How 4000 Physicists Gave a Vegas Casino Its Worst Week (2015) | Hacker News Reader

How 4000 Physicists Gave a Vegas Casino Its Worst Week (2015)

physicsbuzz.physicscentral.com

215 pointsjchanimal98 comments

When I was a physics student in the late 80s, there was a conference in Reno, held in one of the casinos. I think it was just one of the divisions of the APS, and we certainly didn't occupy the entire place.

I recall an amusing anecdote. Of course we didn't gamble. But after the conference banquet, my friends and I were standing around talking shop, and another guy accused us of being nerds for not at least trying out the slots. So he bought us a bucket of quarters and led us into the fray.

After some time spent in sheer boredom, winning little bets and losing little bets, my friend decided to dispute the result of a machine. The attendant came over, and we got to see the insides of a slot machine. It was a garden variety IBM PC-AT with a color CRT, inside a fancy case with buttons and a pull handle. The attendant pulled the keyboard out of its slot, started typing commands, and showed us the result of our game. The machine was of course right, and the attendant smugly shut the lid and walked way.

Eventually we ran out of quarters and all went to bed.

There were some really cool pawn shops in Reno.

And I have sat many times beside a guy who has no qualms against pumping $2,300/min into $100 slot machines, only to walk out of the casino with far more than he came in with.

He talks to the VIP slot hosts, and to other VIP players. Then he picks the machines that are "loaded up" and "are due to pay out". I've tried to explain that this means nothing. But in his mind, it makes total sense.

In the limited experiences observing him, I cannot say his (unfounded) beliefs are not accurate. That said, I've explained that slots are designed to pay out a certain percentage in a year, based on millions of plays. One high roller weekend is just a drop in that bucket, so _anything can happen_.

It's quite a scene to have several machines tied up because they're on the jackpot, and the slot host is busy working on 1099 forms while you play another "due" machine.

None of this is near as exciting as the MIT blackjack stories, but it is still pretty cool to see what $161k looks like in neatly wrapped stacks of 10k.

Not sure the particulars of your friend's situation, but some slot machines have a Jackpot that increases in value the longer it hasn't paid out. If the machine happens to hit a stretch of not paying out the jackpot for a long time, the EV of playing actually does become positive (assuming you can bankroll yourself until you hit the jackpot).

Some video poker machines with jackpots can actually end up becoming +EV, or in other words you win money (in the longrun) with each play. The immediate payout rate for many of these machines is already quite high with a house edge < 1% given optimal play, so in other words on a $100/play game you'd really be losing less than a dollar per play on average even without any additional incentives. Factor in hotel comps, and it can be +EV even without jackpots. Wagering $100k would only be a house edge of ~$800.

Some of these machines (at least in europe) are actually required by law to average out to a certain return percentage over a set number of bets. This means they need to compensate for naturally accuring bad streaks. (And good...)

You wouldn't see that in most US jurisdictions that have proper gaming laws and review (NV and NJ being the big ones).

The software and underlying math is audited by a third party and verified to produce the desired average result over millions of plays, but cannot adjust the odds dynamically to make sure that percentage is hit.

A proper die doesn’t need to compensate to make sure every side is hit the same amount of times.

This was the best story I have read all day! Thanks for sharing!

If they had a convention of HN commenters in Vegas, it would probably be their best week ever. So many people sure they've found some secret way to beat the odds. It's a casino boss's dream come true.

I've usually left Las Vegas with more money than I started with. I don't try to beat the odds, I try to find a $5 table that's full of drunk and/or newbie poker players. These players won't play the expensive tables but are also reasonably good losers when the stakes are lower. Don't try to game the house's portion ... if you do something against the rules they'll spot you. They won't bat an eye if you're taking money from the other guests though.

There are certainly ways to play Vegas and win more than you lose. But doing so in a way that doesn't require a lot of time is not really possible.

Ultimately, it's just another job. But if you enjoy the social aspect, then it makes sense.

Yes ... and obviously at a $5 table I wasn't playing to make a living. It's an interesting way to pass the time when you're stuck there for a conference anyway (back then I went to Networld/InterOp every year). My main point was really that if you want to exploit any system, focus on the weakest link. Blackjack is played against the house so in the long term, the house will win a small percentage of the amount played.

Here's one bonus observation that I learned the first time I was in Vegas - the papers that are handed out inside the convention center doors are generally related to the conference you're attending. The papers that are handed out outside the doors are NOT.

They had RailsConf in Vegas in 2008. I spent half of my time in the poker room, breaking even, but made a key networking connection in a limo that jumpstarted my startup's success.

Is this satire or are you serious?

What sounds satirical about it?

Because many of the key networking connections made in casinos tend to be of a fleeting, salacious nature. Perhaps you have a nice clean mind and didn't realize what you said ;).

Perhaps in a random casino. I'd during RailsConf your chances of finding useful connections are much better, especially if you're looking for them.

I think these "I found the magic trick to beat the casino" websites and movies are funded by casinos.

Where do you think they'll be filming, and who'll be giving them access.

I'm sure some teams will choose a Vegas film over another just for the perks.

If you don't have a very high opinion of HN commenters, you'll probably assume that they get their opinions from Wikipedia. But that means they'll have a pretty good idea of whether a given game is good or bad in terms of odds.

Besides, casinos will often give you free booze simply because you're sat at a table gambling. You don't have to actually win money to profit! You just need to start sensible, and then keep it going long enough to get drunk.

Actually this is also a documented form of advantage gambling:

http://en.wikipedia.org/wiki/Advantage_gambling

But my personal favorite, as a physics major, is the observation that most roulette wheels are unfair, due to slight asymmetries. If only you could get good surveillance of the roulette tables for a few days (hard to do alone), you could make a killing. On roulette!

I recall a case where a trio of Serbians managed to observe the roulette wheel bias and profit in a short timeframe. From memory their measurements gave them a window of 6 numbers to bet on, so 1/6 odds on a 1/37 roulette wheel. The best part is they were allowed to keep the winnings as they didn't actually tamper with the casino's equipment

https://www.theguardian.com/uk/2004/dec/06/ukcrime.gambling

The casinos are very well designed to identify people who win unnaturally often. They will shut you down if you become a threat to their business model.

If you are satisfied in demonstrating at some low dollar amount that you can predict the outcomes reliably, then enjoy the success. But you won't get rich from it.

I have heard that casinos sometimes have balls of slightly different size/weight.

Basically to throw off anyone who might try this technique, as dealers often will throw very similarly for each spin.

My grandfather worked at a casino decades ago, and they interchanged the roulettes between tables, I can't remember if it was daily or weekly

> Besides, casinos will often give you free booze simply because you're sat at a table gambling. You don't have to actually win money to profit! You just need to start sensible, and then keep it going long enough to get drunk.

Yeah, but you have to get drunk with a bunch of random gamblers. I'd sooner go sober.

You must have missed the article and discussion this morning about a now (but not then) YC Alumni that beat the Massachusetts Cash Windfall lotto and literally made millions off it. They are proof enough that some games can be beaten.

[1] https://news.ycombinator.com/item?id=16494280

funny, whenever I'm stuck at a casino (I've had an inordinate amount of "retreats" for work at casinos), I take $100 out of the ATM and hit up the lowest deck table of blackjack I can find, and spend the evening entertaining myself.

when the inevitable "are you counting cards?" question comes, I simply answer, dryly, "no, I just understand math". it's usually at that point they close the table for the night.

The only career advice we got in my physics undergrad was when the quantum professor told us few if any would go on to PhDs, but we could get jobs in Vegas or Wall Street. It was at that moment that I learned to stick with index funds and never, ever gamble. There's no way one of me for a few hours can possibly beat 100s of me working on a problem for decades.

Playing craps is a bunch of fun for me personally. I treat it like entertainment and just play the pass line plus odds. The pass line is just a 1.41% house advantage. And there is 0% house edge on odds. It's actually cheaper than buying drinks if you are planning to. Also much more fun than staying in your room and staring at your mutual funds the entire night!

You’re obviously going to lose money over time, but what makes craps fun is high variance. It’s pretty common to double or triple your money before losing it all and then some. You can definitely walk away a big winner if you’re just playing for a night or a weekend, though chances are, you won’t.

Slots on the other hand is just a depressing one way money drain and I have no idea why people enjoy them.

I'm a morning person and so, even in Vegas (or moreso since there's no clock in the room), I was up at my normal time. Since I'm also normally on EST, this meant I was awake at 0300 or 0300. The first morning I tried to find breakfast early and found that the slots were full of disabled and/or elderly people who were trying to win their way to riches. And it was the slots that they seemed to think was the best way to get there.

Talk about depressing ... I didn't mind spending a few bucks gambling for the entertainment and comaraderie but these people definitely couldn't afford even a single pull of the handle. Talking to a relatively low-level employee at the casino, she said that the floor was full that day because it was the week that public assistance checks were received in the mail ... and that the casinos shooed them out of the building by 0600 or 0700 so that the regular guests were unlikely to see what was happening.

I’m not sure if many people enjoy them. A lever that dispenses random rewards is basically a purpose built addiction machine.

I saw a woman win 15k last night. On an 88¢ bet on a slot machine.

That’s one reason why people play them: they know people do occasionally win big. Why not them?

I go slightly more risky. Putting a little more at risk by placing a few come bets plus odds too.

Same here. Last time I was at a craps table: pass line, come bet, with odds on both. Made something like $500 in less than an hour ($5 table). Granted, this doesn't happen every time. I do lose money sometimes. But, the atmosphere is fun. Chips flying all over the table... everyone drinking and feeling loose... laughs, jokes, chatting with others... the cheer when the shooter hits... camaraderie... it's great. Definitely better than sitting in your room. I see it as a form of entertainment. Before I go gambling, I plan it out. Determine how much I can comfortably play with. I also expect that I will lose it. And then once the money is gone, I'm done gambling.

I've heard this strategy before - go to the penny slots for the drinks, not the money

Slots is the absolute worst thing you can play with highest house edge and expected loss by far. It's also mind numbingly boring.

Not true; some slots are positive expected value

I've known about a dozen people over a decade who make a living gambling. Apparently it's not as hard as it sounds. Most people just don't try. It takes effort and discipline like anything else. And a healthy appetite for risk.

Surprisingly, most physicists I've talked to also seem to have a healthy appetite for risk. Maybe this is a generational thing. The current generation has one eye on wall st and the other on startups.

By my understanding, successful professional gamblers aren't actually gamblers, as they engineer the situation so the odds are in their favour and they are just printing money. Here is my understanding of how a professional “gambler” (arbitrageur?) in Australia made a fortune.

Take a powerball/lotto game, where balls are drawn and punters win a prizes from a fixed size pool. Say the long term rate of return is 90%. The arbitrageur starts placing large volumes of bets, so they make sizeable portion of the pool of bets (say 30%) and have a significant chance of winning the jackpot. They also negotiate a rebate (say 15%) with whoever runs the game, so their long term rate of return is now greater than 100%. The gaming company wins, as they are bringing in an additional 30% revenue each week from the arbitrageur and only paying out a 15% rebate, and the size of the prize pool is fixed. The arbitrageur wins, as their long term payout is more than 100% due to their rebate. The mug punter (without a rebate) loses, as they are winning a smaller portion of the prize pool due to payouts to the arbitrageur. Meanwhile the lottery company is sucking in more mugs by advertising how many jackpots are going off (winners can be anonymous in Australia). In summary, the arbitrageur's rebate is legally transferring money away from those without power (the mug punters) back to those with power (the company running the game and the arbitrageur). Lose-win-win respectively.

The Australian gambler you're thinking of is most likely David Walsh, who made many millions by deploying similar systems in horse races across South East Asia; he is also noted as a collector of eccentric art.

https://en.wikipedia.org/wiki/David_Walsh_(art_collector)

  They also negotiate a rebate.

Uh, right. How does that part work, precisely?

You go to the company that runs the lottery and ask "hi, if I buy 100k tickets, what sort of discount can I get?".

As long as there aren't laws that prevent selling tickets below the nominal price, it's no different than going to a company and asking for a discount when buying bulk quantity of their product.

A discount, in my mind isn’t quite the same as a rebate, even if the net result is the same.

A rebate usually requires the full retail price to be paid up front, with the cost savings deferred to a seperate transaction, often paid by the vendor, when there’s a retail supply chain intervening between purchaser and the supplier.

Sillysaurus is presently downvoted, so as someone who occasionally pays professional gamblers money:

Clearly it is the case that there are some sole proprietor firms which provide entertainment services. While this is generally not the world's most lucrative market, we have no reason to suspect that it is intrinsically a worse business than the least sustainable sustainable enterprise.

There exists a class of specialized entertainment service provider. They are called "professional poker players" and their job is to fill seats at a poker table so that recreational players like myself can get into a poker game [+]. They are compensated irregularly for providing this service; incompetent professionals are swiftly driven from the market. An incidental portion of their compensation comes directly from casinos; a majority of it comes from people like me.

[+] Unlike some forms of gambling in Las Vegas, poker games are not immediately expandable at all margins. A casino with one dealer and three players for a full-ring game is sad. A casino with one dealer and six players is less sad; the game starts and the casino starts making a (relatively nominal) amount of money.

(The economics of poker are not attractive to a casino by comparison to the other games in the casino; it is mostly treated as a loss-leader to bring people seeking entertainment into an environment designed to extract all money in their wallet and then some.)

Basically, 1) People who enjoy gambling only enjoy it when they're betting amounts of money that actually matter to them.

There are billionaires who love playing really high stakes poker against the best players in the world. Doesn't surprise me, I'd love to run a race against Usain Bolt even though I can guess the outcome.

> their job is to fill seats at a poker table so that recreational players like myself can get into a poker game

...now I'm pondering how common the equivalent of this practice is for e-sports ladder tournaments.

My brother lives in Japan and had a friend in college who made a fulltime living playing Pachinko. He would analyze a machine for awhile, and then code see a pattern in how the balls fell. Some people can see things most people cannot, and they can take advantage of it.

My wife met such a pachinko professional. He worked three 12 hour shifts per week. He would not leave a marginally profitable machine during a shift. The light, smoke and noise were terrible for his health and his income was about normal for a salary man.

In 2001, one of the premier theoretical CS conferences, Foundations of Computer Science (FOCS), held its meeting in Las Vegas at the Tropicana. Only about 200 people attend, but it was memorable:

(1) The conference hall right next to the FOCS conference room was hosting the international lingerie trade conference ! Still, the talks were more interesting and well attended.

(2) I, like many others, didnt gamble, but I watched my office mate from MIT (now a brilliant scientist who recently won the Godel Prize) lose his 40$ (a big amount for grad students back then) at the blackjack table in literally 3 hands (35 seconds).

He said, "I cant leave vegas without at least trying" and he proceeded to show exactly how LV works.

I'd pay to see that movie!

Now's your chance! Also on the HN homepage right now: https://news.ycombinator.com/item?id=16515847

it'd be like the big short but the whole movie is the Vegas scene and they lose money in the end

"The year was 1986 ... ended up settling on Las Vegas's MGM Grand."

This immediately made me think this story is made up: the MGM Grand as described and pictured opened in 1993. As noted in the comments, the actual year is more likely 1988 and the casino likely Bally's.

A minimum of googling found this:

http://physicstoday.scitation.org/doi/10.1063/1.2814928

It was 1986, and yes, it would probably be what is now known as Bally's which, at the time, was called MGM Grand.

"...a documentary I’d seen a while back, where some science-minded gamblers proved that a roulette wheel could reliably be beaten with a timer and a pocket computer.."

I suspect those two gamblers were Claude Shannon, founder of information theory, and Ed Thorp, a physicist who developed a card counting method for blackjack [1].

[1] http://nautil.us/issue/50/emergence/claude-shannon-the-las-v...

Documented in "The Eudaemonic Pie" by Thomas Bass. Doyne Farmer is the key guy as I read it. He and other guys went on to Form 'The Prediction Company' (where they used Chaos Theory to play the stock market very successfully – eventually bought by Swiss Bank/now UBS).

I interned in his group (the Center for Nonlinear Studies) at Los Alamos Natl Laboratory between these two events.

I'd suspect not. Rather, I'd suspect it's a reference to https://en.wikipedia.org/wiki/The_Eudaemonic_Pie , which is about work that was probably inspired by Shannon and Thorp.

"It was an unmitigated disaster for the Grand. Financially, it was the worst week they’d ever had. After the conference was over, APS was politely asked never to return—not just by the MGM Grand, but by the entire city of Las Vegas."

"but you'd never guess why."

Sigh. People with basic math and statistics skills (physicists) don't gamble. What a shocker.

Gambling is entertainment, not an investment strategy. I assume physicists still go see movies, go out to dinner, or read books. Those too are entertainment.

I had a friend in undergrad who would gamble on Friday nights (poker, at the casino, if I remember right). Gamble, but not drink. I would bug him sometimes about the money he was throwing away.

Then one day he said "Tony, you know what the difference is between my Friday night [at the casino] and your Friday night [at the bar]? There's some probability that I'll have more money at the end of the night than I started with."

And I quit bugging him!

Engineers have both basic math and statisitics skills and I've seen plenty gamble.

The "do not play" strategy to win in Las Vegas was pioneered by attendees of the AFIPS sponsored Fall and Spring Joint Computer Conferences and the National Computer Conference. Sponsoring organizations included the ACM, the IEEE Computer Society, and others.

Everyone knows the house has an edge. Gambling is a form entertainment. Unfortunately like drinking some people can’t just do it socially. I find an occasional weekend in Vegas fun. Just budget yourself $x you’re willing to spend per day. Play a game here and there, drink some, go to the pool, eat at a buffet, get a massage, whatever. It definitely took me some time at first to ‘get’ Vegas.

I feel like I get the reward-value of gambling... but whenever I think about it, I think about how much more fun it would be to start a casino.

Is there a good one of those "___ Tycoon" business-sim games for starting your own casino? One where things like comping whales and decide whether to break cheaters' kneecaps are part of the gameplay? Because that would be an addictive game ;)

Yea SimCasino would be fun. There is definitely a rush you get in gambling that is different from video games.

I'm under the impression that the general public understands that casinos and lotteries are not charities. Do you really need a PhD in physics to figure this out? The reason people gamble is not because they're dumb.

People gamble for fun, but they get into trouble because of the dopamine rush.

I don't gamble, or find gambling fun, because of the Cortisol rush.

It's more about the pervasive cigarette smoke for me. Even though poker rooms are now smoke free you still have to wade through the rest of the casino to get there and no air filtration system is perfect. The result is you leave smelling like an ashtray regardless.

I wonder if relative lack of gambling also holds true for other Vegas conventions frequented by a generally technical crowd (e.g. Defcon, etc)

I attended a convention with a bunch of probability & statistics experts. A number of them absolutely cleaned up and it wasn't at all surprising to see who - but I suspect the casino's losses to them were quickly subsidized by their less keen peers who were also playing (albeit badly) and the dozens who were watching and ordering drinks like our plane was going down. Honestly if I was hosting Defcon, nerds winning at the Blackjack table would be the least of my concerns.

Vegas is also much less centered around gambling than it was 30 years ago. With the World Convention Center and other venues, there's not even an expectation that convention-goers are going to gamble, and if they want to, they have to go out of their way to get to a casino.

Apparently it was a similar story for a convention of anti-money-laundering/compliance specialists in accounting.

Was just wondering about Defcon actually. Maybe the amount of parties around Defcon and Blackhat makes up for the lack of gambling, in terms of revenue?

"The only winning move is not to play"

Better still, be the casino.

Better still, be a hedge fund.

A while ago, I would regularly go to a kink happy hour at a local bar. I was talking with the organizer and, apparently they were almost not invited back after the first one; because a significant amount of the attendees wouidn't actually get anything from the bar. (There is now an entrance fee redeemable for equal value at the bar).

Summary: they didn’t gamble.

That article wasn’t as fun as I thought it would be.

still was a good chuckle

Fun factoid. Have you ever noticed how absurdly tacky casino carpet is? This is intentional. The idea is to keep people's eyes up and looking at the casino games.

There's also that characteristic smell many casinos have, and I'm struggling to avoid the really tempting jokes there. That derives from quite expensive fragrances and oils filtered through the air circulation. Once again, they're designed to incentivize gaming with a bit of metascience - musk is a popular component in the scents, for instance.

The games begin when you walk through the door, whether or not you realize you're participating.

And you don't even need to collaborate on that. Every person with enough statistical experience (and scientists have to have that) know by intuition that the odds are stacked against them in a casino. So it's really know fun if you can estimate with each dice or ball roll how much money you're losing (on average).

This article gets its facts wrong. It was the 1986 MARCH Meeting in Las Vegas where this happened. I was there.

A friend of mine overheard the saleswomen talking in the lingerie shop (dressed as you might imagine) "I could be stark naked and these guys wouldn't notice me".

I never play the games, either. Once I took the "learn craps" class where the casino gives everyone (fake) chips to learn with. That was relaxing and fun, and they were all gone in ten minutes.

This seems to get attributed to a better understanding of math, because that's the obvious thing about physicists. But how much of it is just flat disinterest, from either being too smart or too interested in other things, to find a game of chance remotely interesting? I mean how exciting is a coin flip for example? Would I bet money on either outcome? No, because who cares? Same thing for any game a casino offers. It all seems geared for an 1800s or early 1900s attention span. But even a modern slot machine - what is it other than a very simple, weird, short, boring video arcade game with only one level?

I think the same exact thing happened with NIPS 2011 in Reno.

tldr; they got a hotel cheap, because hotels+casino outbid each other when larger groups want to rent rooms and they didn't play so the casino lost money.

It was exactly as I expected although the article tried to shove the idea that they gambled down your throat and stretched an anecdote over one page instead of one sentence

[1986]

TL;DR:

the week of the '86 APS April meeting found the gaming floor almost completely empty, leaving the casino with its record-low take by staying at a gambling hotel but obstinately refusing to gamble

Downvoted? Why?

Just so I can learn I'd appreciate to understand the reasons for the downvoting.

I suspect because tl;drs are often frowned upon. The article isn’t so long, and HN members generally strive for substance and more depth (even if they don’t always achieve it) and tl;drs work against that. That’s my guess, anyway. Mindreading is tough, and those that downvote without comment tend not to comment even when prodded (which is one of the reasons comments on voting is against the guidelines).

And how does a TL;DR stop people from getting more substance and depth? You still have the option to read the article...

Whatever the case may have been, the week of the '86 APS April meeting found the gaming floor almost completely empty, leaving the casino with its record-low take; in the (probably apocryphal) words of one casino waitress: "They each brought one shirt and a ten-dollar bill, and changed neither."

Reply on news.ycombinator.com