i enjoy reading taleb but the video was confusing and unclear to me. he was kind of indignant and argumentative the whole time but i didn't know what he was arguing with. he strawmanned a fake 2016 election odds chart that suggested wildly swingy win probabilities from 75% to 10% to 75% over time.

they also made the point several times that you shouldn't price a binary event at 0% if there's still uncertainty and there's a potential in the future that it will not be zero, which seems obvious?

maybe the one semi-valuable takeaway is with high levels of uncertainty a binary option should trend toward 0.5, but i think that's what we've seen this election with both markets and high-profile models like 538 & natesilver. they seemed to be suggesting that 2016 forecasts were naive in deviating significantly from 50% and should have been much closer to 50/50 throughout, but that seems like captain hindsight. i don't think many anticipated the level of correlated error in state polls. that's obviously less excusable this time around.

I obviously didn't mean she wouldn't go on before as a campaign strategy.

I meant she would think something like that is clearly unnecessary and beneath her.

Your strategy sounds OK enough but she seems way too bad at politics to execute it effectively

don't totally discount the possiblity harris would benefit from going on jre, but if i were campaign manager i would not advise.

jre audience will not be blank slate voters that will be won over by above average performance. they're more likely to adjust sentiment based on how rogan treats her... cues & vibes. given his devoted love to elon i can't imagine warm vibes. it's a longform live interview where the only clips taken and replayed after the fact will be any moments where she looks foolish.

I think a lot of people will get cold feet voting for the first woman president.

Salient points of the video:

-a statistician who does a survey (e.g. 1,000 people 3 months before an election) and uses simple statistical sampling methods to price a probability of candidate A winning over candidate B, will likely be wrong;

-projections with no "skin in the game" likely use faulty methods to derive win probabilities;

-mainstream pollsters do not adequately consider uncertainty in their projections.

Shouldn't Harris be a bigger fav to win the popular vote on PM? 64.5-66c range that it trades in is about avg of -190

GOP hasn't won the popular vote in 6 cycles and if that type of shift was going to occur, wouldn't the midterms we just had 2 years ago have shown some type of evidence of that? GOP did horrible that cycle despite inflation being higher then, crime being higher, the stock market being significantly lower and 1st midterm after new POTUS is supposed to be a smash based on last 30 years.


This just doesn't feel any different than the last few elections where it was close and gets decided by like <200k votes in 2-4 states and dems win popular vote by 2.5M+

While that might be true it wasn't true for first black President and Hillary Clinton was a uniquely unlikeable candidate who had been attacked by the Republican machine for 25 odd years instead of a few months as a primary target

I'm not really speculating whether that's true or not but it's certainly a different situation to Clinton/Trump, at the time Trump was a relative unknown vs now he is a completely known quantity in that everyone's mind is made up about him one way or the other - his floor is close to how many votes he got last time, his ceiling is close to how many votes he got last time.

If Harris maintains Biden's turnout from last cycle or improves on it she wins, if not she loses.

Agree Harris prob is better than two thirds to win popular vote, but the last time Rs won the popular vote was 2004, so that's 08 12 16 20 4 cycles in a row Ds have won it if you're talking for president, granted 6 of the last 7 factoring in Clinton and in theory demographically more Gen Z voters are of age and more silents/boomers have died vs 4 years ago and Millenials aren't getting more conservative as we age at the same rate as other demographics - you might be counting midterms as a cycle but they're an entirely different thing as far as i'm concerned anyway idk

Honestly I think it comes down to turnout. Trump is going to do about as well as he did in 2020 in terms of raw voter numbers after adjusting for population growth- if D turnout goes drastically up, Harris wins in a landslide, if it stays the same she should win narrowly, if it goes down Trump should win.

yeah they won 5 cycles ago, not 6.

I know midterm isn't a general but I just feel like if that type of realignment was going to happen, then it would've started to show up in the midterms 2 years ago. Party out of power in the first midterm is supposed to smash and the Dems held their ground quite well. The meat on the bone of the abortion issue still has legs for them as well just like it did then.

Crime is lower than then and has started to regress to the mean after the post-covid/floyd high. Inflation has tailed off a ton since 2022. A lot of the local woke-ness run amok has begun to recede as well. Kids are actually taking the SATs to get into college again, etc don't really wanna go over all that stuff.

But the one no one really seems to be talking about to me is the stock market. The S&P is barely shy of 6k, it is up nearly 60% since Nov 8 2022. Many americans are pretty damn flush on paper right now because of that, many of them are the type of swing voters that are doing well enough to where they could just prefer the status quo over another disruption like Trump. And while I think Harris has many, many problems as a candidate because she can barely speak, I do think the attacks that she is far left have basically fallen flat because it is pretty obvious she doesn't believe in much at all and certainly not any type of progressive economic agenda that she will spend political capital for.

This seems tough for the Dems to run on given some of their stances on wealth and the fact that many of their voters don't really participate in this much, but it does seem like an underrated tailwind to me.

Good post by Domer on the bettor who is pushing around the odds on prediction markets the world over:

https://x.com/domahhhh/status/184659...qC5A7jA94k1vHQ

Shoutout to Biesterfield for bringing this post to my attention.

not sure including midterms makes sense for presidential popular vote odds. trump brings people to polls who don't show up for midterms.

2016 clinton led national polls by 4.5% and won popular vote by 1.5%
2020 biden led national polls by 8.5% and won popular vote by 4.5%
2024 harris currently leads national polls by 2.5%

65c seems reasonable to me

when is that data taken from? like led national polls in late October?

early november by 538 aggregate

turns out i can't math properly and actual clinton polling margin was +3.5% heading into election

it isn't apples to apples, but i do think if it was going to be this trump PV win where he does something no republican has done in 20 years then the GOP (who were the party of trump in the 2022 midterms and he endorsed as many candidates as he could) would've probably had a better showing than a historically awful 1.

That's about how I read it, but as I said, I didn't really have a handle on it. Thanks.

It's true that a lot of wild stuff can happen to totally change the election, which is not as true if we are betting on today's baseball game. To pick an obvious example, the Guardians are not going to bow out and be replaced by the Orioles.

Also, election betting might be closer to a sport like boxing. Where we have an ocean of stats based on these players going against each other dozens or hundreds of times in big team sports, boxers usually don't have many common opponents, it's hard to quantity, this might be their first fight in a year and the result could be determined by judges. So statistically driven prices are less sharp (which is true in boxing, or was back when I bet it).

Lastly, models might be built to tell us what would happen if the election is today, but it isn't. Variance is going to favor the dog. TBF if the model only claims to tell you what would happen today, you can't blame it for doing so. But if you are using models to bet you need to assume the dog has a better chance than suggested.

i honestly think kamala is almost as bad as hilldawg

meanwhile obama was perhaps one of the best candidates we've ever had

Interesting but carries a dem bias that probably wasn't needed for such a writeup imo

I agree that a shift should have shown in the midterms. I disagree with your stock market angle. A couple reasons: there's a bit of a bifurcation in who's enjoying the asset gains from the stock market versus lower-class non-investors. And those non-investors are also very upset about inflation. (Yes, it's come down, but people view it cumulatively, and even still, it's running above target.) So you have richer people, who tend to be older and whiter and more inclined to vote for Trump, who have benefited from stock gains under Biden. But I don't think these people are too inclined to give credit to Biden/Kamala for those gains. On the contrary, I think they are scared that Kamala is going to raise capital gains taxes, maybe even do a wealth tax, and do a bunch of socialist programs and interfere with companies (higher corporate tax, killing energy companies, etc.). So if you're one of these investors who have benefited, you don't want Kamala taking your investment gains away. You want Trump slashing taxes and cutting interest rates and sending your investments even higher. (And you don't really understand that his onshoring and tariffs are hugely inflationary, and that installing a mega-dove Fed chair and cutting rates is likely to blow out the long end of the curve and kill your bonds and probably result in even more inflation. A negligibly tiny percentage of investors really understand the macro-econ here.) So to sum up, the people who have benefited are disinclined to vote for her, and the ones who have suffered (no stock gains and more inflation impact) are her typical base and they are disgruntled.

All this said, I still think Kamala is the bet right now. Pennsylvania elected Fetterman two years ago. Fetterman.

TL;DR for those who don't want to read it. Domer thinks there's one person who is responsible for most of Trump's recent move on Polymarket, betting $25 mil. He thinks it's some random French whale who isn't sharp.

As far as suburban, college educated voters I'd think many of them have benefited atleast somewhat from the gains of the last few years and may not want a disruption.

It doesn't take much of a shift amongst that group in a swing state to really have an effect.

It's kind of a fools errand to try to read the minds of the voting public, imo.

For one thing, a big chunk of them cannot name the sitting VP at any given time. Those who do follow politics are often misinformed and/or project their own views onto the candidate the vibe with, or support on party lines.

Even if none of that was true, I wouldn't really bother trying to figure out something like that with armchair speculation. Well, TBH, I would. But I wouldn't give my conclusions much weight.

Domer take is a dumb one. Broader equity markets , btc etc also pricing in a trump win

People, what is your opinion of polimarkets?

very liquid, great market

edit: assuming you're talking about polymarket, and not some unknown thing called polimarkets

did anyone see Mark Cuban on the overtime segment of Real Time this week? Polymarket came up and he said he is an investor in a fund that invests in it

then he said it is foreign investors, can't invest in the US

I get he has to lie because it is his investment but he must know that like 75% of the site is US VPNs