Couple things:
I have no idea how polymarket works, but theoretically you could make money just on the interest of deposits you hold and people who accidentally leave money in their accounts but never claim it. My understanding is that this is a huge part of how Starbucks earns money: people put money on their Starbucks cards and then Starbucks puts that in a MMF and earns SOFR (like 4.8% at the moment). If you have a billion dollars in pending wagers and account balances, that's $48m in revenue just from the interest. And I don't know how many people deposit money and then forget about it/lose their password/die, but that may be in the millions for a large site also.

ES2, it doesn't seem all that unlikely that people have models that are radically better than the ones we know of. Nate Silver is treated like a god amongst the media and other professional pollsters, and the guy is dog **** at sports betting. He used to post plays on his site using his modeling abilities, and he was a big loser. So it wouldn't be surprising at all for sharp money to be modeling way better than the 538's of the world. It would be akin to listening to the "best" analysts on ESPN, and comparing them to Sigs and Haralabob. Obviously the latter two would mop all the analysts, and the latter two would be more responsible for moving the lines to the correct number.

This is wrong. They never hold your money. They can't take it even if you leave it for 50 years.

So no one deposits on the site? The only money they "hold" is the escrow for the bets? Even then you could get interest from all the pending wagers before they're settled and paid out, and right now that number is in the billions. I don't want to bore people with my ignorance of how this all works though.

It's in smart contracts they don't control. Crypto opens up cool opportunities.

I'm not sure how interest would work. Would be nice if there was a way to monetize that. I don't think it would be polymarket that receives the yield, it would be the individual position holders. Still not sure how that would work and why someone would pay that yield, but something to think about.

Trump wearing garbage man outfit

Betfair is or used to be 5 percent of net profits in the market as a whole.

In the 2012 presidential election name of winner market I was running a market-making bot on Betfair with a few of hundred UKP in my account. Took 20000 UKP in bets (exposure is calculated net, ongoing), made 140 UKP profit, paid 7 UKP to betfair.

Doesn't seem unreasonable.

Good points. I'm totally open to being wrong, as I'm mostly shooting from the hip. I have $ on Harris but don't like either. Maybe I'm coping on the $. But it's just like 1 big bet. Nothing crazy. I do know a few winning players with $ on Harris, including Trump supporters.

But, to push back: the disparity between the markets and polls/models is fairly wide, so that's why I said radically better.

Nates model, as far as I know, weights polls for historical accuracy and blends them. Is there much room for improvement on poll interpretation? Some, sure. But the public models seem to have done fairly well before. I can buy the idea that someone has cooked up a slightly better one.

Part of the point with Connor/Floyd is that the pro gamblers of the world don't collectively have all THAT much money, imo. I suppose crypto has changed that some. Still, when people are betting $25 million at a time... in my mind that is more like wall street money or some insane whale. So IDK, maybe these superior models were cooked up by some finance people. I know nothing of that world. Maybe lots of people saw the line movement and assumed it was inside info and tailed. Especially as these are mostly people who want to believe Trump will win.

While sprots is more sophisticated now, I don't think the type of person who would bet on Connor is. Is it not still the case that Super Bowl lines are a bit off? There are certain events that just get so much public action. There are a lot of Nick Vertucci types in the world.

There also seems to be some uncertainty about the polls this time. I've read some are trying to correct for under representing Trump in the past. Of course, in any case, the pollsters are also working to improve. I don't think anybody really knows exactly how good they are.

I also think politics brings out personal bias like nothing else. I'm sure we all see how unhinged true believers on both sides get. Including people who are usually reasonable. Not nec unhinged politically, but on the polls being rigged, thinking one or the other is a lock for dumb reasons.

I mentioned before, I kind of default to that Taleb view, which is that our bias should be towards it being 50/50. So I tend to think the right prices are between the models and markets. And, though it's now late in the game, randomness favors the dog.

Not that it will prove anything. But if Trump wins decisively, not necessarily a land side, but say he has clear wins in several key states thought to be toss ups, I'll definitely start wondering if someone out there has a superior method.

If someone did crack the code and makes a fortune, I hope we get to read about it one day. But I'm just generally inclined to skepticism.

Bookies have relied on NFL favorite bias for years.

Nate Silver owes a lot of favors and has become compromised. I keep hearing Taleb being quoted. Who has more skin in the game to be accurate to the percent in these elections: Silver or bettors?

Silver has been **** on by the left because he's been perceived to be right leaning. There's so many meta dynamics going on for him that it's impossible for him to separate his biases.

All the markets have Trump around 60% to win except Predict It which has Trump at 53%. There's many reasons to trust Betfair/Poly over Predict It which has high fees and low betting limits. There's many reasons to trust betting markets over someone like Nate Silver.

It's a close election either way.

Nate Silver is trying his best to accurately predict election results

For anyone to act like he is a left wing or right wing hack is ridiculous

Obviously as we know, the betting odds are not always sharp on major markets when a sheer volume of recreational money floods the market, think Conor/Floyd (Floyd was -3xx when he should have been -10000 or so) or literally most Presidential elections.

Also people need to remeber that a -150/+150 election does not mean that the -150 candidate wins 10 out of 10 times, it means they win 6 out of 10 times. Even with the popular vote as it was in 2016, Clinton still wins the electoral college a lot and while the betting odds were slightly wrong in hindsight in that she was too big of a favourite, she wasn't 0% to win there. Trump was a bigger dog in 2020 than the odds thought too in that there WAS a polling error in his favourite and he still lost with a multiple state margin. That isn't the +150 odd game day odds either, Trump was probably more like +300 or whatever if a not guaranteed polling error in his favour eventuated and he would have still needed a decent margin on top of that to get over the line

Obviously these things are easier to analyse in hindsight than beforehand, in that people will say 'of course polling massively underestimated Trump it always does and therefore the market was wrong' or 'of course turnout with women was higher than usual they overturned Roe' or 'of course the market was too skewed towards Trump, it built in an assumed polling error in his favourite to the point where his fair odds require him to overperform his polling by nearly 2% nationwide for the prices to make sense' or 'of course Harris was overvalued, the last female candidate underperformed and it happened again because some people just don't want to vote for a woman when it comes down to it but are embarrassed to say that to pollsters' or whatever

The narrative after the fact is super easy

I feel like my personal narrative leans (that make me think Harris is undervalued because the election is currently flippy on the numbers yet she is a moderate dog in the market) are most along the lines of 'the market is just assuming that Trump will outperform his polling because he has in the past; but pollsters have tried to correct their methodology from last cycle and that isn't being accounted for and it was a two Trump election sample which is too small to assume Trump will overperform his polling in 100% of elections he stands in' combined with 'high turnout is historically good for Democrats and low turnout is historically good for Republicans, everyone's mind was made up about Trump before this cycle anyway so he will get about the same amount of votes he got in 2020, so it comes down to can Harris turn out the same number of votes that Biden did or not, if yes she wins if no she loses' - I lean towards high turnout in America's current political climate.

By no means is Harris a lock and i'm not even invested in the election yet beyond about 0.05 units on Harris -99.5 at a mixture of +500 to +1000ish at this stage and I lost a portion of a unit on Biden nom which was smaller than the amount I actually wanted on so I was fortunate but i'm waiting for complete information right before the polls open and i'll fire my unit/units in and will either win or lose for this cycle once the result is known

For what its worth I think Harris pop vote at less than -200 is definitely +EV but whether it's the most +EV thing I can bet or not idk, there's some chance I end up firing on both candidates -64.5 and -99.5 lines even and end up effectively betting that regardless of which way the election breaks it will be less close than expected, because I think there's a lot of correlation in terms of which candidate overperforms on election day across all of the swing states whether it's polling error or turnout based and I think the market is over-estimating the chances that the election is very close regardless of who wins.

Anyway, not long left, we'll see, and I will definitely post my position/positions when I take them and then we can all monday morning quarterback about who got it right and who got it wrong and so on.

I think Harris is undervalued but not betting or even considering betting Trump. Here's why:

If Trump wins, the only way to make money on him is to have bet on him before the election.

From past experience, on the other hand if Harris wins, it will still be possible to bet her with Q-anon types taking the other side of the bet even after the election, as we saw in 2020.

Obviously the odds are better now than they will be after the election, but I think there's more value to be had around November 8th.

my main big bet is the Harris PV stuff, accumulated in the 58c to 69c range on poli for an avg of around 64 which is what it is at right now

ill just believe there is a huge shift for trump to win the PV when I see it


agree harris +150 is probably betting value as i still think it is a close election but id take trump with a gun to my head if that makes sense, think 55-45 probably closer to right

basically my final thoughts

is that really gonna happen again? feels like a 1 time thing and the sharp $ would be ready to pound the other side if something like harris 90-95c is out there

feels like something that can only happen once

just so many whales on poly now who will shell out Ms for a 5% return in 2 months

https://www.wsj.com/finance/trump-od...=hp_lead_pos12

like I said liberal biases. he think trump win 80%-90%

cool story, thanks for sharing.

if you're concerned your identity will get out, why do you reach out to wsj and video conference them?

nate silver accuses pollsters of herding. hadn't considered that possibility

Journalists tend not to reveal sources.

that's fair but if he trusted them why refuse to provide any proof he was what he said he was

i guess he has varying levels of trust.

ok, i'll talk to you, but not going to share my ID. also don't want video for reasons.

he did prove he was the bettor. but he could very well be an American, or a non-french person. who knows.

former journalist, very common for people to give up plenty of info they don't want out there to you personally just so you can confirm their credentials and trusting you won't disclose it while also not giving proof of other things

wsj incredibly professional - would be absolutely shocking if they didn't properly vet this guy beforehand

Seems some Kamala supporters are now manipulating the market….

If anything he has it in reverse. Trump wins and Elias and his crew roll out the lawsuits.

It could be 1876 all over again.

trump gapping down in every swing state but somehow still fave for presidency

All they really had to vet was that he is the bettor, and they did vet that. The rest is just noise. Who cares if he's French, American, Chinese...he provably demonstrated he's the dude that bet $30M+ on the election, so it's good to hear his supposed reasoning.