3 ms·

Interesting that the Harvard statistician they used as an expert witness says: "the probability of Joe Biden winning the popular vote in the four states indepe

by _red 6y ago

Interesting that the Harvard statistician they used as an expert witness says: "the probability of Joe Biden winning the popular vote in the four states independently given President Trump’s early lead in those States as of 3 a.m. on November 4, 2020, is less than one in a quadrillion." This seems to indicate fraud.

jetrink 6y ago
"[L]ess than one in a quadrillion" given certain assumptions which likely don't include the high volume of mail-in votes due to this election taking place during a pandemic. This election is not like others and if you use models based on past elections to draw naive conclusions, you're going to get garbage results.
- _red 6y ago
  Mail-in ballots with no return address, driven across state lines from NY to PA in trucks, with unverified signatures on forms pre-printed in China. Hopefully you are not in charge of IT security?
  - microtherion 6y ago
    ... according to allegations by people who are also convinced that this is all following a master plan conceived by Hugo Chavez (presumably before he died in 2013).
- tekknik 6y ago
  Isn’t this what every pollster did?
throw-away_42 6y ago
Only if you're clueless about elections and probability.
DennisP 6y ago
Not only does it not indicate fraud, but it was the expected outcome, because Trump convinced Republicans that covid wasn't real and mail-in voting was bad. So Republicans mainly showed up at the polls while lots of Democrats mailed in their ballots, many of which were counted after election day. And why were they counted after election day? In some of the swing states, including Pennsylvania, it was because Republican legislatures outlawed counting them as they came in.
khazhoux 6y ago
Everyone on HN is smart, so let's dissect the statement. * First of all, a quadrillion is a h(u^15)ge number! So right off the top, the notion that Biden's chances of winning these four states were that low, is w(a^15)y out of alignment with the roughly 50-50 split between Dems and Reps. Given no additional information, Biden's chances of winning four 50-50 states are 1 in 16, not 1 in 1,000,000,000,000,000. * Second, this idea that Trump had a lead "earlier in the night" is meaningless, because the vote counting was not evenly distributed. Votes were group by county, by method of voting, etc. Each grouping has a bias. Some counties are heavily Rep or Dem. Mail-in ballots obviously favored Dems because Trump spent the last few months telling his followers not to use them. * Third, why is this "3am" cutoff meaningful? The legitimacy of the ballots does not depend on the order in which they were counted, so wallclock-time is irrelevant. Otherwise, if the ballots had been counted in reverse order I could sit here and claim that Trump cheated because Biden's lead shrank sharply after 3am.
gdulli 6y ago
Not that this point is made in good faith or deserves a serious response, but the different patterns expected in early vs. late ballots was well known and reported in advance of the election, based on well established patterns from past elections.
OldHand2018 6y ago
Not a Harvard statistician. See the appendix on the posted documents for his actual statement (about 20% into the 170 page PDF that has random page numbering). BA Colorado College, PHD Rutgers, post-grad work at Wisconsin in applied economics and econometrics. His career is in economics and environmental studies. Well-educated, but not in statistics. He spent 3 years in the 1980s at the Kennedy School doing energy and environmental policy. This is all in his statement filed by Texas. He states that there is no reason more people would vote for Biden in 2020 vs voting for Clinton in 2016, so it must be rigged. He states that the votes tabulated late at night can't have been from the same sample of voters as the votes tabulated earlier in the evening, so it must be rigged. His argument is that an election is identical to a giant bag with millions of balls, some of which are red and some are blue. Each vote is nothing more than pulling one of the balls out of the bag and looking at the color. The proportion of red balls to blue balls was determined in the 2016 election, and the 2020 election should have the same proportion but more balls. It's completely wrong.

by _red 6y ago

jetrink 6y ago

"[L]ess than one in a quadrillion" given certain assumptions which likely don't include the high volume of mail-in votes due to this election taking place during a pandemic. This election is not like others and if you use models based on past elections to draw naive conclusions, you're going to get garbage results.

_red 6y ago
Mail-in ballots with no return address, driven across state lines from NY to PA in trucks, with unverified signatures on forms pre-printed in China. Hopefully you are not in charge of IT security?
- microtherion 6y ago
  ... according to allegations by people who are also convinced that this is all following a master plan conceived by Hugo Chavez (presumably before he died in 2013).
tekknik 6y ago
Isn’t this what every pollster did?

throw-away_42 6y ago

Only if you're clueless about elections and probability.

DennisP 6y ago

Not only does it not indicate fraud, but it was the expected outcome, because Trump convinced Republicans that covid wasn't real and mail-in voting was bad. So Republicans mainly showed up at the polls while lots of Democrats mailed in their ballots, many of which were counted after election day. And why were they counted after election day? In some of the swing states, including Pennsylvania, it was because Republican legislatures outlawed counting them as they came in.

khazhoux 6y ago

Everyone on HN is smart, so let's dissect the statement. * First of all, a quadrillion is a h(u^15)ge number! So right off the top, the notion that Biden's chances of winning these four states were that low, is w(a^15)y out of alignment with the roughly 50-50 split between Dems and Reps. Given no additional information, Biden's chances of winning four 50-50 states are 1 in 16, not 1 in 1,000,000,000,000,000. * Second, this idea that Trump had a lead "earlier in the night" is meaningless, because the vote counting was not evenly distributed. Votes were group by county, by method of voting, etc. Each grouping has a bias. Some counties are heavily Rep or Dem. Mail-in ballots obviously favored Dems because Trump spent the last few months telling his followers not to use them. * Third, why is this "3am" cutoff meaningful? The legitimacy of the ballots does not depend on the order in which they were counted, so wallclock-time is irrelevant. Otherwise, if the ballots had been counted in reverse order I could sit here and claim that Trump cheated because Biden's lead shrank sharply after 3am.

gdulli 6y ago

Not that this point is made in good faith or deserves a serious response, but the different patterns expected in early vs. late ballots was well known and reported in advance of the election, based on well established patterns from past elections.

OldHand2018 6y ago

Not a Harvard statistician. See the appendix on the posted documents for his actual statement (about 20% into the 170 page PDF that has random page numbering). BA Colorado College, PHD Rutgers, post-grad work at Wisconsin in applied economics and econometrics. His career is in economics and environmental studies. Well-educated, but not in statistics. He spent 3 years in the 1980s at the Kennedy School doing energy and environmental policy. This is all in his statement filed by Texas. He states that there is no reason more people would vote for Biden in 2020 vs voting for Clinton in 2016, so it must be rigged. He states that the votes tabulated late at night can't have been from the same sample of voters as the votes tabulated earlier in the evening, so it must be rigged. His argument is that an election is identical to a giant bag with millions of balls, some of which are red and some are blue. Each vote is nothing more than pulling one of the balls out of the bag and looking at the color. The proportion of red balls to blue balls was determined in the 2016 election, and the 2020 election should have the same proportion but more balls. It's completely wrong.