The Man Who Shook Up Vegas

January 9, 2007 9:54 PM

The Man Who Shook Up Vegasis a fascinating and relatively brief Wall St Journal article about Bob Stoll, a college dropout whose statistical wizardry seems to be forcing the $96 Billion U.S. sports wagering industry to jump through hoops. Even if you don't follow sports gambling at all, the article is a very compelling read. note: since the wsj.com is subscriber only, link goes to a reprint of the article.
posted by jonson (29 comments total)

The Web is thick with ads for sports handicappers, or touts, who offer guaranteed "locks" and claim to have success rates as high as 80%. Their survival is a testament to the gullibility of the average bettor. "It's a very shifty group," says Glen Walker, a veteran sports gambler and line-making consultant for a casino in Antigua. "At least 70% of them misrepresent their records."

There was a 100% chance this made me laugh out loud.
posted by Blazecock Pileon at 10:02 PM on January 9, 2007 [1 favorite]

This is bullshit right? It says that its "costing the casinos a fortune" when people are betting on winners. But I think sports betting is done in a pool and the casinos get a vig based off the total pool and then split the pool between the winners; theres no way casinos are losing money.
posted by Osmanthus at 10:18 PM on January 9, 2007

Off-topic, but related ... this brings to mind:

Jonny Magic and the Card Shark Kids: How a Gang of Geeks Beat the Odds and Stormed Las Vegas

Bringing Down the House: The Inside Story of Six MIT Students Who Took Vegas for Millions.
posted by ericb at 10:20 PM on January 9, 2007

I think sports betting is done in a pool and the casinos get a vig based off the total pool and then split the pool between the winners; theres no way casinos are losing money.

They probably aren't losing money overall, but someone can certainly cut into their profits. They can only do so much to balance betting.
posted by mosessmith at 10:24 PM on January 9, 2007

A few years ago I aced March Madness. I am both unwilling and unable to devote that much time to watching college basketball again. Especially if it involves listening to Dick Vitale.
posted by Cranberry at 10:34 PM on January 9, 2007

mosessmith: They can only do so much to balance betting.
Care to elaborate? If the casino gets a vig based on the overall betting pool, in what way does it affect them who wins/loses?

Seems to me what this article really does is convince people they have a way to win, so more people will bet and the pool will get larger and the casinos will make more money.

[Also I am certain that the books linkedby ericb have made the casinos much more money than they lose by blackjack counters, for more subtle reasons of course]
posted by Osmanthus at 10:34 PM on January 9, 2007

But I think sports betting is done in a pool and the casinos get a vig based off the total pool and then split the pool between the winners; theres no way casinos are losing money.

Well, the article does have a bookmaker say that his run can't last forever -- meaning that if he affects the market too much, either everyone gravitates to his choices, or the bookmakers have to change their spread and that scares away bettors.

In any case, the article also points out that despite the volume of cash exchanged, it's a low-margin business. Ultimately, sports betting can never be as certain for the casinos as blackjack or even poker.
posted by dhartung at 11:38 PM on January 9, 2007

I'm not really sure I believe the part of the article where it mentions sportsbooks will change their point spread if too many people take one side or the other. From reading Amarillo Slim's biography (he had a sports book syndicate in the late 1950s before the Wire Act made it illegal) he indicates he never worried about what side bettors took, or even if every bettor lined up on the same side for a game. Ultimately, he said, as long as you do a good job with point spreads, you'll come out a winner as a bookie, regardless of what side people bet on. Obviously, of course, this depends on having enough cash to cover the bettors when they pick correctly, but over the long term you'll be fine.

A more likely reason the casinos move the line when a whole bunch of bettors pick one side is there's a chance Vegas is wrong about a particular prediction.

Oh, and Osmanthus: Casinos are scared of Stoll because if his picks are better than those of the sportsbooks, when bettors follow his advice, and his advice is right more than 50 % + the vig the casinos are making for each bet, than casinos lose money because of him. Remember, in the long run, casinos either want half the bettors to line up on each side of the field, or on average have half the games go one way and half of them go the other.
posted by Happydaz at 12:13 AM on January 10, 2007

Care to elaborate?

Well, if the vig is say, 10% (I have no idea how much is actually charged, but it couldn't be much more than that), and they collect $3 million in bets, they get to keep $300,000 off the top. But if $2 million was bet on the winner and $1 million on the loser, they still owe the winners $4 million. Or maybe just $3.6 million ($2mil - 10% x 2), I'm not sure. The article discusses how they use the spread and other methods to try to even out betting, but it's definitely possible for them to lose money on a single sporting event. And winning bets always cuts into their profit margin.
posted by mosessmith at 12:14 AM on January 10, 2007

Mossessmith, yes, it's possible to lose money on a single sporting event as the house. But the idea is the house's bankroll is virtually inexhaustible, so in the end, as long as they're around 50/50, it doesn't matter. As for your math problem, if bettors bet $2 mil on the winner and $1 mil on the loser, the casino is out $800,000. The vig only applies for the payout given to the winner, so they're skimming 10 percent off the $2 million payout, not off the $3 million bet.

So they give the winner back his original $2 million, they take the $1 million from the loser, add $800,000 in casino money and give the aggregate $1.8 million profit to the winner for a house loss of $800k.
posted by Happydaz at 1:22 AM on January 10, 2007

Nice article; thanks for the link, jonson. Glad I'm not into gambling; it sounds like a lot of hard work. It's interesting how statistics freaks like Bill James are using formulas with computers to teach the teams how to draft players and play, and this guy is using computers to tweak his own formulas and figure out how to bet on those same teams.

As for books about beating Vegas, long ago there was a wild story with the odd tittle The Eudaemonic Pie, about some guys who spent years building a computer into shoes that you wore to the roulette tables. It could predict where the rolling ball would land while there was still enough time left to place bets.
posted by LeLiLo at 1:37 AM on January 10, 2007

I think it's funny how Casinos whine about people like that MIT group when their entire organization/business is designed to fleece YOU.
posted by evilelvis at 2:18 AM on January 10, 2007

I think what Bob Stoll would tell you, if he's a decent statistician, is that there are lots of people who have been right about sports wagers 60% of the time since 1999. There's an anti-Stoll out there, with his own highly refined statistical models, who has hit 40% of the time over the same period. Random outcomes + a sizable population = outliers in both directions. This doesn't mean Stoll's future picks are going to be better than anti-Stoll's.

This is the same phenomenon that happens with portfolio managers. There are enough of them that you're likely to have some that hit their wagers for a few years running, however imprudent those wagers may have been. These individuals are rewarded by handing them enormous stacks of money to torch.
posted by Nahum Tate at 3:13 AM on January 10, 2007

This is the same phenomenon that happens with portfolio managers.

not to mention, gambling is sinful
posted by matteo at 3:31 AM on January 10, 2007

You bet it is.
posted by Smart Dalek at 5:49 AM on January 10, 2007

Ultimately, he said, as long as you do a good job with point spreads, you'll come out a winner as a bookie, regardless of what side people bet on.

Unless, all the bettors are following a statistician who consistently does an even better job than you do on the point spreads.
posted by caddis at 7:15 AM on January 10, 2007

Why do you need a "smart" handicapper at all? Why can't the sports books automatically adjust the spreads as bets come in? Almost like a stock price moves depending on the demand.
posted by smackfu at 7:28 AM on January 10, 2007

Off-topic, but related ... this brings to mind:

Both good reads, but as a gambling book maven, they aren't the best on this particular topic.

I enjoyed The Odds: One Season, Three Gamblers and the death of their Las Vegas, and everything by Michael Konik is a pure pleasure, but I've only just learned about his latest so I'm still waiting for Amazon to deliver The Smart Money: How the worlds best sports betters beat the bookies out of millions.
posted by PeterMcDermott at 7:36 AM on January 10, 2007

The sports books are counting on the vig, and the rest is cake.

The only reason they make lines is to split the bettors as close to 50/50 as possible, so the losers pay the winners. Everyone pays $11 to win $10, so if the line does its job the losers pay the winners and the house rakes the extra dollar.

A lot of people don't realize that all the casinos share the risks between themselves, too. If one casino's bets are lopsided, they can "lay off" some of the bets with another casino whose bets are more even. That, combined with the fact that the casinos essentially have bottomless pockets, means that they will always win in the long run.
posted by Benny Andajetz at 7:36 AM on January 10, 2007

And, the lines do change based on how the betting is going - all the way up to the event.
posted by Benny Andajetz at 7:40 AM on January 10, 2007

Last year I won an NCAA office pool betting on "George Mason" because I thought it had a funny name.
posted by delmoi at 7:48 AM on January 10, 2007

I'm a firm believer that less information is better when it comes to sport wagering. As Malcolm Gladwell might say, picking winners is not a puzzle, it's a mystery, and when you get bogged down in data you miss the forest for the trees. I never bet games I know "too much" about, or too little. And I wish this paragraph ended "And I am a millionaire because of it."

Hey, it does!
posted by stupidsexyFlanders at 7:51 AM on January 10, 2007

Now, if only that Pacino movie about sports betting was half as interesting as this article. There's certainly some good drama to be had if you piss off enough casinos, right? You don't have to throw in all the ridiculous crap that the movie did.
posted by thanotopsis at 9:19 AM on January 10, 2007

Fortune's Formula tells the tale of the guy who literally wrote the book on blackjack card counting. He later transferred his work extremely successfully to arbitrage and other investment schemes.

It's a good read for those who like geeky stuff, those who like gambling and those who like investing.
posted by Bitstop at 9:27 AM on January 10, 2007

Last year I won an NCAA office pool betting on "George Mason" because I thought it had a funny name.

Nothing's funnier than Gonzaga.
posted by OneOliveShort at 9:29 AM on January 10, 2007

Ball State is funnier than Gonzaga.
posted by spicynuts at 11:02 AM on January 10, 2007 [1 favorite]

I think what Bob Stoll would tell you, if he's a decent statistician, is that there are lots of people who have been right about sports wagers 60% of the time since 1999. There's an anti-Stoll out there, with his own highly refined statistical models, who has hit 40% of the time over the same period. Random outcomes + a sizable population = outliers in both directions. This doesn't mean Stoll's future picks are going to be better than anti-Stoll's.


Except it's a non-random sample. The losing side is stacked with from the article "squares" who will lose slightly more often than they will win, given the books' spread. Tack on the fraudsters and it is possible legitimate analysis could give an edge over the books without being simply an outlier.

Where I depart from Stoll is his idea that sports betting be viewed as investing. Investing in the long term demands low fees and the presumption that the market rises as a whole over the long term. In Stoll's best year, 30 percent of his 'investments' returned -100% with a 'management fee' of up to 2500 to stoll and the books' fees. If any comparison to the market can be made it should be to day trading, with the stark caveat of those many day traders who lost their shirts in the early 00s.
posted by Andrew Brinton at 11:27 AM on January 10, 2007

Osmanthus: I think sports betting is done in a pool and the casinos get a vig based off the total pool and then split the pool between the winners

This is known as parimutuel betting. It is fairly standard in horseracing but is definitely not how the sportsbooks mentioned in the article (i.e.: football, basketball, etc...) are run. The vig is taken off of each bet, not out of the pool of payouts. So, if the pool of bets is unbalanced in the wrong direction, the vig can easily be eaten up by the payouts.

Benny Andajetz: The only reason they make lines is to split the bettors as close to 50/50 as possible

According to this article, while the safe strategy is to split the bettors 50/50, it appears that it is not uncommon for the sportsbooks to attempt to take advantage of underlying biases in the betting population and tilt things one way or the other -- essentially placing a bet themselves.
posted by mhum at 1:01 PM on January 10, 2007

Technical ingenuity in the pursuit of ill-gotten gain isn't anything new. From Strange Inventions Used by Crooked Gamblers, Popular Science, November 1933 [IE users may need to scroll down to see content]:

I learned recently from my friends among the police, signals flashed from packages by electricity is but one of many ways in which science is now used by crooked gamblers. With electromagnets controlling the roll of dice, with specially-compounded invisible ink marking the backs of cards, as well as with new adaptations of daub boxes and chink ink; tap dice and split coins, table bugs and glims, they are fleecing the unwary.

Via Modern Mechanix.
posted by cenoxo at 4:21 PM on January 10, 2007