Lying about your hand

"Did you know that it's OK to lie about your hand, but you're not allowed to tell the truth about it?"

I heard that from a woman at the table yesterday at the Flamingo. It's not the first time that I've heard somebody make such an assertion. Most famously, Annie Duke said essentially the same thing during one of the 2006 World Series of Poker episodes. I suspect that the 700 million re-broadcasts of the series are ultimately the source by which most players "know" this rule.

But it's not true. The fact that so many people hear this and instantly believe it is just one tiny manifestation of the general societal problem that most people have precious little training or experience in critical thinking.* Even if I weren't pretty well attuned to the subject of poker rules, just an overall skepticism about accepting as fact things that people say would have caused me to raise a quizzical eyebrow at Ms. Duke's claim. I would immediately think, "Can that really be true? If so, it doesn't make any sense at all."

Let me address this under three headings: (1) Why the "rule" as stated would be stupid and pointless, (2) What the rule actually is, and (3) What I think the rule should be.

The alleged rule would be stupid and pointless

Suppose for a moment that there really were a rule that said that you could say any lie you wanted to about your hand, but couldn't tell the truth about what you were holding. Two prominent poker columnists have recently addressed this in detail, apparently both believing that there actually is just such a rule. They both do a nice job of explaining why it's ridiculous.

First is Steve Zolotow, writing for Card Player magazine (http://www.cardplayer.com/author/article/all/57/10074):

Someone in the main event told the truth about his hand, saying, "I have a king." His opponent folded, but he received a one-round penalty for telling the truth. Obviously, he should have said, "I don't have a king," then paused and added, "but there is a penalty for telling the truth." Thus, he could have conveyed the same information by lying about his hand. As long as we know that our opponent must be lying, we can assume the opposite is the truth. Unless players are allowed to use some mix of truth and lies, they effectively are telling the truth.

Next up is Mike Caro, who calls this "the idiot rule," because it prevents people from disclosing whatever they want about their cards only if both players are idiots (http://www.pokerplayernewspaper.com/viewarticle.php?id=2351):

But the worst rule in poker states that you can't tell the truth about your hand. For instance, you can't say, "I have three aces," if you actually do. That sucks. It sucks because, by rule, anything I say must be a lie to be legal. So, if I say I have a full house, opponents are given information. They know I can't have a full house or I'd be in violation of the idiot rule. So, I guess, if I wanted to tell someone I had aces full, I'd have to say something like, "One thing I can say positively is that I don't have aces full." If I said that and didn't have aces full, I'd be telling the truth and in clear violation of the idiot rule. For that reason, astute opponents would know that I'm lying--which is legal--and that I absolutely do have aces-full.

Exactly so. A rule requiring that everything you say about your hand must be a lie would be (1) unenforceable (because not every hand gets opened to check to see if the player was telling the truth or not) and, (2) completely ineffective. In fact, paradoxically, the more rigorously it was enforced, the less effective it would become at the presumed goal of preventing collusion.

That ain't really the rule

Fortunately, nobody has yet, to the best of my knowledge, actually been stupid enough to implement the "rule" that Zolotow and Caro rail against. I'm really quite surprised that both of these columnists took the trouble to publish a critique of a rule that doesn't exist.

So where does this poker urban legend come from? As far as I can tell, the source is this language in the World Series of Poker rules (http://www.worldseriesofpoker.com/pdfs/wsop.tournament-rules.07.pdf):

52. Players are obligated to protect the other players in the tournament at all times. Therefore, whether in a hand or not, players may not a.) disclose contents of live or folded hands....

Notice that it doesn't say anything at all about "truth" or "lies." It forbids one to "disclose" the contents of one's hand. The number of ways that one can "disclose" one's holdings are limited only by the imagination. You could show your cards. You could just say what you have. You could tap it out with your feet in morse code. You could send smoke signals (well, except that they've banned smoking at the table now). You could let an opponent keep guessing, shaking your head "no" when he's wrong and nodding "yes" when he's right until he's got it, thus never speaking a word. Maybe if you have a touch of Marcel Marceau's spirit in you, you could pantomime it. It doesn't matter--they're all equally prohibited.

The implication is that you also can't "disclose" your hand by invoking a logical game such as Zolotow and Caro propose: "I do not have XYZ, but you know that I have to be lying about that." As they note, that is the functional equivalent of straightforwardly stating exactly what you have. Both are prohibited under the actual wording of the rule in question, because both are just two of the myriad methods by which one might "disclose" one's hand.

Let me repeat: There is no currently published rule anywhere that says or implies that you can lie about your hand but not tell the truth about it.

Let's look at some other common sources for poker rules, for the sake of completeness.

The Tournament Directors Association rule 12 is nearly identical to, and the probable source of, the WSOP rule:

Players are obligated to protect the other players in the tournament at
all times. Therefore, players, whether in the hand or not, may not: 1. Disclose contents of live or folded hands....

Next is Bob Ciaffone's "Robert's Rules of Poker" (http://www.lasvegasvegas.com/poker/chapter1.php):

The following actions are improper, and grounds for warning, suspending, or barring a violator:

...Revealing the contents of a live hand in a multihanded pot before the betting is complete. Do not divulge the contents of a hand during a deal even to someone not in the pot, so you do not leave any possibility of the information being transmitted to an active player.

Cooke's Rules of Real Poker--for my money the best single rulebook yet--says this (p. 71, rule 10.20):

A player shall not intentionally expose a card or reveal it verbally in order to induce or inhibit action. However, in head's up matches (where only two players are dealt in) revealing a card shall be permitted.

I'm not sure if this is intended to cover literally only games where there are only two people playing the whole time, or if it is intended to extend to full ring games, when only two players are contesting the pot.

Lou Krieger and Sheree Bykofsky, in their book The Rules of Poker: Essentials for Every Game, go even further, and would disallow such disclosures even when heads-up (p. 79, rule 3.9):

Intentionally exposing a card to induce or inhibit action is a serious breach of poker ethics, and in certain instances can be considered a form of collusion. It is never permitted.

Notice that not even one of these sources says even a single word about lying versus telling the truth. So the next time you hear somebody make such a claim, ask him or her to point you to the actual source where the alleged rule is written. It's a safe bet that the speaker will be unable to do so.

What should the rule be?

Let's talk about cash games first.

The overarching reason for any no-disclosure rule is to protect other players. For example, perhaps a player has moved all-in ahead of me before the flop and I have pocket kings. I move all-in, too, and show my cards before other people have had a chance to act, because I would really prefer to have to beat only the guy who is already all-in; large pocket pairs don't play well against, say, five or six opponents, so my action is intended to scare off opponents. But the person who first moved all-in may have something like suited 9-10, which plays much better, in terms of risk:reward ratio, when there are several players in the hand (because the payoff for hitting a straight or flush is potentially enormous). Obviously, the same reasoning applies to saying "I have two kings, so the rest of you should fold," even though the verbal disclosure leaves more doubt about the speaker's truthfulness than does showing the cards.

What about less specific disclosures, such as "I have a really big hand here"? I would outlaw them, too, on the slippery-slope argument. If you write the rule so that it only prohibits stating exactly what your cards are, then presumably it would be acceptable to say something like "I have a pocket pair bigger than queens." That comes so close to what is prohibited that the rule ceases to be useful. So I would move the dividing line all the way down to zero: you can't say anything that suggests--truthfully or not--the strength of your hand. I would also prohibit statements that come in the form of "I might have...," which is a common ploy of some players.

When there are only two players contesting the pot, however, a player disclosing his hand cannot do any harm to anybody except himself, and since I don't believe in protecting people from their own mistakes and/or stupidity, I'd allow it. There are times when revealing one or more cards can work to one's advantage in confusing or deceiving an opponent. Consider this absolutely brilliant ploy by Daniel Negreanu, where he combines showing one card, a statement about what the other one might be, and a scary-looking minimum raise, to get Sam Farha to fold a better hand: http://www.youtube.com/watch?v=d33K96MGzVA. It's an ingenius trick, one of the slyest bits of deception I've ever seen deployed at a poker table.

I was on the bad end of another skillful bit of deceptive revealing by an opponent a couple of months ago. I started with a strong hand (can't remember exactly what it was), but hated seeing three hearts on the flop, when I had none. The turn brought a fourth heart to the board. I bet, my tricky opponent took a long time to decide what to do. While thinking, he turned over the 7 of hearts. He finally called. The river was a blank. I decided that with just a 7-high flush, he must be worried that I had a higher flush, so I moved all-in. He insta-called with the look on his face of the cat that caught the canary (as, indeed, he had). His other card was the king of hearts. He had flopped the king-high flush, and his showing the lower card tricked me into thinking exactly what he wanted me to think: that he just had a low flush and was in a difficult spot, when really he had the second nuts. Well played, sir--you lured me in perfectly.

That kind of clever psychological gambit should absolutely be part of the game, as long as only two players are involved. So showing one or both hole cards, or announcing or hinting at your hand--honestly or dishonestly--is fine by me.

What about in tournaments? The same considerations hold for multi-way pots, but there's an added wrinkle in tournaments that makes the picture less clear for heads-up situations. That relates to the concept of "tournament equity." This isn't the place for detailed analysis, but basically, depending on the payout schedule, late in a tournament, how one player acts in a hand can dramatically affect the amount of money that other players--even those not involved in the current hand--can expect to win. This is at the heart of an interesting debate between Lee Jones and Daniel Negreanu, at http://www.biggerdeal.com/2007/10/05/daniel%e2%80%99s-wrong-i%e2%80%99m-10000-sure/. Be sure to read the comments, too. (One of the commenters is me. I'll let people guess which one.)

I don't know the optimal way to balance the advantages of allowing the kind of Negreanu-esque ploy, which has traditionally been an integral part of the game, against preventing injury to uninvolved players' tournament equity. Perhaps starting the ban on disclosing one's hand, even when only two people are in the pot, when the money is reached, or a few places before the money, or at the final table, would strike the best balance. Of course, even if it works that way, when it gets down to the final two contestants, the rule serves no purpose and could be taken out of effect again.

Back to the start

But all of that is mostly irrelevant to where I started this discussion. I began with griping about how so many players just hear somebody say what a poker rule is, and accept that as the gospel truth, without looking it up anywhere, without thinking very hard about whether the alleged rule makes any sense or not, etc. To compound the problem, these people then repeat it to others, spreading the disinformation. It makes me crazy. It's not that hard to find out what the rules really are.

So when you hear somebody claim that there's a poker rule that says ________, and you've never heard it before, ask for evidence. Hold the thought in suspension until you can verify it. Don't pass it on as it it's the revealed word of God.

And for Pete's sake, don't talk about your hand!


*I risk going off on a full-tilt tangent here, because this is one of my biggest all-time sources of frustration and dismay about the world. In fact, if I had the magical power to cure just one ill of humankind, I think it would be the lack of critical thinking. I would make every person on the planet instantly inclined to critical thinking, and skilled at it. This would immediately solve countless woes, from the relatively trivial (people wasting their money on pills that the sellers promise will cause miraculous weight loss, enlarge male genitalia, or make the user irresitible to the opposite sex) to the truly world-changing (people being led to believe that if they fly a commercial jet into a skyscraper they will be rewarded with a bevy of virgins in heaven). So don't get me started.


Addendum, December 7, 2007

I emailed Mr. Caro to tell him of my disagreement with the premise of his column (Mr. Zolotow doesn't have a published email address). He responded quickly, for which I am grateful. With his consent, I'll share his comment:

About the issue on “can’t tell the truth.” I know that rule is in existence. It only applies to tournaments, though. Many directors have since modified it. Hopefully, my urging had something to do with it. I was, I think, instrumental in arguing against it on several televised events I participated in. They even waived the “can’t show cards” heads-up rule in both the National Heads-up Championship (NBC) and Superstars, when I participated. That wasn’t my influence alone. Many others felt it made for good TV to adopt these changes.

I even had the honor of roasting the person who first put the rule into place at the WSOP.

This is useful information. I can't find any such rule in current, published sources, but it's certainly possible that, as suggested in Mr. Caro's reply, there are tournament directors who use such an unpublished rule specific to their facility, and/or that if I looked at older versions of tournament rules I would find it. I freely admit that I didn't investigate older materials, so if a must-lie rule previouly existed, I wouldn't know about it (having been in the poker world for only a couple of years now).

Phil Hellmuth is bad at math

I thought I couldn't dislike Phil Hellmuth more than I already did. Turns out I was wrong about that. In fact, I'm thinking now that my capacity for disliking him may actually be infinite. Still, I didn't know, until last night, that I would end up including "poor math skills" on my list of reasons for disliking him.

On this week's installment of GSN's "High Stakes Poker" an interesting situation came up. The hand itself was pretty unremarkable, as you can see from the cards shown in the accompanying screenshot. What was interesting was whiny Phil Hellmuth's negotiations for "insurance" on the hand.

In this game, $67,000 pots are bigger than average, but not all that unusual. In most episodes there are one or two pots that climb into the $100,000-$200,000 territory. So to start off with, it's strange for a player to be so afraid of losing that he'd seek insurance on a pot of this size. Nobody else at the table would even think of it--only Phil, who, for some reason, just can't stand losing a biggish pot. This is particularly strange, since he is up by a couple hundred grand for this session, as he has already bragged.

But OK, it's his right in this game to try to reduce the chance that Eli Elezra will catch a lucky card and steal this pot from him. They could agree to run the last card twice, for example, and it would be highly unlikely that Eli would luck out twice in a row, so Phil would be almost guaranteed at least half of the pot. Another possibility is one that Eli explicitly offered: give Eli a quarter of the pot and take three-quarters for himself. I've seen Phil make this kind of arrangement when he's in a dominant position and all-in against Barry Greenstein in a previous season; if I recall correctly, Phil was something like a 90% favorite, and Greenstein agreed to give him 80% of the pot, and they didn't even play out the rest of the cards. In Greenstein's more rational view, he'll take the expected value that Phil is giving up, and not worry about the variance.

For some reason, though, Phil this time wants to go the "insurance" route. That is, he wants to pay somebody a small amount if he wins the hand and, in exchange, be paid a larger amount in the event that Eli draws out on him. This is where it starts to get strange.

I have the episode on my hard drive, so I can play it back and do a verbatim transcript. Phil says, "Queen is a chop. All right, I'm just going to insure against a nine or a jack."

Here's Phil's first mistake. It's true that a queen on the river chops the pot, because both players would have two pairs, kings and queens with a 10 kicker. But Phil fails to notice that a 10 also produces a split pot, because it gives both players tens and kings with a queen kicker.

Phil continues: "So it's eight outs.... 36 to eight is the actual odds. I'm a 4 1/2 to 1 favorite."

That's his second and third mistakes. Eli actually has only seven outs to win the pot outright: the four remaining jacks to make a straight, and the three remaining 9s to make a better two pairs. Additionally, Phil is strangely counting the cards that make for a chopped pot as being on his side. (Sam Farha folded a 9 and a 10, which would affect the calculation if Phil knew about them, but he doesn't, so we'll consider them as unknowns, treating them the same as the ones still in the dealer's hand.)

The actual odds, not counting the queens and 10s, are 31 cards for Phil and seven cards for Eli, which is 4.4:1 (which, by coincidence, happens to be close to what Phil said, though he was wrong in how he got there). If you account for the pot-splitting cards by adding them to both players' sides of that ratio, it's 37:13, or only 2.8:1.

Phil asks for insurance at 4:1. "I'm 4 1/2 to 1. I'll take 4 to 1. I'm letting you have above 10%. You're getting 12 1/2% juice. That's pretty good."

Brandon Adams takes him up on it for $40,000 to $10,000. That is, Hellmuth is to pay Adams $10,000 if he wins the hand, and collect $40,000 from Adams if Elezra sucks out on him. Then Farha joins the deal for an additional $8000 to $2000.

The dealer puts out the river card: a queen, and the pot is split.

Now Hellmuth says, "It's a split pot, nobody has to pay any insurance."

Farha objects, but then both Jamie Gold and Daniel Negreanu remind him that Phil specified that he was only insuring against a jack or a 9, and Farha, for some reason, accepts that.

But that's what is most strange here. You would expect Adams and Farha to collect their insurance payment, because the event against which they were providing insurance did not occur. Phil said he wanted insurance against a jack or a 9, and none came--so he owes his "premium" to his insurers. What's more, he was unambiguously including the pot-chopping cards as being for him in his proposal, when he claimed that he was a 36:8 favorite. He never said that a chopped pot would mean no money going either way as part of his offer.

So one of two things happened: The insurance deal was made with the understanding that a split pot meant Phil owed insurance, and he reneged on his end of it, or he misrepresented the terms of the offer. By coincidence, he happened to state the odds approximately correctly for the situation in which a chopped pot meant no payment either way, but his explanation clearly included the implication that he would pay up in the case of a chopped pot. It's a very strange situation, and I wonder whether debate among the participants continued after the taping stopped, and whether any money ended up changing hands. (This was the last hand of the night.)

However they worked it out amongst themselves, we're left with these conclusions: Phil Hellmuth doesn't know how to count an opponent's outs. He's a high-stakes dweeb who, uniquely among these players, can't tolerate the swings that inevitably come with playing for these amounts. Finally, he either misrepresented the offer he was presenting to Adams and Farha or welched on it after the hand was over.

If you didn't have enough reasons to dislike Hellmuth before, there's at least three more for you.

Addendum, October 30, 2007:

I've been discussing the math of this with a bunch of other people at the http://www.twoplustwo.com/ forums, and one of them just posted a link to this entry from Hellmuth's blog, apparently made soon after the taping of the show in question: http://www.philhellmuth.com/phil-hellmuth-poker-blog.html?id=1866 In this post, he says, "Later on, I figured out that I owed $12,000 to Brandon ($10,000) and Sammy ($2,000), because they insured the very last hand of the night. By the way, Brandon doesn't even know that I owe him $10,000, because I decided that I owe him after the fact! He will be pleasantly surprised, that's for sure."

Glad to know that he finally did the right thing. Still hard to know why they didn't get it sorted out on the spot.


Addendum, November 4, 2007

Brandon Adams just posted a note about this situation to the twoplustwo forums:

Yo, I should have commented on the Hellmuth insurance thing earlier. First off,
Phil is one of my favorite guys in poker. He's not cheating anyone. I'm going
off memory here....The deal was only partially clear in my head at the time of
the insurance. it was obviously a good deal for me. when we chopped, i didn't
know what we were supposed to do, though since insurance naturally is protection
against something bad happening (in this case, a bad beat) and something bad
didn't happen, my first thought was that a chop went to me. but then i started
thinking about it and i realized that the deal was good for me even if a chop
was a chop. so i was confused and decided to let it go. this is just what i
remember thinking... as of now, i have no recollection of the cards and i
haven't read the above. The next day phil runs into mori at the coffee shop and
tells mori that he should have paid me. Phil emailed me to tell me he owed me
10k. I thought about it some more and told him not to pay me. I said this b/c if
i had won with a chop, then that implied some ridiculous vig for me (i still had
a decent edge counting a chop as a chop). He insisted on paying me, I told him
we'd flip a coin for it, then he said he'd pay me $6k. he paid me $6k the first
time i saw him at the series. ba

This is pretty reasonable of Adams. Here's the math on what he's saying: Suppose the deal had been explicitly made that no insurance money went either way in the event of a chopped pot. There are 31 cards that win the whole pot for Phil, 7 for Eli, so the odds against an insurer having to pay out are 31:7, or 4.4:1. That's what Adams is calling the "decent edge counting a chop as a chop." On the other hand, if a chopped pot meant that Phil still had to pay his insurance "premium," then there are 37 cards on which he has to pay it (the 31 that are an outright win, plus the 6 cards that result in a split pot), and only 7 on which he collects insurance. 37:7 is 5.3:1, which is what Adams is here calling "some ridiculous vig for me" (32%, actually, compared to the 11% he decided was "decent").

In any event, it sounds like they ended up settling it like gentlemen.