Hacker News new | past | comments | ask | show | jobs | submit login
Sorting Out the Law Behind Phil Ivey's Edge Sorting Debacle at Borgata (pokernews.com)
94 points by andrewljohnson on Apr 22, 2014 | hide | past | web | favorite | 45 comments

So if I understand this correctly, the casino alleges:

1. The casino was using cards which, when viewed edge-on, were not rotationally symmetrical. (A separate complaint against the card supplier.)

2. Over multiple rounds of play, as players handled the cards, they rotated them to encode a high/low value before returning them to the dealer.

3. The players requested an automatic shuffler so that the cards rotations were not randomized by the shuffling process.

4. Players observed the edges of the "shuffled" card stack to determine whether incoming cards were high/low, using this information to improve their odds.

Players were not allowed to handle the cards. In this case, they were playing mini-Baccarat. In traditional Baccarat, players are allowed to touch the cards and even bend and turn them all they want, but all 8 decks are thrown out after every shoe (i.e. touched cards are never played again). In mini-Baccarat, players do not touch the cards, and the cards are reshuffled instead of thrown away. This type of scheme only works in mini-Baccarat (and Punto Banco - a mini-Baccarat variation), and the casino would only comply with the unusual requests that made this scheme possible for very big players. Casinos are very tolerant of high rollers - Caesars Palace was fined $250,000 a few years ago for allowing a man to jump up on a Baccarat table and dance before placing his bets.

In this case, Ivey had an Asian woman with him, and she told the dealer that she was superstitious and asked that the dealer turn cards of certain values a different direction. That she was Asian was important because the casinos are claiming that Ivey requested that his dealers speak Mandarin Chinese, and that the instructions to turn the cards were given by the woman in Mandarin so that supervisors wouldn't hear and immediately object to the request. In any event, the dealers complied. This incredibly simple strategy apparently worked at Borgata and Crockfords in London, to the tune of over $20 million in combined winnings. Because the casino could have simply said "no," I find it extremely unlikely that the Borgata will prevail here.

This was much more of a social engineering hack than anything else. Interestingly, because of this issue, casinos (at least in Las Vegas) have purchased cards with a new type of background that looks like pixelated noise - much harder to do this with.

Do you have any links to the pixelated cards?

This dancing story is nuts:

  > A complaint filed  by the board against Caesars Palace  says a customer was
  > playing baccarat in the high-limit baccarat room on Oct. 10, 2009. On three
  > separate occasions, the man climbed onto the baccarat table from his chair,
  > walked on the table and made a  bet before returning to stand on his chair,
  > eventually sitting down, according to the complaint.
  > On the  second occasion, the player  performed a dance on  the table before
  > returning to  his chair, the  control board said. The three  incidents took
  > place over a 45-minute period, according to the complaint.

(I think this is what you are referring to. But it is from 4 years ago.)

Link to complaint: http://gaming.nv.gov/modules/showdocument.aspx?documentid=30...

Yes, that is the incident I was referring to...I guess I just read about it last year. Here is the new card back: http://prntscr.com/3cjo5v (here's a closeup: http://prntscr.com/3cjlqk). This is much different than the diamond design they used until recently that enabled this scheme - http://prntscr.com/3cjnne

They do not punch a hole/heart out of the middle of these decks anymore? When I was younger my grandfather would always bring me decks of cards and they all had holes punched out of the middle.

No, these days on the cards they sell in the gift shops they shave 2 of the edges off. They still stick out if put in with other, non-shaved cards in live play at the tables, but are more useful than cards with holes in them :).

What a ridiculous lawsuit. It's barely removed from a bank suing a customer because the customer walked in and said, "Can I have a million dollars?" and the bank gave them a million dollars.

Minor technical correction:

Players are not typically allowed to handle the cards in punto banco, so would have needed the dealer to turn the cards for them.

This, I understand, is one reason they preferred a Mandarin-speaking dealer.

Sometimes you're given the option to deal out the cards yourself, but this would be more likely to break Ivey's careful edge sorting. If it was offered, I would strongly tend towards declining it.

Other thoughts:

Traditional card counting might still have been useful, assuming only the top card would be readable.

Finally, this is only what Ivey has admitted to. I politely suggest that there are other ways of getting an advantage.

This. A previous article stated that Ivey and his friend talked the dealer into rotating some cards (6s 7s 8s and 9s) prior to the first shuffle.

I don't think any dealer or casino would agree to a total re-orientation before the first shuffle/game.

My impression from prior coverage was that the rotation occurred as the cards were revealed during play, so that on each next shoe, more and more (potentially all ) of the intended cards were reversed.

The way dlss suggests is the safest and most sensible way.

Usually you'll start play by breaking the seal on new decks which will be fanned out to check them: it's effective and efficient to get the dealer to turn a few sequences of cards here while they're still in the right order.

Otherwise, you have a lot more work to convince the dealer to rotate the cards one-by-one: "rotate that 9 of clubs, not the other 9 of clubs" is way more suspicious.

Having said that, it is a bit surprising that "dealer collusion" spidey-senses were not tingling. This is the first thing that casinos should watch out for. Maybe they realised something was going on but couldn't figure out what, or how it would be effective, in time.

Sure, if the dealer/pit agrees, that'd be the best way... it just seems even more blatant, and the idea the casino would comply even more outlandish.

Whereas, making requests during the game, that seem a bit erratic (or triggered by in-round results, like "let's flip that one so it doesn't burn me again!"), seem like they might escape scrutiny longer.

Cards almost always come pre-shuffled in the packs that casinos use for baccarat.

Can you link me a source to that article? Phil Ivey is perhaps the greatest living card player in the world. Do not underestimate the near autistic savant level of perception and mental capacity that he and other genius card players have. I can imagine Ivey still being able to gain an exploitable edge merely by simply recognizing the direction in which specific cards are returned into the deck.

"The lawsuit claims that Ivey and his companion instructed a dealer to flip cards in particular ways, depending on whether it was a desirable card in baccarat. The numbers 6, 7, 8 and 9 are considered good cards" from the article linked by https://news.ycombinator.com/item?id=7580399

FWIW I think almost anyone could gain an exploitable edge by recognizing the direction in which specific cards are returned into the deck if they took it seriously and practiced... it just sounds like a lot of work, and a way smaller edge. I personally think it's more impressive that he socially engineered the dealer.

It's kind of like "good artists copy, great artists steal" only for statistical edges. maybe "good gamblers work hard for an edge, great gamblers have an edge before they sit down"

Yeah, this is extremely interesting to me either way. By knowing which way only one card is facing, you still tilt the deck in your favor ever so slightly. After using the same deck over multiple rounds, I assume you tilt the edge in your favor significantly without even having to give directions on specific cards. It would add to the casino's case for them to argue that he manipulated the dealer to make it seem more akin to an inside job, but I can see how Ivey could gain a significant advantage without even needing any "social engineering."

"it just sounds like a lot of work, and a way smaller edge." If you read Ben Meizrich's Bringing Down the House, you wouldn't believe the degree to which these guys work to get an advantage over the house. There's a part in the book where he talks about how members of the MIT Black Jack Team so thoroughly practiced cutting a deck that they were able to cut stack of decks exactly 52 cards from the bottom. A seemingly improbable feat to do regularly but one that they perfected to the point that it became an integral part of their strategy.

I completely agree the details are interesting -- there are very few things that focus my attention as completely as statistical decision theories.

If you haven't read Ed Thorpe's The Mathematics of Gambling I think you'd love it. It's about finding these edges regardless of the game. Thorpe is fascinating too -- he invented one of the early blackjack counting systems, and figured out how to beat the game of roulette (as crazy as that sounds). He was also one of the first stat arb guys, figuring out how to price warrants and making a killing for the UC Irvine retirement fund.

Also, if you haven't heard of the Kelly criterion, I think it'll basically blow your mind. I think it's the least appreciated house edge there is, and a very cool piece of math.

Thorp is pretty interesting. With Claude Shannon he invented the first wearable computer:

"It's with Shannon that Thorp would revisit a question he had considered years earlier: whether he could apply mathematics to beat the game of roulette as he had done with blackjack. Thorp and Shannon would develop a friendship and, in the process of answering that question, build what is widely regarded to be the first wearable computer."


Naturally, they are returned in a uniform manner. To get any rotation (for edge differentiation) would require the ability to handle the cards (which he did not have) or the ability to convince someone to rotate cards (which is what it is alleged they did).

Additionally, Phil Ivey is an astonishingly good poker player (20M+ club, 8 braclets), but is known more as a degen on other games.

That was my understanding as well.

5) profit

you understand it correctly.

I think the writer deserves a ton of credit. This article was informed, technical, and funny.

What I don't understand is how much information the players got. Did they have to learn each half of the deck? That seems absurdly hard. I get the feeling Phil Ivey was just along for the ride. My guess is the chinese guy could watch how the cards were sweeped, adjust in his head which cards were in what half, and recompute odds based on the backs. That is a bat shit crazy talent.

Deserves a ton of credit for record length sentences. A lot of those paragraphs are two 50+ word sentences.

They no longer award credit for that, after a high profile incident involving a couple of poets out of jersey who were caught gaming the system with single-sentence sagas.

I remember that case. Semicolon and Sons sued Sammy "Sentence" Splice, saying his single-sentence sagas seemed somewhat salaciously un-salubrious.

Hah, there was a similar case in Vegas where the face cards had a darker back than the non-face cards[1] and some players noticed it, and started exploiting it. Of course no suits were filed, they just told the players not to come back.

[1] The mechanism was root caused to printing the front of the card first, then the backs, the card company did this to save money (fronts were all the same but the backs were different) but all of the face cards were on one side of the sheet so as the sheet went through the rollers the face cards (which have more ink on them) slowed the cards down slightly giving a slightly darker back (I couldn't tell the difference but folks said they could)

I have to wonder in such cases as what you describe, and the lawsuits against Gemaco: who owns the card company? Or who might have compromised the company's manufacturing and quality-control?

(There was another lawsuit against Gemaco, for not providing decks that were pre-shuffled as required: http://www.pokernewsdaily.com/atlantic-city-casino-loses-law... )

  > baccarat players  have a  long and  tawdry history  of introducing  some truly
  > bizarre superstitions  into gaming  parlors. Gently blowing on  the dice  at a
  > craps table is downright sensible behavior compared to some of what goes on in
  > baccarat pits on a daily basis,

Can anyone provide some examples of these tawdry and bizarre superstitions?

See the comment above https://news.ycombinator.com/item?id=7631481 about casino allowing dancing on the table - it's a few levels above blowing on the dice.

How helpful is linking to a comment that I wrote? I take it that you are less familiar with baccarat than I am, otherwise you would surely be able to provide an example that I did not write. Do you think the table dancing is a common occurrence? If you read the official complaint that I included in the comment that you linked to one gets the impression that the table dancing is not a common occurrence at baccarat tables.

The core problem here is management at the Borgata is being pressured to increase Atlantic City revenues, despite competition from regional casinos and an overall downturn in gambling.

To try and boost numbers they will accommodate strange requests for changes to games without fully understanding the reasoning behind the changes. This is the same method Don Johnson used to win $5m from them at Blackjack.

This has some similarities to recent high-profile "hacking" cases:

"Other claims are interrelated, which is to say that certain claims depend on the success of others to survive. And, at core, most of these lose all oxygen if a court finds, as I suspect it will, that 1) there is no contract between a player and a casino; 2) a casino deviates from its own well-oiled protocols at its own risk; and 3) exploiting house vulnerabilities is not a form of “swindling and cheating” any more than an automatic shuffler is a “cheating device” (please cast aside memories of one particularly memorable scene in Ocean’s Thirteen, and appreciate that no one is alleging Ivey or Sun to have planted a corrupt shuffler in the Borgata)."

So what if we compare this case with, say, accessing a website (e.g. the Weev or Aaron Swartz cases)? Do you have an implied contract with sites you access over the internet, are online resources published "at your own risk" and is 'exploiting house vulnerabilities' (e.g. disclosing information publicly that you intended to be private) a swindle?

On all three counts, it seems clear that the same principles apply.

If anyone is interested in the advantage Phil Ivey had in the Baccarat game, it can be found here (disclosure it's my site): http://www.uspoker.com/blog/putting-iveys-baccarat-session-m...

Thanks, this was a fairly informative number crunch!

If you're interested in the mathematics behind this kind of stuff you should check out Eliot Jacobson's blog [1]. He does mathematical analysis of edge sorting, hole carding, card counting, loss rebates, etc.

[1]: http://apheat.net/

Now, what I don't understand is that Phil Ivey is one of the top 5 poker players in the world. He depends on action in many of these casinos, some of whom will no longer even allow him on the premises.

He's endangering his own livelyhood. Is he so far in debt that it's necessary? Did he blow his bankroll and then some? What's the backstory that he's willing to risk this much?

Now, Binion's is going to allow him in the WSOP, but how much action will he lose?

Most of his action these days is in Macau. Also, a significant portion of Ivey-level action in Las Vegas takes place in the Ivey room at Aria. I seriously doubt they will bar him from playing in the room that literally bears his name. He wasn't cheating, and I don't think this will seriously impact his playing opportunities.

"When you're smart, people need you." - Real Genius

I would bet that on average, any given casino wants Ivey more than he wants them, because he's a draw, and there are more casinos than super-famous poker players.

Particularly with poker, where the house has no stake in the game itself.

It occurred to me that the main analogy to card counting is poor, regardless of the law, but I found the article interesting.

In card counting, you never have precise knowledge of what cards will come... it's always statistical, and you don't manipulate the cards with your hands to create the signal.

It sounds like with this method, you literally know the next card to come and base your bet on that knowledge. That sounds more like marking cards... like if they were playing poker, and they were denting the Aces with their fingernails.

This scheme, also, was imperfect: it made a few high-value cards look different, but only over time (and I'm not sure the dealer's indulgence rotated every card, or just some).

Even if all the high-value cards were the only ones rotated, it didn't reveal exactly which value would be next, only that it would be one of the more-interesting ones. And that much information didn't guarantee a win, just a better chance of a win, when all the other (unpredictable) cards came out. (There are no choices in this game other than how much to bet, on either the 'player' or the 'dealer'... so only one card's worth of extra info, the top card in the shoe, at the beginning of each round.)

So it's all still just statistical.

Also, apparently Ivey never touched the cards: he requested the dealer reorient them, and under the full observation and assent of the casino. (It wasn't a secret side conspiracy with the dealer.)

He didn't say it was like card counting. He said that it exists on a continuum, somewhere that is clearly less legitimate than card counting but not obviously cheating like bringing loaded dice.

Your analogy is actually worse; the cards came pre-marked. So they didn't mark the cards, but they did go out of their way to exploit the existence of marked cards.

It can't be considered marking cards because they are not physically altering them.

It is considered "edge sorting" because that is exactly what it is. This isn't new, and if you mention the term to any gaming regulator or supervisor they know exactly what you are talking about.

lol donkaments

nice downvote. Lots of donks here DIAGF.

Registration is open for Startup School 2019. Classes start July 22nd.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact