That’s basically what we find. In Democrats vs Miscellaneous, the picture is even more crazily skewed than before. Democrats improve relative to Misc. in 520 ward/race observations. They tie 89 times, and Misc. improves in relative terms just 3 times. That’s not a typo. (21/N)
4
122
0
695
This corresponds to p-values between 10^-73 and 10^-177. The fraction of Democratic “wins” here (520/523), excluding ties, is a ludicrous 99.4%. (22/N)
3
152
2
712
So how do Republicans compare with Miscellaneous? While they’re not exactly the same, they’re far closer to each other than either is to the Democrats. Other than a few outliers (as Misc. has very few votes in total), the distribution is fairly symmetric around zero. (23/N)
5
108
1
628
Republicans improve relative to Miscellaneous 179 times, Misc. improves 251 times, and there are 74 ties. The p-value you get depends greatly on how you allocate the ties. Give them to M, and it’s 10^-11. Give them to R, and it’s 0.55, almost exactly chance (253 vs 251). (24/N)
3
88
1
620
Excluding ties, the R “win” percentage is 41.6%. So under some measures, they look slightly worse, but this is affected by questions of rounding and the small vote totals for M. What’s incontrovertible is that D looks wildly, wildly different from either of them. (25/N)
3
95
0
610
This is exactly what we’d predict if votes before look like votes after, which for R vs M, they do. This is also inconsistent with the driver being something Trump did, like telling all his supporters to vote in-person. (26/N)
1
87
0
618
If so, why do changes in Miscellaneous votes look about the same? The important difference after Tuesday night, whatever you think it is, is coming on the Democrat side. (27/N)
2
90
1
620
Could there be other reasons than fraud why ballots might be different before and after? If the ordering is random and drawn from the same pool, no. But if wards count different types in a different order (votes cast at 9am vs 4pm, in-person vs mail-in), then possibly. (28/N)
4
81
0
585
Whatever is changing vote distributions before and after, it’s overwhelmingly impacting Democrats, not Republicans. If you think it’s about in-person vs postal voting, Republicans must be similar to Miscellaneous in this respect. This is possible, but not at all obvious. (29/N)
3
99
0
639
There’s another more important aspect. If Democrat increases are partly fraud, we would expect that increases should be larger *when the fraud is more likely to impact the race*. We have lots of down-ballot races like State Assembly Representatives we can test here. (30/N)
3
86
0
581
Sometimes the Democrat is way up after early counting, so improving the margin doesn’t matter. But if the Democrat is down early on, adding votes is much more important. I’m assuming fraudsters would like to win as many races as possible with the least amount of fraud (31/N)

5:08 PM · Nov 7, 2020

11
94
2
652
The comparison is now between two different races at the same ward. A Democrat voter comes to the ballot box or mailbox, and sees a number of races. For some, like President, it’s going to be a close call. For others, it might be a heavy favorite for the Democrat. (32/N)
1
63
1
399
The voter is a Democrat, so presumably he’s inclined to vote Democrat for both. We can compare within a given ward which of the two races showed bigger improvement for the Democrats in that particular ward after Tuesday night. (33/N)
2
62
0
393
Indeed, the increase in Democrats relative to Republicans is significantly higher when the Democrat is doing worse overall in early counting. Within each ward, late votes break more heavily to Democrat in exactly those races where they are likely to affect the result. (34/N)
39
458
59
1,221
How big is the effect? There were 8 races where Republicans were ahead on Wednesday morning. By Thursday night, half had flipped to Democrats. By contrast, there were 19 races where the Democrat was ahead, and not a single one flipped Republican. (35/N)
2
145
7
666
It’s not that the races flipped because heavy Democrat wards started reporting in. More votes started coming in for Democrats relative to the ratio for that exact ward the previous night. The votes also skewed more for races that Democrats looked like they might lose. (36/N)
1
106
7
573
This is surprisingly hard to explain with common explanations for why Democrats pulled ahead overall. E.g. mail-in ballots are counted late, and these are more heavily Democrat. In general, this doesn’t explain why some races later skew Democrat more than others. (37/N)
10
85
0
524
The key is that for each voter, voting by mail is common to all races. A single voter can’t vote for some races by mail, and others in person. So if the overall D skew is a mail ballot effect, most versions of this predict that all races should be equally affected. (38/N)
5
79
4
504
Consider a simple example where everyone votes straight ticket. More Democrats vote by mail, and these are counted late. This would predict overall Democrat improvement, but it should be the same for all races, regardless of whether the Democrat is ahead or behind. (39/N)
3
70
1
495
More ballots come in Democratic, they each vote for every Democrat, so all Democrats increase in the same percentage terms. This isn’t what we find. In the data, within a ward, the important races go up more than the unimportant races. (40/N)
4
101
5
513
The prediction that all races should be equally affected holds for many variations. Does the answer change if every Democrat voter has a 90% chance of voting for each Democrat candidate, if this attitude is the same those who vote in-person vs by mail? No. (41/N)
2
59
0
419
The answer also doesn’t change if Democrat voters generally vote less for shoo-in candidates, but vote more in tight races. If holds equally for Democrats who vote by mail vs in person, there should be no difference across races in how much they break towards Dems. (42/N)
1
60
0
407
You need something complicated to explain it. Dem voters vote less for Dem candidates if they know they’re know are going to win anyway, AND this instinct is somehow larger in Dem mail-in voters than Dem in-person voters, AND Dems vote more by mail overall. (43/N)
1
62
4
438
If this sounds confusing, that’s kind of the point. We’re a long way from just Dems voting more by mail. It’s not impossible, and we can’t rule it out. But if it’s about mail-in ballots, there must be some difference *within Dem voters* between mail vs in person. (44/N)
2
64
1
460
Races swung more towards Dems exactly where the Dems were down on Wednesday early morning. To explain this with mail-in ballots needs a very complicated story. To explain it with fraud needs a very simple story – you commit fraud more where the fraud matters more. (45/N)
8
210
18
788
This is why the evidence suggests fraud to me, but your mileage may vary. I’ve tried to stick to the facts, as I don’t have any special ability to interpret the numbers above. Whatever is going on is crying out for explanation, and the simple alternatives don’t do it. (46/N)
5
88
4
544
A final question to ponder. What should our null hypothesis be? When we say “there’s no evidence of fraud”, we’re claiming “no fraud” as the null hypothesis. To me, the system of vote counting is so broken that this is very difficult to justify. nitter.net/shylockh/status/… (47/N)
A metaphor for the likelihood of voter fraud, for people who insist that it's a conspiracy theory, or there's no evidence of it. (1/7)
Show this thread
2
75
0
494
I find the possibility of voter fraud entirely plausible, and that belief has nothing to do which party you think is doing it. At a minimum, I feel strongly that this possibility needs to be investigated more seriously than it is, given the evidence above. /fin (48/N)
21
99
2
651
One postscript that people have asked about - why Milwaukee? Why not analyze everywhere else too? I picked Milwaukee after initial strange reports of ballots being found at 4am. But it took about 48 hours of work and writing to put together just this much.
11
72
2
511
Do you still believe that most voter fraud will be caught in time before it decides the election? Really?
10
59
2
446
Most people assume that there must be thousands of highly paid people working on this problem right now. I suspect the answer is more like ten, and I personally know about seven of them.
17
88
7
591
At least for doing this kind of analysis, not for the broad voter fraud question. But a similar issue applies, I suspect.
49
46
1
460