[quote=“sbright33, post:17, topic:925432”]
I could transcribe a few of his key paragraphs. Or you could speed up the video to save time. I could make a link so you only have to listen to 5 minutes to get the idea.[/quote]
Or you could link to an unbiased source of information for once?
Just a thought.
Nooooo-Information first, discussion next.
It’s math. I could do the same math. You could too, but surely you don’t have time to waste.
That’s another topic. Unrelated. Source is our government data.
Your assumption is that the ballots that were returned and the ballots that were later sent out were from and for the same voters. You have yet to prove that.
YOU brought it up. 
OK, so I watched the bit about the Oakland County precincts data.
The claim is that as the percentage GOP the precinct is (based on party registration, I assume), the larger the deficit between Trump and the down-ballot candidates (he doesn’t say which specific candidate he is measuring against).
So, basically, he finds it unbelievable that the more GOP-leaning the district the more ticket-splitting there was.
I find this wholly unremarkable, and would need to compare it to previous elections before I’m even remotely piqued by it. He did compare it to Wayne County, but that doesn’t really help because there are basically no precincts in Wayne County that are heavily GOP.
Considering the title, I ask that you show your math.
So Techdirt discredited his claim to be the ‘inventor of email’?
Then he sued Techdirt, and the case was thrown out of court. But he’s still claiming it, so he’s blatantly dishonest.
See his Twitter account (with a warning message from Twitter about this current claim of election fraud).
He made a previous claim about election fraud, which was recently discredited by Reuters, no less.
https://www.reuters.com/article/uk-factcheck-election-ballot-massachuset-idUSKBN26N2AF
If you can’t see by now that ‘Dr Shita’ is a snake-oil salesman and phoney, or alternatively a mental case, then it’s a waste of time arguing with you.
Isn’t that what we would expect? A lot of Republican leaning people split the ticket to put a check on Trump but keep local Republicans in office.
Sounds like much ado about nothing.
Okay, I skimmed through it and assuming they picked their best possible illustrations of their point, they have nothing. Their supposed “no cheating” county is very democratic, their supposed “cheating county” is rather republican. The linear regression for the “no cheating” county is rather suspect, since most of the precincts are in the “very low number of republican voters” blob, in fact if you look at the “cheating county” you can recognize a similar pattern if you look only at the most democratic precincts.
My hypothesis is that in districts with very few republicans, the ones that are R are more likely do be diehard Trumpians, whereas the more republican a district is the larger the percentage of reluctant supporters of the administration. Fits the pattern, with zero cheating. And considering there is no evidence of widespread cheating, and lots of evidence it would be very hard to do, my hypothesis is the more parsimonious.
It’s pseudoscience by a know kook.
OK, I highly doubt that this will help our OP, but I went ahead and downloaded the data from Oakland County.
I plotted the % of the vote for Trump by precinct against the Trump deficit compared to James. This was at the precinct level, the same as the video did.
There is absolutely no linearity or anything like what is shown in the plot in the video. I will try to upload a picture if folks are interested.
So I’m really curious what the video was using as the x-axis of his scatter plot. It doesn’t appear to be % of the Trump vote. Maybe registered voter percentages? Or maybe it was just made up?
ETA: OK, now I see that he used straight-party vote as his x-axis. I have that data so I will update with that.
Thank you so much Jas! Where can I find that raw data?
https://results.enr.clarityelections.com/MI/Oakland/105840/web.264614/#/detail/190
Sorry, this. Downloadable reports are on the side.
I agree with your theory 12m ago. But why is it so smooth and linear? This does not apply to past elections.
All that would mean is that Republicans had more reason to not vote straight party in this election and vote split ticket as opposed to past elections.
Do you think there is anything about this election where that might make sense?
OK, I think I was able to re-create his chart from the data, at least somewhat. He is missing some relevant data in his plot (perhaps things were updated after he made the video). And the two things he is comparing (% GOP SP vs. %Trump - %GOP SP) is sketchy for a few reasons (mainly because subtracting percentages with different denominators never sits right with me).
I have the 2016 data, but it’s not in as nice a format to make a comparable plot. Maybe if I get bored I will. But probably not, because why bother at this point.
But either way I find this totally unremarkable.
Basically the statement is that the higher percentage of the straight-party vote went for Trump, the more he under-performed with non-straight party voters. But that is extremely intuitive.
Why would a GOP voter in a heavily GOP area not vote straight party? Because they want to vote against Trump. And why would the deficit get worse the “more GOP” a precinct is? Because the more GOP voters there are, the more ticket splitting there will be.
Thanks for taking one for the team (and for doing the work the OP was too lazy to do). Did you see anything to explain the ridiculous “ballots returned before ballots sent out” claim?
Only on some crazy websites that I’m a bit scared to even trawl very deeply. Let’s just say that it was from an anonymous data researcher submitted to a Chinese website. They claim it comes from publicly available data here: https://data.pa.gov/Government-Efficiency-Citizen-Engagement/2020-General-Election-Mail-Ballot-Requests-Departm/mcba-yywm
A quick summary shows that many of them were likely ballots picked up and returned in-person the same day. The ones with previous dates are almost certainly just data entry errors. A quick look through the list shows a mix of Democratic and Republican registrations, with a D skew about in line with what you would expect since Democrats used mail-in much more heavily than Republicans.
ETA: There was also a claim in there about birth dates older than the oldest known person in the state. Just to cut that craziness off, false birthdays (typically 1/1/1800 or 1/1/1900, but sometimes other dates) are used for anonymity, often in domestic abuse situations.
Time warp. Wormhole. Infinite Improbability Drive. Or just a really hot cup of tea.