My adventures on Intrade

In Everything is Obvious (Once You Know the Answer), Duncan Watts mentions attempts by the McCain campaign to shift the Intrade market during the 2008 election. Apparently, they were able to shift the odds, but only briefly. Presumably by lowering Obama’s odds they created a cheap buying opportunity for other people who (quite rightly) thought that Obama was going to win.

I’m sorry, how is Perry ‘toast’? The one who’s a joke within the conservative movement is surely Gingrich, not Perry. The likeliest outcome is surely that the ultra-conservative vote shifts behind Perry, as it was pretty much always going to once he entered the race.

To join in with Doctor Slack @3, Perry is following a theory of how to campaign that he used in Texas – pouring money into particular kinds of street-level organising, holding off certain kinds of TV and PR for now, etc.

One can certainly argue that it won’t work, but I’m not sure his numbers at this point in the cycle are a good basis to do so.

The trick with “information markets” is that they aggregate the views of the participants. Since the majority of the participants on InTrade are political junkies, that’s the bias you get in the prediction. Throw in the thinness of the market and it’s typically good for the big question, but shakier on the breakdowns. Particularly if something doesn’t fit the pattern of previous occasions.

I don’t know John… A 1 in 14 chance of Gingrich getting the nomination? Even aside from the polling, the relative size of donations and campaign war chests aren’t very supportive of that optimistic an assessment. Gingrich raised a little more than Rick Santorum in the third quarter, and it looks like he’s about maxed out available resources…

It looks to me as if Intrade’s results generally overestimate the chances of weird, improbable things happening involving people with well-known names.

They’ve got Hillary Clinton as 2012 Democratic nominee at 1%. I could conceivably imagine things getting so dire for Obama that he pulls an LBJ and declines to run (not likely, but maybe more than 1% probability), but I doubt his Secretary of State would be the leading candidate under that scenario; the probability should be dominated by “someone else”. Biden (0.6%) at least has all the horror scenarios in which he becomes the incumbent President; I’d rank him above Hillary Clinton. I suspect she’s overpriced because she’s well-known and a former presidential candidate.

The number of junk emails from Newt’s campaign in my inbox has tripled during the past two weeks. He is clearly making his move.

First, thanks (I assume) to US laws against online gambling, it’s quite difficult for Americans to participate in the market, which is, at least for legal purposes, based in Ireland.

I expect that you’re half right — the chilling effect is not from the law against Internet gambling itself (which has been around for 5 years), but rather from the new interpretation and hyper-aggressive federal prosecution this year:

http://sports.espn.go.com/espn/poker/news/story?id=6362238

… for instance, djw’s experience @12, if generalized, would be pretty decisive evidence against semi-strong efficiency.

I would add that while I haven’t followed intrade closely over the last year (and no longer have an account), when I have paid attention to it, it seems that the political prediction markets are following a somewhat less moronic strain of conventional wisdom than they did in 2008, which was a bonanza of free money for anyone with half a clue.

Well, LP, I did say “most of the time”.

OK, I should have added a caveat at the end, like “in this case.” But is there any reason to think that semi-strong efficiency would be less likely for the US presidential race than in other cases? I would think that as one of the most widely followed political prediction markets, with (I assume) some of the highest trading volumes, it would be among the most likely to come close to efficiency. If semi-strong efficiency doesn’t apply here, why assume it applies to other prediction markets?

I agree with djw that these markets tend to over-react to news.

OK. But that’s just another way of saying, they don’t display semi-strong efficiency. :-)

I haven’t been following Intrade, but if Perry is a good buy now, he’s the one you want to back, not Gingrich.

anon/portly,

I live in the US, I consider myself a conservative, I am Republican, I pay a moderate bit of attention to the process, I want Romney to win and in my opinion Romney “is the emptiest of empty suits, a haircut masquerading as a man.”.

“I live in the US, I consider myself a conservative, I am Republican”

If we lived in a more civilized world this pronouncement would be viewed the same way as announcing that one likes to fornicate with sheep.

Romney is smart and competent. What drives the “empty suit” narrative is that, even by the standards of national-level politicians, he seems to have no persistent principles whatsoever. He sounds like a pragmatic moderate, a quasi-liberal or a wingnut ultraconservative demagogue depending on what he perceives to be an electorally advantageous position at any given moment. You have to be this kind of chameleon to some degree to make it, but with Romney it’s unusually transparent.

I’ll remind people that Romney, unlike most GOP candidates, has actual accomplishments.

Sure; he made a multi-billion dollar fortune transferring jobs from the US to China.

Definitely an accomplishment, just not one I like.

@Walt: Okay, but it’s not the market for actual presidents; that’s something people pay $2000/plate of rubbery chicken for a tiny speculative share of. But we talk about InTrade as if it was as predictive as the actual market? What’s the basis for believing that InTrade aggregates all the knowledge available to the economy on the candidates’ chances? Surely most of that knowledge is held by people who are not at all invested in InTrade, and they’re applying that knowledge to completely different (and probably larger, certainly more numerous) investments the prices of which InTrade doesn’t capture. Do we assume that once InTrade’s a certain size it’s a statistically valid sample? I’m not trying to argue here; I genuinely don’t understand the principle.

But inside Romney’s is the perfect vacuum of outer space. McCain and even Bush were much more substantial.

If you compare Romney to November 1999 Bush, you might call them a wash, empty-suit wise (I wouldn’t, because Bush’s business and gubernatorial record seem less substantial and because his father was much more obviously responsible or potentially responsible for his success). But then you have 8 years of Bush as president – are those eight years, and all we know about them, evidence of substance? The first president in my lifetime where people actually talked and thought a lot about the vice president? I would have said no. Maybe I was overly influenced by reading all those bits of evidence to the contrary adduced by Delong. Maybe there were other bits of evidence I wasn’t reading.

And the comparison with McCain surprises me even more. My impression has been that detailed analyses of his senate career show a man who embraced the “maverick” label without actually demonstrating any “maverick” tendencies. I thought the smart, informed view of McCain was the one being made here about Romney, as sort of the quintessential empty suit.

We know relatively little about Mitt, at this point.

particularly Repubs have to dematerialise somewhat to make the transition from primary to general, and their suits tend to flap a bit as a result

But I think this just makes the forming of an accurate impression really hard. If Mitt looks too substantial, maybe that means he isn’t really very smart. See comment 28 (RP).

Google tells that its widely shared on both the Dem left and Repub right.

I would say such views re a more moderate Republican are simply generic – their informational value is nil.

“slightly-obsessed-with-America foreigner” is a little more comprehensible

But still accurate! I didn’t mean it as an insult, merely as a good short description. I still think your paragraph 6, the one about how Newt might actually win, may demonstrate the difficulty of understanding American politics too deeply without living here. I lived in a Commonwealth country for a few years, and though I could have voted, I never did, feeling that I didn’t really understand their politics well enough. Then again my votes in the US have often turned out so poor I have considered not voting here too.

Even if he survives the harassment scandal (which looks less likely now than when I entered the market), Herman Cain isn’t a viable candidate.

What scandal? He’s a Republican. It’s OK if you are Republican. Maybe he’s a viable candidate, maybe he’s not, but scandal has nothing to do with it.

Romney, unlike most GOP candidates, has actual accomplishments. For instance, my family gets health care that we could not afford if we weren’t in Massachusetts.

Since I’ve already made a habit here of repeating, at possibly tedious length, that legislation is done (or not done) primarily by legislatures, I may as well say it one more time, in a state rather than federal context. How much involvement did Romney really have in this, and how should it be interpreted? Massachusetts is a very blue state and might plausibly have gone single-payer like Vermont; instead it has a centrist half measure very similar to the one for which Obama gets so much criticism. That outcome may or may not be near the liberal end of what is achievable in federal politics, but it’s clearly much further right on the scale of what is achievable in Massachusetts.

It seems like a waste to balance out two candidates with zero chance (Paul and Huntsman) with one with zero change (Gingrich). You’re basically making the claim that Gingrich’s odds are higher than Paul’s & Huntsman’s combined. as Casino points out, you might as well go with Perry as the most undervalued of the realistically possible candidates.

It’s really interesting to see the contrast between Watson’s doctrine and jfxgillis’s concrete experience. Watson knows from first principles that bubbles Just Cannot Happen, so he needn’t learn anything about the operation of actual prediction markets.

Gingrich left his wife (with cancer in the hospital) while having an affair with another woman. He is unelectable, and it would appear that most Washington insiders think he is in this for the continued notoriety, which translates directly into private speechmaking money in the future. He is currently up to about $150,000 per appearance.

Romney is a welfare state liberal. His basic problem is that footage exists of him taking both sides of every issue, back to the beginning of his career. There is a bounty of contradictions to use against him in the general election. A month ago, he stated that he has always believed the SAME things, then a week later he stated that it is right to CHANGE your mind, if the facts change — so he has actually flipflopped on the issue of flipflopping. Now he is taking the tack that he is just another human being, which may be some sort of precedent in U.S. elections… Usually, these things would not be good for getting conservative Republican votes, nor is being a Mormon good for getting fundamentalist Christian votes, but there is ample polling evidence that the Far Right is moving into “Anyone But Obama” mode, and may be convinced to overlook such problems (consistency was never their strong suit anyway — something you might get Romney to admit, if you could infuse a few chocolate liqueurs into him).

Latest reports show that Cain is losing points in the polls, due to the scandal.

I would guess that Perry is Romney’s most dangerous opponent, running for the “Whom would you rather have a beer with?” sweepstakes by trying hard to be Reaganesquely charming and folksy, while remaining indecipherable on any issue. (Being a vacuum is essentially identical to flipflopping.) Perry is so low in the polls that he may seem out of the running, but imagine, if you will, that Cain is knocked out because we find that he has propositioned dozens of women: which is the easiest prediction of his behavioral pattern. Then, his poll numbers will be transferred to someone else, and that someone else doesn’t have to be Romney. It is significant in this regard that the Cain camp has accused the Perry camp of bringing the scandal to the public’s attention.

I continue to hold that we are witnessing the symptoms of the Republican Party’s historical cul de sac. The Republicans often anoint their chosen one early and all fall in line, but this time they are slugging it out, and what we are seeing is the manifestation of the internal contradiction: the political-economic propaganda of Hayek-Ayn Rand-Reaganomics is revealed to be practically unworkable and intellectually bankrupt against the altruistic needs of the real world. At the same time, it is a strong ideology which still appeals, not only to the Right, but sometimes to the 7% of U.S. voters who are truly Independent, and who often decide the outcome of elections because the country is so evenly divided. (The other 23% or so of people who RESPOND as “independent” are actually regular party-line voters who very rarely, if ever, switch sides.) The most reliable Republican voters are blue-collars who ended education with high school. They are amply supplied with their fears and resentments from listening to talk radio.

@Watson:

???

I thought a bubble was an asset price appreciation based on speculative expectations of further appreciation.

And I think they should use the chart I linked in #44 in textbooks as a representation of it.

Correct me if I’m wrong, but I think you think the expectations were rational, not speculative. But I disagree. I think it was always the case that the 2000 election was approximately a 50/50 proposition, and you can see the market oscillated in a narrow range around that figure for most the year (rational) until the market got way out of range in both directions (speculative).

Actually, I’ve taken a small position in Bachmann (she’s .005 bid, .007 ask on the IEM). My reasoning: Romney will win Nevada and New Hampshire. A conservative will win Iowa. That conservative will then become the anti-Romney going in to South Carolina. Neither Cain nor Gingrich has any hope of winning Iowa — organization is required. Bachmann and Perry therefore, regardless of polls and funding problems, are the sole possibilities. While Perry is the more probable, he’s not that overwhelmingly more probable.