vinivici9586:

As a self respecting librul, I'm going with your map not so much because I believe Biden will win both Texas and Georgia, but because I'm hoping and praying that turns out to be the case.

There may be some basis for our optimism. I was listening to Amy Walter discussing Texas and the enormous number of Texas voters who have already voted. She noted that nearly as many Texans have already voted as the total number of Texans who voted in 2016! This may be wishful thinking on my part, but I sense that huge numbers of voters are fed up with Trump and want him out - the sooner the better. (Ted Cruz is on the record stating that he believes Republicans could suffer losses of "Watergate proportions" in tomorrow's election, so even Ted recognizes that tomorrow could be very bad for the GOP.)

The kind of massive nationwide voter turnout that appears to be underway suggests to me that Biden is going to win in a landslide. It may not be comparable to the 1972 Nixon/McGovern landslide - Nixon won all but one state - but it will be a humiliating defeat (and stunning rejection) of Donald Trump.

If Trump is defeated in line with your prediction, (mine too), it will be interesting seeing how quickly the remaining Republicans turn on Trump denouncing him for the GOP debacle. It didn't take Republicans long to turn on Newt Gingrich after the 1998 mid-term debacle. Gingrich got the blame for leading a fool hardy effort to impeach and remove Bill Clinton from office. For that mistake, Gingrich felt compelled to resign as Speaker of the House.

If Luckbox turns out to be right and a calamity occurs, (i.e. the pollsters are wrong again and Trump somehow gets reelected), we'll have to suffer through four more years of Luckbox (and Mason) gloating.

It is cute Bundy5 doesn't understand the difference between majority and minority. Plurality isn't necessarily a common word, but calling a minority a majority, geez.

Shy Trump voters again, with a slight move in the polls in the last few weeks towards Trump. Also if there is a move, it probably will move in a similar direction in OH, FL, NC, GA, PA, IA (NE3 is similar to IA) .

I look at this as one coin flip with Trump getting all 6 or probably close to zero of the states. Not 6 independent coin flips.

Biden, 291. If Trump wins, the people publishing the polls are in the tank and should not be trusted again.

All of the betting lines seem to have factored in trump outperforming polls by a margin similar to what he did in 2016.

for instance, he outperformed the national popular vote by 1.5%, and something very similar in battleground states on avg.

and so in states where biden is up 1.5%ish, the line is about even money on predictit. ie: arizona, where biden has implied pr of winning at 52% with a polling edge of about 1.5% on a composite of most recent polls (via real clear politics and 270towin). in north carolina polls are almost dead even and yet trump has an implied pr of winning of 54%. supposedly the pollsters improved their methods a bit since 2016 - color me a skeptic. when you look at how the same pollsters are doing relative to the right leaning ones like rasmussen, you see that the gap is pretty similar so i expect the polls on the whole to underestimate trump by a similar margin (1.5%).

some state odds are a bit out of line if we just deduct 1.5% from the poll composites but msot are close and clearly show that consideration has been given to the bias found in less cycle.


most likely path to victory is trump winning all of the states he's slightly favored in (via predictit), and then also arizona and pensylvania. that would be 279. wouldn't be shocking to see him take nevada either. but this requires everything to fall into place.

biden really just has to win 2 of the closely contested states, or one of the big ones (florida). if he takes florida it's basically game over.

i like the biden price as it stands though it's close. would probably bet trump at +225ish.

We'll see, there are no guarantees in politics. The polls are certainly not in his favor and it won't be a big surprise if he loses. Still, Trump has a massive outreach. Election day and subsequent counts will determine whether he has managed to capitalize on that. We live in untraditional times of politics.

There is a saying in my country, "don't sell the hide until the bear is shot". We'll know when we know.

this is way too confident.

granted he's up 5% and 7.5%ish in those states respectively but the margin of error on these polls is quite high even before accounting for the methodology understating trump support.

there were a few states in 2016 where final poll composites were off by 7%. this is something smaller states are particularly vulnerable to when the polling is done with fairly small samples. the margin of error on the wisconsin polls for instance is about 4%, shows wild inconsistencies between pollsters, and from the same pollster in subsequent periods.

we should expect to see biden/trump outperform the polls by a similar margin in at least a couple of states. the betting line now implies biden has a 70% chance of winning. think it's better than that? go make some money.

I got a text today from Mercury Opinion for a poll asking R, D, T-third party, or not vote.

They could also be betting that police pepper spraying voters who responded "Biden" to that poll and preventing them from voting could hand Trump the victory, which is not as bad of a bet!

it could be that there are new methodological flaws in the polling above and beyond what we expect from 2016, and we could see something dramatic happen in the next 24 hours that'll shift public opinion, but without any specific reason to believe those things exist/will happen the extent to which the results deviate from betting lines is going to be an artefact of statistical noise (which will be independent).

The problem with these electoral college prediction threads is that 99% of the people predict a Democrat win, because this politics forum is very heavily left wing biased. I think last time, there was either no one or almost no one who predicted a Trump win on this thread.

what site are you using to filter polls? how do they grade them?

real clear politics has trump +0.5 and 270towin has biden +1.

at those margins the betting line makes complete sense.

but of course they include trafalgar and rasmussen, which i'm assuming the site that grades pollsters you're using is not a fan of.

Just for the record.... Rassmussen had it within 0.4% as they called Clinton +2 in their last poll and Clinton won the popular vote by +2. She lost PA by 44,000 (not "by like 10,000) votes but other than that.... quality posting.

there were several people here vocal about betting on trump. whether they thought trump was a favorite i have no idea, but thinking trump was the favorite to win was a much bolder position and quite uncommon (you can't really frame that in terms of the left/right dichotomy). this is why the line settled in where it did.

Fair assessment.

I'll predict that Trump wins by one vote. Amy Coney Barrett's.

Im the odd one out. I have decent chunk on Trump #leggo

Text voting is probably more accurate to someone phoning you and asking you.

First of all with all the polling measuring the last 3 weeks of polls is about as good as asking Ptolemy about the order of the planets. Secondly the "polls" in 2016 underestimated Trump in NC by 2.7%

https://www.realclearpolitics.com/ep...nson-5951.html

Betting looks fair....

Also have to ask this question has there been more polls done compared to 2016?

If you consider Ms Barrett following the US Constitution and upholding the primacy of the PA legislature making the voting rules and not the fifth column of the PA Supreme Court re-writing the election laws, then you may be correct....

355 Trump

the interesting thing is that if you modify your criteria ever so slightly you can find estimates all over the place.

If i include B- or better pollsters from the past few days, i'm looking at biden +0.5.

And ironically if i include low graded pollsters over the past 3 weeks, it actually increases bidens edge (mostly because of survey monkey).


i'm not a big silver fan and have issues wrt transparency in terms of how he uses the data to arrive at a score. there're some with A+ ratings that have nearly identical mean reverted bias scores and percentage scores for accurately predicting the winner as others who got a C-.

like - i can see how something like rasmussen gets a lot of flack. but he lumps trafalgar in with them, even though their results when back tested only slightly overpredict republican votes.

https://www.thestreet.com/mishtalk/p...polls-as-crazy

he goes nuts because he dug through the tabs of one obscure poll where it showed results for a small subset of voters that he thought were highly improbable, and it was then later removed. was it an error? did they mess up the sampling? was it just a very small sample size? ... is this really why they're graded a C-?

it's unfair to look at the differential for any given state as a predictor of that states differential going forward. sample size, yes, but more importantly the pollsters that are producing data for that particular state at that particular time often won't overlap and their methodology is different.

what we could do is use each pollsters overall party bias as a deflator to create an adjusted average.

I agree the states are correlated, but I don't agree that it's an all or nothing proposition for these six states. That would only make sense if the polls were saying the same thing in all of these states. Biden is ahead significantly in PA, by much less in FL and NC, OH and GA are basically tied, and he's losing in IA.