Originally published at: https://boingboing.net/2020/11/03/final-election-polls-and-forecasts-released.html
…
Long night ahead
Long weeks ahead. Long years, really, given that whatever the result America’s fascists aren’t going to just vanish.
What would it take to push for electoral reform?
Reminder to all: it may be weeks or months before we know anything. There were more mail-ballots cast in Texas alone than in the entire last election. All those have to be counted by hand. There’s concern that it may not even be possible to count them all by the Safe Harbor cut off date.
The Democrats to grow some metaphorical balls.
More brains and less balls in the Oval Office.
That too, but they need to have the political will to do it. It will get them a lot of shit, so they must have the commitment and conviction to follow through.
The will of the people, and changes to our Constitution. It’s baked in at the moment.
More useful in the short term: voting reform. A single bill that will fix voting in all fifty states simultaneously. The Supreme Court tossed out the 60’s era voting rights bill because it targeted only a few states. And the Constitution allows for Congress to provide rules and guidance over state law.
An aggregator of aggregators weighs in…
Nate Silver also points out that polling error isn’t necessarily favorable to Trump, and a 3-point error in Biden’s favor would result in a Biden landslide.
Right. If polling errors were obviously favourable to trump then they would have been fixed. The pollsters at least try to eliminate the biases.
How are pollsters accounting for the ‘hidden racist’ factor that put Trump over the top 4 years ago, i.e. people that won’t admit they support Trump when asked by a pollster but vote for him because the actual voting is anonymous? That’s always an error in the racist candidates favor.
Personally, I don’t believe that there are many Trump-voting racists who are so ashamed of their choice to be a Trump-voting racist that they keep it secret. There are plenty of extremely vocal Trump voting racists to provide comfort to those who are “shy.” But maybe I’m wrong—it’s kind of hard to prove or disprove the existence of “secret Trump supporters” because they are “secret”. We could just as easily talk about “secret Biden supporters”.
That’s not quite accurate. In their final forecast they give Trump a 10% chance of victory. Biden a 89% chance of victory, and a 1% chance of an electoral college tie that would be a nightmare for all of us.
If I may be pedantic (and this is the BBS, so I may!), the 1% in 538’s model is technically a combination of rounding error and the probability of a tie—in other words, their model puts the chance of a tie at less than 1%, but it’s getting rounded up. It’s still a non-negligible probability, and that nightmare could happen
I nominate @ugh for the Pedant Pendant badge
I also don’t know that polling errors are going to be the issue. Polling is based on likely and previous voters. Massive turnout and lots of first time voters isn’t something polling models can really figure.
It can’t be months. The electoral college meets December 14th.
Most mail in ballots are machine readable. And in most states both early votes and mail in votes are counted ahead of time. While those ballots have to be processed/approved by hand before they can be counted, even in Texas counting starts before polls close on election day and is done by machine.
Delays on that front come with come with large numbers of votes arriving on or after election day and GOP fuckery around whether those can be counted.
Hand counting is mostly a plan C for tight districts where multiple machine counts don’t give a clear answer.
Legal experts who know more about this than me have said it could be, depending on what goes to court and how long that takes. Say, for example, there are conflicting certifications of results in a state, or a state refuses to send electors. There are plenty of avenues for delay.
And COVID related staffing shortages and procedural hiccups.