The sound you hear may be the world's smallest violin

January 2, 2013 3:06 PM

"Las Vegas bookmakers make their money by balancing their risk, but sometimes they simply come out on the wrong side of too many bets." With the regular 2012 NFL season now over and the playoffs about to begin, please take a moment and shed a tear -- or more likely, raise your beer -- as you consider the terrible beating Las Vegas sports books absorbed in 2012. (LAT link, so potentially behind a paywall depending on your number of previous visits in last 30 days.)

The losses for sports books were especially bad back in Week Nine of the NFL's regular season (the weekend of October 27, plus/minus a few days) when many popular NFL teams beat the point spread, resulting in big payouts for bettors who not only bet their favorites, but bucked the odds and parlayed those individual bets into winning tickets with much higher odds, resulting in big individual payouts.

Of course, casinos not only have the math on their side, they also set the odds at which they offer these bets* and thus aren't supposed to lose, which is what makes this story interesting.

Bonus multi-part history* of sports betting in America: Part 1, Part 2, Part 3, Part 4

*Some of the above content comes from sites that promote or otherwise support sports betting, which may make the starred (*) links NSFW.

posted by mosk (30 comments total)

"casinos (...) aren't supposed to lose"

I thought about this when I first heard about this and imagined that Vegas was actually just coming close to breaking even.
posted by boo_radley at 3:11 PM on January 2, 2013

Here is a counterpoint to the LA Times article, which seems to use actual facts to argue that the books didn't lose, but just made less than normal.
No The Sports Books Did Not Get Killed During the NFL Season
posted by split atom at 3:20 PM on January 2, 2013 [10 favorites]

I don't understand why the casinos didn't alter the lines to balance the bets. Unless they have a gut instinct too and want to gamble with the gamblers money.
posted by Keith Talent at 3:22 PM on January 2, 2013

I don't know if there is any line that will get more bets against Notre Dame than for them. And Notre Dame won a lot this year.
posted by smackfu at 3:27 PM on January 2, 2013

Interesting link, split atom. I would have included that in the FPP had I seen it. This post mainly stemmed from reading the LAT article, which just plain surprised me. But then, I'm a sucker for "man bites dog" stories...
posted by mosk at 3:28 PM on January 2, 2013

use actual facts to argue that the books didn't lose

OTOH, they only have numbers for the NFL season through end of October.
posted by smackfu at 3:28 PM on January 2, 2013

The LA Times, Yahoo, USA Today, and ESPN are not my go-to sources for anything concerning Vegas.
posted by Ardiril at 3:33 PM on January 2, 2013

Yeah, I thought the whole point was that the books balance the betting lines. That's what betting lines ARE; they're not educated guesses, they're a numerical expression of how much money is risked on either side. If they're doing it right, they don't "win" or "lose"; they take their cut off the top either way. It's the bettors who win or lose.
posted by Fnarf at 3:39 PM on January 2, 2013 [3 favorites]

split atom's link is worth reading. I don't think they really can lose, can they? Aren't they more or less the middlemen between bettors? That is, if 7 people want to bet Packers for every person that wants to bet 49ers, don't they basically offer 1:1 odds for Packers fans, and 7:1 odds for 49ers, minus a shave?

Or is it more complex than that? I thought the whole idea of being a bookie was never to risk your own money, only other peoples'?
posted by Malor at 4:02 PM on January 2, 2013

Heh, I should have read Fnarf's comment before posting. In other words: what the guy said, right above me. :)
posted by Malor at 4:03 PM on January 2, 2013

I don't know if there is any line that will get more bets against Notre Dame than for them. And Notre Dame won a lot this year.

Every line I've seen has Alabama a 9-10.5 point favorite, so it seems early money is overwhelmingly against the Irish.
posted by BitterOldPunk at 4:10 PM on January 2, 2013

I bet on sports fairly regularly, and I generally can keep a little ahead of the bookies' cut in my W-L record. I'm in agreement with split atom's link. The numbers aren't out, and the bookies might have actually lost money on Week 9 of the NFL season. But I would put down my life savings that the bookies won on Weeks 10-16, and that they are doing just fine for the NFL season as a whole.

Also, the bit about:

Now, Rood is sweating out hundreds of Super Bowl winning bets made on teams such as the Seahawks, who opened at 75-1 odds.

I assure you that they are not sweating these out. Seahawks are currently at 9-1 on winning the Super Bowl. There is a very good reason they had 75-1 odds at the beginning of the season: no one was betting on them save a few homers. Even still, the Seahawks are 9-1 odds, the sixth favorite out of 12 to win the Super Bowl.
posted by Mister Fabulous at 4:10 PM on January 2, 2013

Yeah, I thought the whole point was that the books balance the betting lines. That's what betting lines ARE; they're not educated guesses, they're a numerical expression of how much money is risked on either side. If they're doing it right, they don't "win" or "lose"; they take their cut off the top either way. It's the bettors who win or lose.

The books have to set a line before they can take any bets at all. They adjust it based on betting, but they can take a bath on early bettors.
posted by Etrigan at 4:22 PM on January 2, 2013 [3 favorites]

Notice how in the LA Times article, Kornegay doesn't really say anything at all. His quotes don't really refer to books losing money. This is the kind of press leading at which Vegas marketers excel.
posted by Ardiril at 4:25 PM on January 2, 2013 [1 favorite]

I seem to recall an article by someone (maybe one of the Freakonomics guys?) who went and researched how sports betting lines are set. They found that sports books often didn't try to do the obvious thing (i.e.: balance the bettors on both sides of the line). Apparently, the vig alone wasn't enough for them. What they were actually doing was setting up unbalanced lines, essentially investing in/betting on themselves and against the wagering public, pitting their own (presumably superior) knowledge against the masses. While there are, of course, sophisticated gamblers out there, there were/are enough chumps out there to make this a profitable enterprise.

The only other thing I remember from this article is that one of the times the sports books don't do this is for the Super Bowl. Something about it being the most scrutinized match-up of the year making it hard to have that kind of information edge.
posted by mhum at 4:33 PM on January 2, 2013 [1 favorite]

This is my schadenfreude face. :-D
posted by Splunge at 4:38 PM on January 2, 2013 [1 favorite]

I thought bookies adjusted their odds by waiting until I made a bet, thereby guaranteeing the opposite result would occur.
posted by fullerine at 4:42 PM on January 2, 2013 [2 favorites]

mhum, I've read one sportswriter make that claim, but he didn't do a very good job of explaining it. It's something I've wondered about, though. The sportsbook I, ahem, browse gives stats on who's betting what. I often see numbers like 70% of bettors taking one side.
posted by lost_cause at 4:45 PM on January 2, 2013

There is no way for sane bookies to lose money guys. You set odds based upon the betting, not the game's expected outcome, aka use the martingale measure. In fact, anyone bookie who made less than usual drank their own cool aid and tried betting against the plebs.
posted by jeffburdges at 4:53 PM on January 2, 2013 [1 favorite]

For example, only 27% of the action is on Alabama -10 in the BCS championship game. It seems like the line would adjust more quickly. 27% of what is unclear - number of bets or total money- and of course, they could just be making the numbers up. Still, it looks weird if all they're really trying to do is balance the numbers and take their 10% off the top. Maybe the math does actually balance over all their bets offered on everything, and they just tweak the numbers one way or another on a given game to pull in suckers.
posted by lost_cause at 5:08 PM on January 2, 2013

Ah. Here it is: Why Are Gambling Markets Organised So Differently from Financial Markets (pdf) (google cache)

From the paper:

First, I demonstrate that the bookmaker does not appear to be trying to set prices to equalise the amount of money bet on either side of a wager. In almost one-half of all games, at least two-thirds of the bets fall on one side of the gamble. [...] A rationale for this failure to equalise the money emerges in the paper’s second finding: bookmakers appear to be strategically setting prices in order to exploit bettors’ biases [...]

My findings provide an explanation for that empirical regularity: it is profit maximising for the bookmaker who sets the spread.

posted by mhum at 5:09 PM on January 2, 2013 [8 favorites]

I don't think they really can lose, can they? Aren't they more or less the middlemen between bettors? That is, if 7 people want to bet Packers for every person that wants to bet 49ers, don't they basically offer 1:1 odds for Packers fans, and 7:1 odds for 49ers, minus a shave?

A little math correction: They'd offer 1:7 for the Packers and 7:1 for the 49ers. This is an "odds" or "moneyline" bet, not a spread bet. Let's say this were the case where there were 7x the people betting on the Packers than the 49ers. The odds would be pushed closer to 1:10 for the Packers and 6.5:1 for the 49ers so the book gets their cut. If every one of these eight fans bet $100 there would be $800 in the pot. If the Packers win, the Packers fans would get their $100 back, plus ~$10 in winnings. That leaves $30 left over for the book to keep. If the 49ers win, the 49ers fan gets their $100 back, plus $650 in winnings., the book keeps $50.

In spreads (what this largely talks about) the books set the a point total where they think it will make the people bet equally across both teams. The underdog is given more points, and the betting proceeds. The general rule is that the bets are referred to in the American moneyline form of -110. This implies that to win $100, one must bet $110. Let's say two people bet, one on each side. The total pot would be $220. The winner would receive his initial bet of $110 back, plus the $100 in winnings, totaling $210. The other $10 goes to the book. If the bets become lopsided they might set the moneyline to -120 (bet $120, win $100) for the favorite and +105 for the underdog (that is bet $100, win $105).

Or is it more complex than that? I thought the whole idea of being a bookie was never to risk your own money, only other peoples'?

The issue that arises is that the bets don't always come in even despite what the books might want. In some cases, the bets will wind up somewhere like a 3:1 lopsided bet. The books want to keep the money coming in that might be scared off by putting down a spread that would be scared off by an otherwise reasonable 20 point spread. In the Week 9, some of the games included:

Arizona 17 @ Green Bay 31
Tennessee 20 @ Chicago 51
Denver 31 @ Cincinnati 23
Baltimore 25 @ Cleveland 15

I list these particular ones because many of those teams that won are favored very heavily by fans. Fans will pick Green Bay or Dallas or Chicago or Denver versus a spread no matter how high that spread gets (within reason). As a result, the books have learned that if they set the spread high enough, the odds turn in the favor of the underdog. In the long term, the underdog in these games becomes the winning bet. If the bets are lopsided, the books have no choice but to be the "bettor" that supports the underdog.

This particular Week 9 had a lot of teams with 1. strong fan bases 2. were all strong favorites and 3. won their games handily and beating the spread. The bets were heavily lopsided towards the favorites, and favorites went 10-2. In many of these games, the books were on the hook for a big chunk of the underdog bet. The books had to pay out.

As for the books complaining: they all should know that statistically speaking they will lose their ass every so often. It happen to hit Week 9 2012 of the NFL season. It will happen again some day.
posted by Mister Fabulous at 5:16 PM on January 2, 2013 [23 favorites]

Thanks, Mister Fabulous! That's an extremely cogent explanation and one I wish I had found when I put this FPP together.
posted by mosk at 5:20 PM on January 2, 2013 [1 favorite]

Here's a really good article about a professional gambler living in the Philippines who makes a mint betting against gamblers in Hong Kong.
posted by awfurby at 5:55 PM on January 2, 2013 [2 favorites]

Just a note here. Legal books have a lot of cash to back their bets. Illegal books have different ways of covering their losses. FYI.
posted by Splunge at 7:21 PM on January 2, 2013 [1 favorite]

Sportsbooks are essentially never balanced on NFL lines. Some books don't even strive to be balanced, since they know that they can exploit systematic biases in the betting public, but most books would love to collect a smooth 4.5% on every game with perfectly balanced action, but that is impossible. They also have significant limits on their ability to adjust the line. The central problem for them is that if they initially set the line at a certain number like Washington +3, they run a huge risk by moving it. If they move it to +3.5 and the game ends on a 3 point differential, they have to pay all the +3.5 bettors and they end up with a push with all the 3 point bets. This could result in huge losses. Sports bettors say that the house has been "sided" when the final score ends up exactly on one extreme of the spread. Roughly 15% of all NFL games end up with 3 point differences, so there is a huge incentive for books to avoid moving lines that are close to 3 points.

Even worse than being sided is being "middled." Imagine that everyone is betting on the Vikings +7.5 such that a sportsbook has an aggregate exposure of $1 million on Minnesota. They move the line to +7, but money continues to pour on the Vikings and they end up with an imbalance of another $1 million on Minnesota, increasing their cumulative exposure to $2 million. Desperate to even out their risk, they move the line to +6.5 and finally the big Packers backers show up and they get the full $2 million exposure balanced out and they now are "even" between the Packers and the Vikings on net. Unfortunately, the Packers end up winning by exactly 7 points. This means that the $1 million on the Vikings +7.5 gets paid, the $1 million on the Vikings +7 is returned as a push and the $2 million on the Packers -6.5 is also a winner, so the book has now lost $3 million on the game. The net result is that there is significant inertia on lines and that books almost never want to move a line at 2.5/3/3.5 or 6.5/7/7.5 because of this problem.

When online sports betting was easy in the US, I used to have a computer program that searched for these kinds of discrepancies between different books and would make large wagers on both sides simultaneously in order to try to find these risk-free opportunities. They normally only existed for a few minutes until one book or the other would move back into line. Certain sportsbooks, like Pinnacle, would clearly be willing to step out of line with the other books to attract money on one side or the other. Essentially, they were courting one-sided action because they believed that the line was not accurate. I noticed that my arbitrage bets almost always involved Pinnacle on one side and that the money flows were incredibly uneven. On average, when Pinnacle took a line outside the norm, they won and the money flowed from the other books into Pinnacle. So, I'm quite sure that at least some sportsbooks actively court uneven betting.
posted by Lame_username at 7:25 PM on January 2, 2013 [9 favorites]

I came in to LOL about that Onion article, and got the most intense, concise and cogent explanation of sports betting in pretty well all of its permutations and business models I've ever seen or heard anywhere, in one clean hit. I mean, Wow! Thanks, MeFites! Come for the Sizzle, stay for the Steak.
posted by riverlife at 11:38 PM on January 2, 2013

Splunge: " Illegal books have different ways of covering their losses. FYI."

Ooh, story time! Do tell. Did you get a side of beef instead of $500? Or a car? Was it drugs? Coke?
posted by boo_radley at 9:35 AM on January 3, 2013 [1 favorite]

For one thing, they often have other income streams and flexible accounting practices.

And not that I have any experience in this, but a friend of mine was once told flat-out by his bookie that he couldn't pay him, so he'd have to wait or roll over his winnings into the next week. Period. No apologies, no bargaining, no incentive offers, just, "I ain't got the money. Maybe next week." The part about "What are you gonna do, call the cops?" wasn't vocalized, but it was definitely there.
posted by Etrigan at 10:02 AM on January 3, 2013

Heh. Well this guy once told me...

A bookie who will tell you that he just can't pay you may have one of several reasons. He's new and alone. He has a very small bank and overextended himself. He's not connected to anyone and he's running the chance of getting hurt very badly. Lethally even.

He knows that you are a noob. You have no juice and he's not worried about you blabbing to anyone.

He is connected and he took a bath. But the guys upstairs want their cut and don't give a fuck if he sinks or swims.

Otherwise he'll get backing and pay it back at a ruinous rate of interest (to him).

The biggest illegal books have lots of employees and a big bank. They also give everyone different odds depending on if they are high rollers or nickle and dime schmucks.

Or so I have heard.
posted by Splunge at 4:25 PM on January 4, 2013