2742
Comments (47)
sorted by:
36
Marshall 36 points ago +36 / -0

I love the usage of IMPROBABLE when talking about statistics. There is no such thing as statistical certainty, just HIGHLY improbable approaching certainty.

11
Proda 11 points ago +11 / -0

Yup, I was discussing something like this with my dad some time ago, and told him: "do you think it would be possible for all the oxygen molecules in a room to move in the corner, occupying a 1 cubic meter volume on the opposite side of you and have you suffocate?" He answered with: "of course it is impossible!"

Well, no, it isn't, in their random movement they are just extremely unlikely to move all in the same direction and remain relatively stationary there for a sizable amount of time without any external force, and I mean, so incredibly unlikely that getting out and finding the love of your life, then marrying her, then having a lightning strike kill her in front of you all in the same day is more probable.

Though, if something that should happen once in a million times, happens together with more million to one chances, like it seems to be the case this US election?

Fuck it, it's very likely an external force is influencing this shit.

2
Highfyingcrow 2 points ago +2 / -0

I would argue that it is impossible due to the diffusion gradient. There isn't enough inherent energy in the system overcome that type of entropy.

Furthermore it would need to retain that state for enough time for you to suffocate.

It is so far off the realm of possibilities.

1
Proda 1 point ago +1 / -0

Molecules move randomly, entropy in this case very strongly favours an even concentration of them throughout the room. But moving randomly and independently from one another doesn't completely exclude the remote possibility of them all moving towards the same point. It's not completely impossible, but it would be so damn unlikely that I don't expect it to ever be witnessed in the whole life of the universe.

2
Dahnald2020 2 points ago +2 / -0

Never thought I'd see some statistical mechanics on TDW, yet here we are :D

4
Proda 4 points ago +4 / -0

I'm a Chemist, had an exam on that in uni, hated it at the time, but gotta say, thinking back on it it was fun.

17
MightyTheme 17 points ago +17 / -0

MEDIA: Whoever takes Florida will almost certainly win the election! DJT takes Florida. MEDIA: Whoever takes Ohio will almost certainly win the election! DJT takes Ohio. MEDIA: Oh well, just print 50K votes, Biden wins. How dare you question our democratic process, bigot.

1
hansgruber7 1 point ago +1 / -0

Absolutely, but pretty sure it was hundreds of thousands of votes.

8
Bullet3250 8 points ago +8 / -0

MORE SUPERPOWERED AUTISM HERE -

"The R2 value of 0.9815 is very close to 1.0 (1.0 being perfect linear correlation), so the odds that the data are not linearly correlated are astronomically low. The only logical conclusion is that approximately 750,000 votes were flipped from Trump to Biden by the computer systems in Georgia. - - "

https://drive.google.com/file/d/1qZsdM34SULa2BEt72rDOdHSba_z8dQr-/view

2
IcedCovfefe 2 points ago +2 / -0

When I was in my college stats class I’m sure the first sentence in this paragraph would have made perfect sense to me.

Now (20 years later) it sounds like a foreign language.

2
buzzly6 2 points ago +2 / -0

Don't feel bad, I thought it sounded like a foreign language then.

1
Bullet3250 1 point ago +1 / -0

The link walks you through the data... well worth it.

8
TrumpTrainChoooChoo2 8 points ago +8 / -0

What’s funny is when I went to a Google news search to see how they were framing this, all the articles just say “oh well I guess we can’t count on bellwethers” with NO mention of how mathematically improbable it is that ALL these counties just happened to get it wrong when that hasn’t occurred in a hundred years. As someone with knowledge of statistics, it was pure propaganda. No different than saying “well I guess we flipped heads 1,000 times in a row, must just mean there’s no more tails.” NO that means that something was clearly rigged you hypocritical pieces of shit.

The media is truly enemy number 1 at this point.

3
VoteCyborgTrump2040 3 points ago +3 / -0

They truly are. Without them, none of this corruption would be possible.

5
Bernd_Lauert 5 points ago +5 / -0

Most people don't realize just how improbable this is. If we assume that the bellwethers fail 1 out of every 5 elections (which is not true, they're much better than that, but just for the sake of it), the probability of 18 out of 19 bellwethers being wrong in the same election is 0.000000000398459%. This is not a typo or anything. It's borderline impossible.

1
Hunni_bee 1 point ago +1 / -0

Wow!

5
Giant_Robot 5 points ago +5 / -0

And, the only bellwether he won was in his home state.

6
RexCollumSilvarum 6 points ago +6 / -0

Totally consistent with a landslide-losing candidate. He wins his home state and not much else.

3
WishdoctorsSong 3 points ago +3 / -0

Revenge of Walter Mondale.

4
deleted 4 points ago +5 / -1
13
fjobb 13 points ago +13 / -0

It's not impossible though. However when there are so many signs that would indicate Trump is going to win, and they are desperate to not look into anything, or do basic integrity measures, it is not going to sit well with people who are honest with themselves and care about integrity more then naked power grabs. When it’s the side that wanted to investigate Trump for 4 yrs over ridiculous claims of being a Russian spy, but they don’t want to spend a few weeks to even wait for court cases to play out here, it just makes it more obscene.

7
theultimatesean 7 points ago +7 / -0

Just like it is not “impossible” that all the atoms in your chair “happen” to all move in the same direction at the same moment and lift you right off the ground.

But in the real world, we would be pretty comfortable saying such a thing is impossible. Just like all the bellwether counties being wrong. They weren’t wrong.

3
MuadDon 3 points ago +3 / -0

But in the real world, we would be pretty comfortable saying such a thing is impossible.

You just defined improbable!

3
fjobb 3 points ago +3 / -0

Don't get me wrong, I certainly don't think biden won, and I will never believe there was not fraud (largescale fraud, there will never be any election ever with absolutely 0 fraud) in this election no matter what anyone says, short of Trump coming out and saying it was all a scam and I lied, or if it was somehow proven all the evidence we are looking at here daily was all faked. Especially when you add other evidence like biden doing worse than even hillary most places other then absolutely critical cities that he had to win, the fact no one liked him and he only got out of their own primaries through shady means, the turnout numbers being so high that they are unbelieveable etc. I agree it is functionally impossible that he actually won. However if you have some guy analyzing stats, and he claims that since bellwethers went for Trump so that means it is impossible that he lost, that is just simply not true and can be used to discredit, hence why I think the "improbable", or "extremely improbable" is the correct term.

1
Marshall 1 point ago +1 / -0

What you have concluded is that it is implausible. On this point I agree. Manipulated input is a much more plausible statistical probability.

4
Dtom13 4 points ago +4 / -0

Trump also won both Florida AND Ohio. Only three times in our nation's history has a presidential candidate won both Florida and Ohio but lost the general election.

1
arandoperosn 1 point ago +1 / -0

When were the other two? It has only happened once since 1900. The country was very different demographically before then.

1
Dtom13 1 point ago +1 / -0

Kennedy and I forget the other one. Florida and Ohio are both populous states with a lot of electoral votes, and have voting demographics comparable to most other states. So if you can't carry one or the other or both, you most likely can't carry enough other states to win.

2
arandoperosn 2 points ago +2 / -0

I went through the list from 1900 till now. From 1900 until the 1920s (ish) there was a strong north/south split, so Florida went Democrat and Ohio went Republican pretty much every election. From WWII onwards, Kennedy won in 1960 without winning either, and in 1992 Clinton(?) won the election winning only one of them. Other than that, the election winner has always won both, since WWII.

That's why I'm saying the demographics were very different back in the 1800s to early 1900s. It was the very strong north/south split.

1
Dtom13 1 point ago +1 / -0

True, so I guess that presidential candidates always won one or the other or both, but the reasons for that were different as time went on.

2
arandoperosn 2 points ago +2 / -0

Yes. And to compare Biden to Kennedy is somewhat of a stretch.

1
Dtom13 1 point ago +1 / -0

I have heard in places that Kennedy stole that election. Not sure how true it is, as I have never done any research on it. I do know that his assassination wasn't carried out by some "lone gunman"

1
arandoperosn 1 point ago +1 / -0

There are conspiracies about that? There was a good point made on TV... (by someone who must not be named here anymore) that many conspiracy theories have been born out of a lack of transparency from the government and the media that covers them.

2
SayItAintJoe 2 points ago +2 / -0

All this does is piss me off even more. It's so fucking blatant.

1
Rickshawrick 1 point ago +1 / -0

what was it in 2016?

1
HocusLocus 1 point ago +1 / -0

First... Steve Cortes doesn't list the actual counties he used. That is unforgivable in this high-paranoia environment. His claimed criteria is "perfect since 1980"

I find only 11 bellwether pre-2020-election counties with this criteria.

Of the 11, Trump lost one (Clallam County, Washington) so with his criteria it is only ~91%.

  • 53.8% Valencia County, New Mexico – perfect with the electoral college winner since 1952 (longest current perfect streak)
  • 56.3% Vigo County, Indiana (county seat: Terre Haute) – 2 misses (1908, 1952) from 1888 on, perfect since 1956.
  • 53.7% Westmoreland County, Virginia (county seat: Montross) – two misses since 1928 (in 1948 and 1960), perfect since 1964.
  • 60.9% Ottawa County, Ohio (county seat: Port Clinton) – one miss since 1948 (in 1960), perfect since 1964.
  • 52.9% Wood County, Ohio (county seat: Bowling Green) – one miss since 1964 (in 1976), perfect since 1980.
  • 54.2% Essex County, Vermont – one miss since 1964 (in 1976), perfect since 1980.
  • 63.9% Juneau County, Wisconsin – one miss since 1952 (in 1960), perfect since 1964.
  • 56.3% Sawyer County, Wisconsin – one miss since 1952 (in 1960), perfect since 1964.
  • (Trump lost) 47.1% Clallam County, Washington – two misses (1968, 1976) since 1920.
  • 56.7% Hidalgo County, New Mexico (county seat: Lordsburg) – one miss since 1928 (in 1968), perfect since 1972.
  • 59.2% Washington County, Maine – one miss since 1972 (in 1976), perfect since 1980.
2
deleted 2 points ago +2 / -0
1
GabenOfficial 1 point ago +1 / -0

1 in ten quadrillion

1
VetforTrump2 1 point ago +1 / -0

We know, just convince the judges and hang the traitors

1
Ruinerdown 1 point ago +1 / -0

Gtfo...... data is racist!!!!!

-1
Walbort -1 points ago +1 / -2

It's only a statistical improbability if the conditions were relatively similar to past elections. Mass mail-in voting, by itself and not even considering fraud, could fundamentally change the election dynamic and make bellwether counties statistically irrelevant.

2
LiberalismIsTheVirus 2 points ago +2 / -0

Except that Bellwethers essentially resonate with what the majority of what America wants. That's their entire purpose. It is only improbable if the changes were fraudulent which it was.

0
Walbort 0 points ago +1 / -1

More specifically, bellwether counties resonate with what the voting demographic wants. If mass mail-in voting somehow got more people to vote than would otherwise, more specifically getting lazier people to vote, this could easily change the voting demographic dynamic. This is just one theory, as is fraud another theory. Another theory could be that Trump got more people to vote than ever before, but got more new people to vote against him rather than for him, thus changing the voting demographic dynamic.

1
Marshall 1 point ago +1 / -0

This is true and why allowing unsolicited mail out of ballots without strong audit controls was known as an invitation to fraud from the very beginning.

1
VoteCyborgTrump2040 1 point ago +1 / -0

How would that change the bellwethers? It wouldn't at all. Unless you're saying that mail-in voting caused there to be massive amounts of fraudulent ballots. Then it would change things.

And for the record, Trump didn't just win the bellwether counties, he fucking smashed them. With double digit leads. One of them was a 23 point lead.

-1
Walbort -1 points ago +1 / -2

My point is that it wouldn't necessarily change the bellwether counties at all, but enough other counties get a much higher Dem turnout. The bellwether counties are a statistical anomaly. There is a predictable correlation between how they vote and how the rest of the country votes. It's not unreasonable to assume the voting population demographic could change drastically, but affect different counties/states/regions differently, potentially having little affect on bellwethers, but having a massive effect elsewhere.

This explanation is plausible, even if it's not likely. Since there are other plausible explanations, the statistics and probability in the OP post don't tell the whole story, and can't be taken at face value.

1
VoteCyborgTrump2040 1 point ago +1 / -0

Yeah, that's ridiculous. It's not plausible at all. It might be technically possible, but it's highly implausible to the point that no reasonable person would even consider it. The voting was not very different, despite the mail-in ballots in most places. The anomalies only exist in important swing state counties. In most counties, the turnout is consistent and normal. Then you have certain fraud-ridden democrat counties, such as some in Detroit that have 80+% turnout because of obvious fraud.