Ok, so a couple of questions:

1. What margin of error do you have to build into polls to account for this? Are we talking 1 or 2 percent, or much higher figures?

2. Does this polling error only apply to right wing parties, or can it be a problem for left wing parties also?

You can't build a margin of error for it. You can only do that with things like sampling error. This is not a "known unknown" it is a "unknown unknown".

So far it has only applied to right-wing parties. Could it apply to left-wing parties. Yes, I believe so. But you'd have to find a left-wing party which a) appeals to people who don't normally vote and b) who reflexively hate all manifestations of authority including pollsters and would refuse to speak to them. I'm not sure that applies to the supporters of any left-wing party.

The most plausible left-wing organization to benefit from the same bias is Jeremy Corbyn's Labour Party in the UK. He has a very passionate fan base and has built the largest support network in Europe. The media universally hate him and his supporters are naturally anti-establishment-on the other hand they don't seem bashful about speaking up.
I'm watching elections involving Labour closely to see if there is any systemic bias: it could be profitable if it materializes.

I believe a lot of 2020's result will come down to "is there a recession beginning within 12 months of Nov 2020?"

I'm not sure that matters any more. Economic growth is just going to the super-rich.

If unemployment jumps a couple points during Trump's last year, you don't think he'll get dinged for it ? He only won those midwest states by a slim margin.

The aim is really to find 'populism' and if it really is catching on. Jeremy Corbyn isn't going to win anything. Look at the recent by elections. It isn't because of his politics its because he runs away from the media when they ask him the simplist of questions. He's the only labour leader not to have a positive rating. Also he's popular in a demographic that doesn't turnout i.e young people.

As a US comparison, if Bernie Sanders was the candidate he probably would of turned some red states blue. He's a competent and popular politician unlike Corbyn and would have increased turnout.

Finally, there's more succesful right wing politicians as the older demographics turnout more.

Well you bet on Trump, didn't you? You must have said to yourself 'if Trump gets within X percent of Clinton in the polls in the final few days, I'm betting on Trump'?

For Jeremy Corbyn, I believe there's no evidence (as of now) to suggest the country will vote for him in a general election.

Anyone here in the popular vote margin contract on Predictit? Seems weird that they expired the contract a few days ago, but haven't settled it. It's still there but can't be traded. I guess I don't have real extensive history on there, is this normal?

For more than simple Yes/No contracts they take a while to resolve. I think part of their process is winding down some stuff manually because there can be some weird dependencies. Don't worry, it's not instant.

Market was closed/expired by them 8 days ago, and still not settled. Seems pretty weird. Why would they stop trading on the contract, if they aren't ready to settle it?

There wasn't any evidence of Trump winning either. The pollsters are right in the vast majority of cases. It was a rational piece of guesswork. Making correct betting decisions doesn't require cast-iron scientific proof.

People confuse these things all the time. Say I've developed a cancer drug. I have to be pretty certain it cures cancer and doesn't have side-effects which kill people because the consequences of that could be catastrophic.

With betting the worst possible scenario is that I make a bet and lose the expected value of the bookie vig, say 10%. So if the bookies are saying Trump is 5-1 and I think he might be evens with a 20% probabilty then I make the bet. I will be wrong 80% of the time but if I am right even a small % of the time I will make money over enough bets.

This is why I love political betting. People like you form markets on the basis of wish-fufillment.

You make a number of demonstrably false statements:

1) Saying Jeremy Corbyn won't win anything is demonstrable bull****. No one knows anything with 100% probability. No one knows Man U will not lose to a league side in the cup. It may be likely-the difference between the things is as night and day.
If you really don't understand this then you should never, ever consider betting because you will lose money. I mean that with deadly seriousness, you are a danger to yourself.
If you were just using imprecise and sloppy language then you deserve to get slapped down for it on a gambling forum composed of intelligent people.

2) By-elections have zero correlation with anything. This is an easily demonstrated statistical fact.

3) Corbyn won't win because he "runs away from things" or something. Instinctive rationalization of an arbitary preference. Means nothing in gambling terms except for people fading you. For this reason it is generally a terrible idea to bet on candidates you vote for or fade those you oppose.

4) Corbyn doesn't have a positive rating. Again, means nothing statistically. People do not vote for a party because they "like" a leader, it is far more complicated than that.

5) Corbyn is popular in a demographic that doesn't turn out. That type of reasoning works until it doesn't. Trump was popular in a demographic that doesn't turn out also. Guess what-they turned out.

6) In the hypothetical scenario of Sanders winning the dem nomination he would win the presidency. Guess what-he didn't. Corbyn did win the nomination for his party.

Running through all this is your basic ignorance of how betting works. Corby does not have to be a short odds favourite to win the election. For betting purposes he simply has to have better true odds than the betting odds imply adjusted for vigorish.
If he goes off at the GE at 200-1 he's a fantastic value bet if he has a 1% chance of winning.

Maybe with unemployment. This is rather different from economic growth.

We've had a lot of economic growth since the seventies but no improvement in real wages. The media keeps pretending like it matters that billionaires doubled their wealth last year-it doesn't really affect the rest of us.

Unemployment is something which genuinely affects people. However I'm not sure exactly how this works in the States but here in the UK it is obvious the government just manipulates the data to make up a bull**** statistic that means nothing. "Real", for want of a better word, unemployment, yes I think that will affect the result.

So what are people betting atm? I got crushed on election night but have made almost all of it back, it seems like a goldmine post-election with all the trump money being tossed around.

ambassador to china market somehow dropped back down to ~83c. I've been maxed since the 60s and it was recently 90c+. Not much volume though since I assume most people are already maxed and waiting it out, but ~17% return in 3 weeks is 1 of the best bets right now imo

Also trying to max trump false statements in january 3 or fewer but there is very low volume

I don't think Ted Cruz is going to be the next Supreme Court Justice.

trump tweets is a new, very active market

I'm in the China ambassador market as well. It dropped off the highs when Trump team announced all current diplomats are dismissed on day 1 without exceptions. It also looks like Trump and the Rs are going to try to steamroll everything through without even bothering to get the ethics paperwork done first. It makes me a little nervous, but I'm still holding.

Why do you think you might have an edge betting on such things?

Betting on Trump tweets seems to be roughly equivalent of a bet on wags of a dogs tail on a given walk. More or less impossible to predict.

I'm not betting on tweets?

Although that twitter market is GLORIOUS to follow

They now say they are waiting on FEC report to settle, but why close trading on the contract then? It's now been closed 2 weeks and they don't even know for sure when FEC report will be released only that in the past it was released in January. The turnout market is also waiting on the FEC report to settle but it is still trading. Just kind of annoying because I have a large % of my predictit roll tied up in that contract and would have probably sold a little at 0.99 to participate in some other markets but instead have to wait weeks now.

I agree they shouldn't shut down trading until they are ready to close out the positions. I have seen decided but not official outcomes trading as low as 96-97c, so maybe they don't want people trading out of positions to affect "predictions" ? I can't think of any other reason.

Been making a mint in these tweet markets

Yeah they are quite profitable. Every time I think I'm maybe going to get stuck with shares he comes through with another meltdown.

How'd you do this week Mrage? Huge swings and volume the past couple days. Took a turn of work but I ended up making a nice profit

Hi, mostly going to lurk this thread. I have an occasional bet on politics stuff.

Obviously I haven't read from the start so I don't know if this has already been discussed, but do we like the value in Betfair's "Trump to Leave by End of 1st Term" market?

Currently
Yes Back 2.28 Lay 2.6
No Back 1.62 Lay 1.77

I backed "No" today at 1.75 (put it on in the morning and waited for it to be filled.) As an added bonus it's void if he ends his term as a result of an assassination attempt - so he has to resign, be impeached or the medical stuff.

I mean it's pretty rare for US presidents not to complete their first elected term (the last being JFK but that would be void) but then Trump is pretty unusual.