Stossel the Fool

Townhall.com ^ | November 16 | John Stossel

[Kaslin]

,/center>

I was so dumb last week.

I wrote my column Tuesday -- before election results were in. I assumed Hillary Clinton would be president-elect.

I looked so stupid.

On Facebook, commenters pounced: You owe Trump an apology! I'm sorry for the lies you continued about him! You were never fair! You're nothing but another left-wing mouthpiece. You're a washed up, anti-American gutless TV host!

I was wrong because I trusted the bettors.

That's usually not dumb. The best predictor of things has been betting markets. They are more accurate because they reflect the wisdom of crowds. Crowds can be an ignorant mob, but crowds do have wisdom. Know the TV show "Who Wants To Be A Millionaire"?

When contestants are stumped, they may ask the audience for help or an expert. The experts are often brilliant specialists. The audience -- well, they are the kind of people who wait in line in the rain to watch a game show. Still, the audience gets the answer right 91 percent of the time, the experts succeed just 65 percent of the time.

With betting markets, the crowd is made up of people willing to put their money where their mouths are. That makes them extra careful.

Most of these "prediction markets" are based overseas because, useful as they are, American law calls them "illegal gambling."

So producer Maxim Lott and I converted European betting into an easy to understand website, Electionbettingodds.com, and I've come to trust it. Again and again, betting is more accurate than pundits and polls -- until this election.

I'm not the only one who got it wrong. The Huffington Post's statistical model gave Clinton a 98 percent chance of winning. The prestigious Princeton Election Consortium gave Clinton a 99 percent chance.

People just lie to pollsters when they think the pollster will sneer at them if they say they're voting for someone smugly described as racist and sexist.

This was the second time this year that betting markets were wrong. Most bettors thought Brexit would never happen -- people in Britain would vote to stay in the European Union. Again, British voters lied to pollsters because they were embarrassed to admit they would vote for Brexit after months of the elite telling them they were xenophobes and racists if they wanted a change.

Relying on the betting markets, I also wrote that it was sad that freedom-loving senators like Wisconsin's Ron Johnson lost to command-and-control bureaucrats like Russ Feingold.

Oops, wrong again.

But the prediction markets are right most of the time.

Consider what happened early in this year's Republican primary. Ben Carson surged to first place in polls, but the bettors knew better. They never gave him more than a 9 percent chance. In 2012, when Rick Perry, Newt Gingrich and then Herman Cain surged to first place in polls, prediction markets correctly said Mitt Romney will win. In 2008, bettors correctly predicted results in every state but two. In 2012, it was every state but one.

The markets even predicted when Saddam Hussein would be captured. Right before his hideout was found, the odds on that date tripled in price. Somehow, people with skin in the game pay more attention and intuit the right outcome.

Even last week, when bettors were wrong, the betting odds still adjusted faster than pundits on TV did. The bettors saw what was happening and quickly hedged their bets, while many in the media -- mostly Clinton supporters -- still clung to their failed expectations.

My failure won't make me abandon prediction markets and go back to trusting pundits or opinion polls -- or internet commenters who had fun trashing me:

"Dewey beats Truman ... oh wait."

"I can't laugh enough at this article."

"I liked Stossel ... but he is as clueless as the liberal media."

I sure was! But I will still trust prediction markets over everything else.

There is wisdom in crowds, especially crowds that put their own money on the line.

[Kaslin]

Shut up!!!

[DoughtyOne]

I made screen captures on this site starting on 09/14/16. I captured the screen anywhere from one to five times per day.

Why? Because I knew this would happen. I knew the site would either disappear or the figures would be fudged.

I am taking this information from my 11/08/16 17:55 EDT screen capture.

Hillary 84.4% up 11.7% (lst wk), Trump 15.3% down 11.2% (lst wk). This is a little off from Stossel’s time capture.

My next capture was at 11/08/16 22:53 EDT.

Trump 81.6% up 55% (since 11/01), Clinton 17.7% down 56.6% (since 11/01)

As you can tell, the percentage moves in the five hours were in excess of these amounts.

[Kaslin]

Oh he probably did. Many of his colleges at FNB are Trump supporters

[litehaus]

“This story is bizarre. What point is he trying to make?”

Ah,com’on...That he should be able to keep his job..simple...

[going hot]

I bet on d trump in October of last year, 7:1 odds, made a nice profit.

[JediJones]

I think the predictions were wrong because the polls failed to account for the popular/electoral vote split. Everyone was looking at national polls over and over. State polls were done and discussed much less frequently, and even less frequently were they compiled together by anybody to make an electoral prediction.

Granted, the people who did compile state polls still tended to show a Clinton win. However there were far fewer state polls to work with and they showed much more radical swings in results than the national polls, indicating unreliability.

It was a safe bet that the national polls wouldn’t switch to Trump’s favor rapidly enough for him to win the popular vote. But it was not a safe bet to think that Trump could not micro-target a few states right at the end and get his numbers up in a narrow area enough to win the electoral college.

[schurmann]

John Stossel and the other prognosticators failed to predict events because the event they were trying to predict was then still in the future, which (at the time of their guessing) had not happened yet. Nothing is predetermined.

As a pofessional pursuit and a pastime, prediction is irreducibly probabilistic. No amount of data collection, no application of statistical methods, no reading of tea leaves or sheep entrails, no flipping of tarot cards, will overcome that truth.

Humans vary, and are inherently unredictable.

[riverdawg]

“Who bets $84.00 to win $100?”

Someone who thinks the outcome they bet on is almost a sure thing.

There were many Trump voters, including myself, who thought he was a long shot to win the Electoral College. One can hold that opinion and still bet on a Trump victory if the odds are favorable enough.

An important difference between a betting site and an election is that the former registers the intensity of preference by the participants (bettors), whereas in an election each participant (voter) is counted the same.

I think a simpler explanation for Hillary! being the odds-on favorite than that her supporters were “gaming the system” is that her true believers thought they had a near-sure thing and bid the odds to 5 or 6 to 1 in the process.

[going hot]

I see you made some financial arrangements as well! Nice cleanup D1!

[China Clipper]

Easier to read

Thank you!!

[econjack]

He did win OH, but I was driving around before the election.

[pa_dweller]

People just lie to pollsters when they think the pollster will sneer at them if they say they're voting for someone smugly described as racist and sexist.

This.

[sig226]

I don’t know if they purposefully twisted the polls. I don’t accept the claim that people wouldn’t answer them honestly. It’s an anonymous phone call. I’ve done phone polls, honest ones and push polls. Push polls are pretty obvious and I think most of us hang up on them, but if I have time, I’ll talk to the honest ones.

I was polled for this election and I hung up on them. I was doing something else and I didn’t feel like dealing with them, so I don’t know if the poll was legitimate. I also didn’t want anything to do with the media during this election. They polled me in the middle of the wikileaks releases and I was thoroughly disgusted by then.

My cell phone has a Palm Beach County area code. I wonder if that’s why they called. I also wonder how many people refused to even talk to them.

[econjack]

It’s an anonymous phone call.

Really? So there's no way for the caller to identify the person? I kind of doubt that.

[Rome2000]

Ain’t nobody got time for that!

[DoughtyOne]

I think that’s a decent alternate opinion, but I still don’t think people bet $840 dollars to win $1000.

There’s to much to lose, and these folks got their clocks cleaned.

[DoughtyOne]

I hope you did as well. Easy money...

The best part, is knowing it isn’t sitting in a Leftist’s pocket anymore.

It’s sitting in mine.

[going hot]

gravy, LOL!

[DoughtyOne]

“;^)

[Gene Eric]

It was crunch time, Stossel, and you behaved like indolent teenager. Were you wasted the entire month of October?