I've seen it happen. Stew in that cesspool of "jokes" and humor long enough, and it becomes how you think. Your brain comes to anticipate the jokes and the comments so you're part of the in-crew. And you start to think "I mean, it's kind of true". A subset of those people go on to be truly hateful.
Ideological homophily is a bitch. I actually tried arguing voaters out of their antisemitism for a while. As you might expect, it was a thankless task. I really think that examples like voat should make us skeptical of online communities in general. No one likes to be told they are wrong, or tell others they are wrong, so you get echo chambers.
Not for this, but for the "obviously crazy" Trump beliefs we have a useful proxy. PredictIt ran a large number of betting markets (far more than a traditional bookmaker) for the 2020 US Presidential Election.
Of course the markets about Trump have lots of people spouting outrageous nonsense. You could dismiss that as just "for shock value". But PredictIt is real money (albeit not the sums involved in the "real" international betting markets) and the actual bets placed reflect those same crazy beliefs.
About 24 hours ago the market for how many Electoral College votes Trump will get closed. The answer, as we've known for about a week now, was 232. In a normal healthy market, even if people are writing comments saying that the US military will force people to vote again at gun point, the price for Trump gets 232 votes would be $1 (the way PredictIt works you're getting one dollar for each correct prediction, so if you are absolutely certain of something it's worth precisely $1 and you wouldn't sell it for anything less).
The closing price, less than 24 hours ago, was 96¢. For 96¢ you could buy a dollar, just by being sure of a well known fact about Donald Trump's loss. Because the people "trolling" the comments really believe he didn't lose.
I'm skeptical of this because Withdrawals are subject to a 30-day holding period after your initial deposit and a 5% processing fee. (https://www.predictit.org/terms-and-conditions) If people have trouble getting in and out of trades those are the situations in which you expect market prices to be somewhat wrong. 5% is actually spectacularly high in most situations, imagine how much of an edge you need at predicting to beat that.
The price on predictit was ~85c for a very long time after the election. But it's not necessarily because people are unreasonable, there are limits of arbitrage as well. In order to make any profit at all you'd need to lock up a sizeable chunk of money, say $1000, for something like a month, pay 5% commission, plus I imagine the hassle of it all, and all you get is a few dollars. In economics literature things like this are described as limits of arbitrage, to explain situations (like deep out of money options) where it seems like an arbitrage opportunity exists but where everyone still thinks it's a bad trade.
You also have the non-zero risk of the site shutting down. Predict-It's been around a long time now, but I still wouldn't wire them more than I could afford to totally lose. 4% on what I can afford to lose really isn't worth the hassle.
I haven't checked for PredictIt since I am not an American and can't bet there. But the usual financial construction for a betting market where you can take both sides is the same as for a law firm, or a share dealer, all your money held by them is legally separate from their money so even if they get into horrible debt problems their creditors have no possible claim to your funds and you just get it all back in a few weeks after the paperwork is done.
It's clearly not as truly safe as government bonds or insured savings where the government is ultimately on the hook to make you square - but it's pretty safe. I wouldn't put next week's rent money (if I rented) on "Trump lost the election" at PredictIt. But money put aside for the new car, for next year's vacation or for a birthday gift in February? No sweat.
But predictit is not a usual financial construction. For example, the first thing it says is:
> PredictIt is intended and offered as an experimental research and educational facility of Victoria University of Wellington, New Zealand (“Provider” or “We”), not as an investment market or a gambling facility. PredictIt is not regulated by, nor are its operators registered with, the U.S. Commodity Futures Trading Commission (CFTC) or any other regulatory authority.
I don't know precisely what that means for segregation of funds, but I don't want to just assume they work the way I would imagine them to work.
PredictIt, IIRC, also has limitations on total bettors on an issue, and amount an individual can bet in total and per issue.
Those constraints can produce odd results. Say, if the market was “full” of investors so no one new could enter, and also had reached a point where the people who weren't irrationally holding out expectations of Trump has already maxed out their positions, the only available trades would be between relative dead-enders.
Plus it costs money to withdraw funds from the market (5%, from another comment), so it looks like even if you could get in, you could actually only by 95¢ for 96¢.
Don't be surprised that there are people who believe such things for real. There're organizations for the 1 in 50 or 1 in 1000 brightest people, but there's the other end of the spectrum too.
A friend/acquaintance of mine has Alex J and Info Wars as her primary (only?) news source, and she sounded proud about this, as if she's discovered something others didn't know about.
I can’t imagine people actually think like that. They must be doing it for the shock value, or because of some echo chamber effect.