with this stuff and can skip ahead.
Generally, states control voting and set the election laws in their state. Counties *supervise* voting, and create procedures to manage elections. Counties are usually divided up into precincts where people physically vote, excepting early voting, mail-in voting, and drop-off voting, which happen at other designated places. Each precinct is managed by a ‘clerk’ or ‘supervisor.’
Here’s a very basic diagram for you:

Depending on the state, early votes, mail-in/absentee votes, and dropped-off ballot box votes are usually already counted by the time the polls close on Election Day. Those are usually the very first votes recorded, and are the first ones to show up on the reports right after the polls close.
The real-time vote data that streams in over the hours after polls close usually comes from Election Day voting, as data from the precincts comes in to each county supervisor’s office for counting. Those results are then transmitted up to the state.
Precints report to county supervisors of elections, and county supervisors of elections report to the state.
Precinct information doesn’t always immediately come in when the polls close. For example, some states require anyone standing in line when the polls close must be allowed to vote, even if that takes two hours after the official cut off time.
In some states, precincts use dial-up modems or wifi routers to upload their cast ballot data. Others transport memory cards taking from precinct scanning machines to the supervisor’s headquarters for loading into the tabulation software system.
Because of the electronic machines, votes come in to the supervisor’s office in precinct-level batches. So a large heavily-democrat precinct can cause an apparent ‘spike’ in votes, and would show results like in the example above. You’d want to see a smaller spike for the Republican candidate too in that case.
Cheaters know this too, so the existence of any large spike probably requires explanation, even if everything else looks right. It’s the candidate’s job to ask those questions in real time. Later, in most states, citizens can get reports via public records requests showing the vote breakdown by precinct.
On the other hand, a spike like the one shown above could signal something more problematic, like a tardy batch of mailed-in ballots, which would raise serious questions like why weren’t they already counted at the beginning of the evening?
That’s what Walker’s team might need to find out about Warnock spike.
Candidates carefully watch precinct-level reporting, which is what Lindell’s team should be doing, rather than watching state-level data coming in from news networks. The ’spike’ might have come from a well-known heavily-blue precinct and could have been expected, saving a lot of speculation.
There are several ways to “smell test” precinct-level cast vote reports. For example, well before the election, candidates know exactly how many registered democrats and registered republicans live in each precinct. Going in to the election, candidates usually already know how many people in each precinct already voted right before the election. Most of that information is available from public records in most states.
So, there are easy ways to do rough tests on by comparing a precinct’s totals against the known registered voters, to see if the reported totals makes sense. It’s not perfect, but candidates aren’t operating completely in the blind, either.
That said, early voting and — especially — mail-in voting pose problems, because there are just too many opportunities for bad actors to manipulate the system and too few ways to double-check the data.
For example, here in Florida, we recently discovered a ballot-harvesting operation in densely-populated Orange County, where harvesters were paid $10 each for ballots collected from low-income folks. And because it was trivially easy to request mail-in ballots, we suspect that harvesters were really the ones making the requests, to make sure everyone in those neighborhoods got a ballot that could be harvested.
I would be SHOCKED if this wasn’t happening all over the country. Harvesting is what I think is happening in those cases you hear about where a supervisor automatically mails ballots to every registered voter in the county. It’s not complicated.
Historically, it would have been too hard and too expensive for harvesters to swing a national election. Pre-Covid, mail-in (absentee) ballots were only provided to people showing a good reason, like living overseas, being disabled, and so forth. And it’s expensive to pay a lot of harvesters.
But now, with new covid rules, and with Soros cash and Zuckerbucks, it’s not only feasible, but likely. In fact, we’ve already confirmed at least one large, well-organized group operating in Orange County.
And that’s only ONE example of organized cheating.
The solution is same-day voting using serialized, watermarked paper ballots that are counted by hand.