Why Intrade failed

March 12, 2013, 12:07 AM UTC

FORTUNE — So it’s the obvious joke.

The one thing that Intrade.com, the company that up until Sunday night allowed its traders to bet on everything from Saddam Hussein’s capture to the next American Idol or Pope or U.S. president, did a very poor job of predicting, it seems, was its own demise.

Although it kind of did, or at least its founder did. John Delaney, who Fortune profiled along with Intrade back in 2005, regularly complained that U.S. regulators had it out for his company, which was based in Ireland, because either Americans are generally prudes when it comes to gambling, or regulators didn’t like people betting on U.S. elections or a combination of the two. The official explanation from the CFTC, which effectively banned U.S. citizens from using the site in November, was that Intrade’s contracts kind of looked like futures contracts and we have decided that in the U.S. things that kind of look like futures contracts should be regulated (except, of course, the type of bespoke credit derivatives that blew up in the financial crisis, which Wall Street continues to argue are too special to be regulated).

So if it’s a witch hunt that finally did in Intrade then perhaps it’s for the best that Delaney didn’t get to see it. The founder of Intrade died climbing Mount Everest, reportedly within 50 yards of the summit, nearly two years ago.

MORE: Why Dell doesn’t like Icahn’s plan

Those who believe in rational markets and the so-called Wisdom of Crowds, which is what Intrade and prediction markets in general are based on, have taken a beating in the last few years. First there was the financial crisis, which neither the robust market for subprime mortgage bonds or the stock market in general saw coming. Then there was the rise and quick demise of the so-called Twitter hedge fund, which had planned to beat the market by analyzing millions of tweets and deciding whether people were feeling good or bad about the world. It lasted less than a year. Now comes the end of Intrade.

But the rise of Nate Silver and his FiveThirtyEight blog was perhaps the biggest blow to Intrade. In 2009, a study found that Silver, who uses statistical models, rather than markets, to predict elections, did a better job forecasting the outcome of the 2008 U.S. presidential race than Intrade. During the last election, the predicted chance of Mitt Romney winning on Intrade skyrocketed after President Obama’s poor performance in the first debate. Silver’s statistical model barely budged. One guy proved to be a whole lot smarter than the crowd.

But Intrade allowed traders to bet on many more things that just elections. And other events, like the naming of a Pope, which Intrade got right the last time around, might be much harder to build statistical models around.

Joe Weisenthal over at Business Insider essentially lets Intrade off the hook for its miscalls. He says that Intrade, like all markets, reflects conventional wisdom, and gives you the ability to bet for or against it. But that misses the point. The theory behind prediction markets is that they would do a better job than conventional wisdom. As more people placed their bets the market would move toward what the actual outcome would be.

MORE: How Citigroup won (and Goldman lost) in the Fed’s stress tests

But that’s often not how it worked out, though it’s not clear why. Intrade’s biggest failure was perhaps predicting that Obamacare had an 80% chance of being struck down by the Supreme Court, up until the moment it was announced that it was upheld. It’s worth noting that failure happened a little over a year after Delaney died.

While he was alive, Delaney was able to make a cogent argument why Intrade was good for the world and not a den of thieves. (Intrade, on Sunday, said that the site had to close because of “financial irregularities,” so Delaney and Intrade may end up being remembered as the latter.) Delaney reasoned that what people were doing on Intrade should not be considered gambling. That seemed to keep U.S. regulators at bay, for a while. But nonetheless, Intrade’s legality was always in question, which kept some users away and made it a less successful market than it could have been.

So the real reason Intrade failed may not be because prediction markets don’t work. But because, based on our current laws, it’s still more profitable and unquestionably legal for people like Nate Silver to use their statistical models to launch blogs than to spend their time and efforts on Intrade. Until that changes, we are all probably better off without it.