(Soundbite of TV show "Hardball with Chris Matthews")
Mr. MIKE BARNICLE (Political Analyst, MSNBC): It's basically, potentially, the end of two dynasties in this country, the Bush dynasty and the Clinton dynasty, and it's all happening right here this evening in New Hampshire.
MIKE PESCA, host:
Well, that was pundit Mike Barnicle sounding the death knell for the Hillary Clinton campaign way back in January, right after the New Hampshire primary. So, Barnicle is paid to be a "Hardball" bloviator, but sober-minded elder statesmen were basically saying the same thing. New Hampshire was to be the end. Instead, it was like an Electoral Lords, as Hillary Clinton won big. She rejuvenated her campaign, if you remember back, and she made the so-called experts look pretty foolish.
But pundits make mistakes. The polls make mistakes. That means people make mistakes. So, who's not making mistakes, or making a little fewer? Maybe the question is, are we not invested enough? There is a theory behind why electoral-prediction markets, where the candidates are treated like stocks, or more precisely commodities, often do a better job than polls and pundits.
Barack Obama is trading pretty high right now after his victory in North Carolina. You should have bought last week during the Jeremiah Wright hullabaloo. On the line now is John Delaney, CEO of Intrade. Intrade has been called the king of prediction markets, and it's in Dublin, Ireland. Hello, John.
Mr. JOHN DELANEY (Chief Executive Officer, Intrade Prediction Markets): Good morning.
PESCA: Good morning. Tell me very quickly, how does this prediction market work?
Mr. DELANEY: Well, you explained it quite well. We will list a market on each candidate in a very similar way to a stock, and people trade the candidate, and the price the candidate trades at equals the probability of success. So, for example, if Obama is trading right now at 90, that means he has a 90 percent probably of winning the nomination, whereas Clinton is trading at 10. That means she has a 10-percent probability of winning the nomination.
PESCA: And these people are trading real money? Like I could either invest in IBM or Mike Gravel?
Mr. DELANEY: They absolutely are, and that's what I suppose differentiates a real money market and an opinion poll. When you are asked in an opinion poll survey who you think will win, you may say who you might like to win, or you may say the first thing that comes into your mind to - just to, you know, answer the question. When you sit down in front of your Intrade account, or you sit down to do this, and you have a 100 or 10,000 dollars, or whatever figure you have invested into that person, who you might like to win is very secondary. Who you think will win, and generate a return for you, is the most important thing.
PESCA: Now, my understanding is that election-prediction markets began as a tool created by colleges and universities who wanted to get a better insight as to, you know, human nature and decision-making, but you run it as a business. You're not, I mean, it's nice as a side note if you can offer the public some insight, but you're trying - you get a little money off of each trade, and are just providing a market for people who want to play stocks, and those stocks happen to be Barack and Hillary?
Mr. DELANEY: Yes. That is our business motto, but it is only part of our business motto. We firmly believe that there is value in the information, whether that information is used for, you know, investing or speculation like a stock exchange, or for managing risk. Let's say it will benefit your business if a Democrat wins versus a Republican wins in the presidency, and you might like to know, what is the probability of that?
If you want to know that, you want to get the best possible information, and prediction markets on issues like this prove to be more accurate than opinion polls. So yes, our business is about providing a transaction platform and charging a fee, but it's also about supplying the best possible information.
PESCA: And you can make money not just by guessing right which candidates are going to win, but during the course of a campaign, I could have bought Hillary low and sold her high. She still may not win the election, but I could make money on your site, right?
Mr. DELANEY: Absolutely, the very same way she would with a stock or a commodity, I think, as you compared these markets to earlier. You can short an individual candidate. So, maybe earlier in the contest you didn't know who was going to win, but you were absolutely sure that Gore wasn't going to run.
PESCA: Let's compare the investors, your investors in Intrade, with the public at large. Two weeks ago, CBS and the New York Times did a poll, and they revealed that 51 percent of the public thought Obama would win the nomination. Thirty-four percent thought it would be Hillary, so that would mean that an Obama contract would be trading at 51 cents. But as of two weeks ago, so this was before the big Indiana and North Caroline vote, what where the traders on Intrade pricing Obama at?
Mr. DELANEY: If you give me an exact date, I will give you an exact figure.
PESCA: I think the poll was on the 3rd of April. So, that was actually - oh, no, I'm sorry. The poll was on - let's take, I don't know, April 29th.
Mr. DELANEY: On April 29th, our market gave Obama a 79.6-percent probability of winning. That is profoundly different than the poll you just mentioned.
PESCA: Does this - now it looks like, I mean, we just played a lot of tape. All the pundits are saying Obama is going to win. Does this mean your people were right, or they just got lucky, or a combination?
Mr. DELANEY: Well, that's a great question, and the key is in the term "prediction market." They are not guaranteed, but what I can tell you is based on academic research by some very smart guys from George Mason University and others, prediction markets such as Intrade are more accurate than opinion polls. Let me give you one statistic.
The error from an opinion poll like Gallup, their exit opinion polls, prediction markets, have a 43-percent lower error. Now, it doesn't mean they are perfect, but it means they are significantly better than I suppose the next-best alternative. They are not always right. Sometimes they are wrong, and there is still a 10-percent probability that Clinton could win this nomination, but that's a small probability right now, and it seems to be getting smaller.
PESCA: Historically, in terms of candidates, has anyone with a 10-percent probability ever won in the time you've been setting markets?
Mr. DELANEY: Oh, yes, indeed, and you know, logically that should be the case. If you run a market, you know, a hundred times and, you know, at least one of the one-percent chance is perhaps should actually come back to win this. A one-percent probability continued to do it should turn up at some stage.
PESCA: And do you think - we started this segment by slamming the pundits, but do you think that your people betting Intrade are greatly influenced by the pundits? Or are they smarter than the pundits?
Mr. DELANEY: Again, that's a really good question. Markets are aggregators of information. So if a pundit like Scott Rasmussen comes out with an opinion, our traders will aggregate that information. They will listen to him. They will be influenced greatly, or not at all, or somewhere in between.
PESCA: And they have the ability to discount or put credence into what he's saying, and it's the wisdom of the crowds.
Mr. DELANEY: Precisely. So, just as the New York Stock Exchange will aggregate all the analysts' opinions on IBM or Microsoft...
PESCA: Got it.
Mr. DELANEY: So would our market aggregate the opinions of the political analyst or pundit.
PESCA: John Delaney is CEO of Intrade. Thank you.
RACHEL MARTIN, host:
Stay with us. This is the Bryant Park Project from NPR News. Transcript provided by NPR, Copyright NPR.
