Right now the electoral college markets are tracking the polls pretty closely. I think we’ll see some divergence when we get close to the election since the polls can’t keep up. In past elections the markets were much better than the polls within a few days before the election (though not on election day itself which has too many rumors).

Other thoughts:
- There is a long-shot bias —states which are above 85% (for one candidate or the other) reflect a probability closer to 100%.
- The state markets are strongly correlated. There is a small but non-trivial chance that many states will be way off this year. And then people will be reluctant to trust the electoral college markets in the future.

So, I have (at least) one answer to my series of provocative questions: Electoral college prediction markets are more useful than the state polls towards the very end of the presidential campaign (but not on Election Day). Interesting. Thanks.

PS: The discussion about this post goes on in the comment area of another post.

- The bad point is that, at the time he wrote up his column, Virginia governor Tim Kaine was the favorite of the InTrade VP prediction markets. The others were, in decreasing order, Evan Bayth, Kathleen Sebelius, and then… Joe Biden. So, the critic reading his column today could say that the prediction markets are oversold to a gullible public and that a prediction market bubble ready to pop up is forming under our very nose.

- Now, we know that Barack Obama made his decision while vacationing in Hawaii (less than 2 weeks ago). That’s only from that date that the VP prediction markets started generating probabilistic predictions worth quoting. The trick is that Justin Wolfers (and the other prediction market analysts) didn’t know that, on August 1st. (PDF file)

- I don’t regret my decision not to publish about the VP prediction markets. I’d look like an idiot today.

[...] Mr. Obama’s decision had as much to do with Mr. Biden’s appeal among white working-class voters and compelling personal story, and his conclusion that the Delaware senator was “a worker.”

The important information in the NYT piece is that Barack Obama personally called governor Bill Richardson “late last week” to announce him that he was not considered anymore. That’s around the time the Joe Biden rumor began to have more weight in the media circles —see the InTrade chart below.

Bo Cowgill, back in May 2008 (when I started to act as a prophet of doom):

This is dumb. Cover them if something interesting happens. Maybe your theory will turn out to be wrong. Anyhow: Although the decision is made in secrecy, the Presidential nominees have a number incentives which we have plenty of information about. Specifically:
* They want someone who will balance their tickets in terms of geography, race and class.
* They want someone who will help with weak areas of their campaigns.
* They want someone who will be a good campaign surrogate — giving good speeches and attacking the opponents effectively.
* They want to avoid a VP who will de-motivate or offend the base.
* They want to avoid someone with a bunch of skeletons in the closet such as angry ex-wives, out-of-wedlock kids, etc.
* Etc etc.
Anyhow, I don’t see any reason to ignore these markets in case something interesting happens. I read Midas Oracle so that I don’t *have* to read a whole bunch of other websites!

Bo Cowgill was on the right track, now that I think of it —in a society where everything leaks out.

On the opposite of the spectrum, Tom Snee was too much extreme in his view:

According to Tom Snee of the Iowa Electronic Market, at Iowa University, futures markets need more hard information than they get in the veepstakes, to reliably predict a result.

Markets are very good at predicting elections, he says – but not choices being made inside Barack Obama’s or John McCain’s head.

Justin Wolfers was more measured.

So, Bo Cowgill and Justin Wolfers are the winners, on that one.

I was partially wrong. I am a bit too extreme, sometimes. (Did someone else notice that? ) I need to learn more about… granularity.

-

PS: On the Republican side, now…

Who will be the Republican Vice Presidential Nominee?

Some blogger says his wife is fantastic.

New York Times portrait of Joe Biden.

UPDATE: Barack Obama‘s speech + Joe Biden‘s speech

-

I think it is the worst pick ever. What a blunder. Joe Biden (a D.C. insider) is unpopular and gaffe prone. Plus, that choice shows that Barack Obama is insecure when it comes to foreign policy. An emphasis on the economy and, thus, on a successful gubernatorial experience would have been better.

Kathleen Sebelius was the one to pick. She is my vice president. (And Ron Paul is my president. )

-

I have over-estimated the secretiveness of Barack Obama’s decision process. The chart above obviously shows that the Joe Biden narrative leaked out to reporters was beamed out for a purpose: testing the Obama-needs-a-VP-who-is-strong-in-foreign-policy argument, and letting the Press do the final vetting on gaffe-prone Joe Biden.

InTrade CEO John Delaney (along with the HubDub and BetFair people) will now brag on his marketing material that his prediction exchange did forecast Joe Biden as the Democratic vice president nominee.

What’s bad in all that (other than I have an egg on my face [*] ) is that we won’t have a public debate on the different quality of the various primary indicators, and how that conditions the accuracy of the prediction markets.

[*] I have an egg on my face, but Caveat Bettor has a whole omelet on his.

According to Tom Snee of the Iowa Electronic Market, at Iowa University, futures markets need more hard information than they get in the veepstakes, to reliably predict a result.

Markets are very good at predicting elections, he says – but not choices being made inside Barack Obama’s or John McCain’s head.

Thank God for the BBC.

Thank God for the Iowa Electronic Markets.

Shame on John Delaney —over 3 generations of Delaneys.

Other than Tom Snee (the IEM spin doctor), Chris Masse and Justin Wolfers are the only prediction market analysts to have sent out warnings about the VP-candidate prediction markets.

• creating suspense (sometimes false) so as to generate free publicity;
• sending a positive message to the supporters of each VP candidate;
• letting the Press do the vetting of the VP candidates;
• flattering the political journalists by leaking to them;
• sending out false leaks so as to preserve the surprise for the scheduled announcement day;
• sometimes, buying time to impose the head of the VP search committee as the most serious VP candidate (remember Dick Cheney in 2000). [Psstt... Funny enough, in the 2008 election, Michael Moore is pulling for Caroline Kennedy. ]

All that means that there are no good primary indicators for the prediction markets on the Democratic and Republican VP-candidate selections.

I want to offer 6 remarks:

1. Not all prediction markets are created equal. Some have good primary indicators (e.g., the prediction markets on the presidential elections, thanks to polls), while some other prediction markets have unreliable primary indicators (e.g., the prediction markets on who will be on the ticket).
2. The prediction exchange executives (like InTrade-TradeSports CEO John Delaney) will never tell you that, because their job is to sell their wares, of course.
3. The public needs prediction market analysts, who can judge the quality of the primary indicators of one particular prediction market, so as to separate the grains from the shaft —reliable prediction markets from unreliable prediction markets. (A prediction market analyst has also other functions, which I will blog about later on.)
4. A prediction market analyst should have a dual competency —in a vertical (in our example, US politics), and in prediction markets.
5. The expertise in the vertical (here, politics) should be a major, and the expertise in prediction markets should be a minor. Take a look at these 2 mainstream media news stories: the one written Jack Shafer in Slate (which I linked to at the top of this post), and the one written by Justin Wolfers in the Wall Street Journal. Obviously, the one that shows the most mastering is the one written by Jack Shafer, an American professional journalist who follows US politics for a living.
6. The consequence of that for prediction market journalism is that the writer should be an expert in a vertical, and the editor should be an expert in prediction markets —and not the other way around.

That said, I wish the very best of luck to our good friends Caveat Bettor (who is betting on Tim Kaine) and Nigel Eccles (who is predicting Joe Biden).

UPDATE: My (informal) Democratic VP-candidate bet is on Kathleen Sebelius. Hint, hint.

UPDATE: Gawker says that Joe Biden would be a horrible choice. I agree. Plus, he has denied to be the pick. He could have lied to reporters, though.

UPDATE: New York Times publishes portraits of all VP candidates.

DEVELOPING…

NEXT: While InTrade CEO John Delaney is deceiving the journalists to sell his wares, Tom Snee of the Iowa Electronic Markets is telling them the truth: BEWARE THE VP-CANDIDATE PREDICTION MARKETS, THEY JUST AGGREGATE RUMORS.