The Wisdom of Crowds Redux —Bookies and Bettors

February 5, 2017
Posted by Jay Livingston

Several posts in the early years of this blog (e.g., here ) looked at the “wisdom of crowds” – the idea that the collective wisdom of large numbers of interested people is usually more accurate than the guesses of a few experts. Each post focused on a single event, usually a football game, where the public favored one side while the “smart money” (a small number of professional gamblers) favored the other. 

My thesis was that at least in sports gambling, the crowd was not so wise. If it were, it would have put a lot of bookmakers out of business.

The only data I had, unfortunately, was anecdotal – a few games, like the 2010 Superbowl, where the public heavily favored one side and lost. But this season, I’ve compiled a more complete data set – all NFL games. My indicator of the crowd’s opinion is the change in the point spread late in the week – from Friday to kickoff.*  If the spread goes up, it’s probably because the public is betting the favorite. The bookies are raising the line to attract more money on the underdog and thus balance their books. (On most bets the bettor puts up $110 to win $100. The book with equal amounts on both sides – say $1100 on the favorite, $1100 on the underdog –  is guaranteed a net of $100 no matter who wins.)

I looked at games this season where the line moved by at least one point.** Here are the results.

If you had bed against the wisdom of crowds, you’d have won 54 bets and lost 32. Putting up $110 to win $100 on each of the 86 games, you’d have come out $1880 to the good on a total investment of $9460 – about a 20% return. And except for the first week – 1 Win, 3 Losses – the whole season you’d have been in the black, playing with house money.

As for today’s Superbowl, there has been no movement in the line. It opened at 3 two weeks ago and has stayed there.*** Small bettors are tending towards the Patriots, larger bettors towards the Falcons, so the money is about evenly distributed. Of course the deluge of bets in the next few hours could change that balance.

My own hunch is that the Falcons will win it on the field.


----------------------
   
* Line changes early in the week are usually caused by large bets from “sharps.”

** The change in the point spread is not a perfect variable. For one thing, different books put out different lines. I used the consensus number. For another, bookmakers now respond to betting imbalances by changing the odds rather than the point spread. For example, in today’s Superbowl, the point spread is 3. If a book is getting too much action on the Patriots and needs more Falcons money, rather than raising the line to 3½, they will adjust the “vig.” Bettors usually think of the vig as a tax on losing bets. If you win, you get $100. If you lose, you pay $100 plus the tax – usually 10%. But to balance the bets, a book might raise the vig on the Patriots to 15% or 20% and lower the Falcon bettors’ rate to 5% or even 0%.  In this case, the unchanged point spread would be misleading. The public would be betting on the Patriots, but the line remains at 3.

*** Books are very reluctant to change a point spread of 3. It’s the most common outcome – out 10% of games are decided by three points. If a book raises the line to 3½ and gets a lot of action on the underdog, and if the final score is 20-17, the book loses all those 3½-point bets while not collecting on the 3-point bets. That’s one reason that when the line is 3, books are much more likely to adjust the vig rather than the points.

The Language Anachronism That Nobody Notices

January 27, 2017
Posted by Jay Livingston

The opening of “Bridge of Spies” shows us New York, 1957. Federal agents tail Rudolf Abel as he walks through the streets and now into the Broad Street subway station. Here is a screenshot.

Hollywood does this sort of thing so well. Every period detail is perfect – the cars, the clothes, the street signs and advertisements, the subway station signs, the shoeshine stand,* even the candy bars inside the candy machine though they are on screen for less than a second. When the Feds come to arrest Abel a few minutes later, his Brooklyn apartment breathes the same authenticity. Ditto his false teeth (Abel is just coming out of the bathroom in his underclothes). The script continues.

One of these two lines is an anachronism – the equivalent of having someone drive up in a Toyota. It’s “need to.” I’ve mentioned this before, but once I became sensitized to it, every time I now hear “need to,” the actor may as well have shouted it.

Before 1970, “need to” was not an imperative. We told people that they “had to” do something, or that they “should” or “ought to” do something. You’ve gotta remember, this is 1957.

This chart from a post in The Atlantic by Benjamin Schmidt about the language in “Mad Men” shows  the relative use of “ought to” and “need to” in selected scripts all set in the 1960s. Some of them were written in the 60s, others in this century. The simple need/ought ratio is all you need to figure out which is which.

I checked a couple of those old scripts (“The Apartment,” “The Hustler” – both are great movies). The “need to” count is basically zero. And if Schmidt had used “have to” instead of “ought to” the differences would have been even more exaggerated.

My own speculation (here)  on why “need to” became so widely used starting in the 70s is that it was part of a general shift from a language of morality to a language of therapy. But I have no idea why the change went unnoticed. The lead scriptwriter on “Bridge of Spies,” Matt Charman, is only 37 years old. He grew up in the “need to” world. But the other writers, the Coen brothers, are in their sixties, and Spielberg, the director, is 70. They too were ignorant of the change from the language of their youth.

“Need to” appears fourteen times in the script. One of these lines manages to use it in tandem with yet another anachronism. Donovan (Tom Hanks), the American lawyer enlisted by the CIA to negotiate the spy exchange, is speaking with a Russian official.

“Conversation” – in the sense of a full exploration of issues and positions and options – is, I think, very recent. In 1957, governments may have had “discussions” or even “talks,” but they did not have conversations. 

-----------------------

* The shoeshine stand is on the platform where people stand waiting for their train. I wonder what happened when the train came in before the shoeshiner had finished. Of course, this is the Broad St. station, and on the BMT lines, there was probably plenty of time between trains. (And by the way, if anyone knows what year it was when the subway system finally stopped using the IRT, BMT, IND designations, please tell me.)

You Can’t Argue With a Joke

January 23, 2017
Posted by Jay Livingston

Of all the responses I’ve seen to the Trump/Spicer claim that the inaugural drew the biggest crowds in history, this one – from a hockey game in Dallas –  was by far the most effective.

Whoever runs the Jumbotron for the Dallas Stars deserves a Peabody. The attendance figure pokes fun and deflates Trump’s assertions but without being derisive. The factual criticism that followed Trump’s and Spicer’s performances can be disputed, as Spicer tried to do.  Even if the “facts” that Team Trump presents are false, at least there’s an argument about who’s right. Besides, Kellyanne Conway may have gotten some sympathy for the way that journalists pounced on her “alternative facts.” How would you feel if a bunch of smart-ass reporters checked your every word? 

The Jumbotron avoids those traps. You don’t notice it right away. So a second later, when you do notice the attendance figure, you feel like one of the in-crowd that gets the joke. You’re on Jumbotron’s side. If you laugh – how could you not? – you already share the assumed story behind the humor: that Team Trump is lying about the numbers. Game over. If Trump and company argue with it, they come off as tedious and tendentious. Imagine Trump ranting about how the Jumbotron is the most dishonest scoreboard in history by the way. Imagine Spicer and Conway offering alternative facts about the hockey game attendance. They’d just be digging themselves in deeper while showing that they are utterly humorless.*

-------------------
* I make no predictions. Trump may still tweet something about this.

Why Deny the Obvious Lie?

January 22, 2017
Posted by Jay Livingston

Is there method in Trump’s megalomania?

Why would Trump say things that are obviously untrue? Not only untrue but easily demonstrated as untrue, like his claim that his inauguration had drawn the largest crowds in history. Photographs clearly showed that the crowd on the mall at Obama’s first inauguration was larger.

This screenshot is from an interactive graphic (here ) that allows you to slidethe dividing line back and forth to see the whole mall for both inaugurations.

Trump, as we have come to realize, never admits that he was wrong. And now he has a press secretary who does the same. Yesterday, Sean Spicer repeated the false claim.

This was the largest audience to ever witness an inauguration – period.

He criticized other estimates, saying,

No one had numbers, because the National Park Service, which controls the National Mall, does not put any out.

Spicer did not take any questions, so we can’t know whether he realized the contradiction between these two statements. If no one has numbers, how can Spicer be so sure that Friday’s crowd was the largest in history?

Spicer went further.

We know that 420,000 people used the D.C. Metro public transit yesterday, which actually compares to 317,000 that used it for President Obama's last inaugural.

Reporters checked with the actual Metro statistics. Spicer was lying.*

Why would Trump leave himself open to headlines like this? (Click for a larger and clearer view.)

The psychological answer is that he can’t help himself. He really believes that everything he does is the most stupendous, and he makes up evidence to support his beliefs. To protect his ego from contradictory evidence he launches vigorous attacks on those who provide contrary factual evidence, even – as with his “landslide victory” or the size of the crowds at the inauguration – when the truth is easily available to anyone. Other politicians would be embarrassed to have their statements exposed as blatantly false. But Trump cannot be embarrassed by the truth because he cannot be embarrassed by anything. He is shameless.

But Ezra Klein at Vox thinks the attack on facts is not just psychological, it’s also strategic.

the groundwork is being laid for much more consequential debates over what is, and isn’t, true.    Delegitimizing the institutions that might report inconvenient or damaging facts about the president is strategic for an administration that has made a slew of impossible promises and takes office amid a cloud of ethics concerns and potential scandals.    It’s not difficult to imagine the Trump administration disputing bad jobs numbers in the future, or claiming their Obamacare replacement covers everyone when it actually throws millions off insurance.

The strategy certainly works among Trump’s supporters, the folks who get their news only from Fox or right-wing Internet sources. For the rest of the public, it will depend on the strategies that the media take for reporting on an Administration with so little respect for facts.

-------------------------------
* The New York Times (here) has documented a few other falsehoods from Trump and Spicer, including Trump’s false assertion that during his speech the rain stopped and the sun shone and that “it poured after I left. It poured.” In fact, the rain was light and continuous throughout –  no sunny skies, no downpour.