How does physics impact poker?

April 14, 2008 9:44 PM

Does the theory of the multiverse in contemporary physics mean that the top card in a deck of cards isn't "set" until it is seen? (ideal answerers have a background in physics, but all answers appreciated.)

After a weekend in Atlantic City (I came out even, after expenses, so no harm done) I'm curious about whether the multiverse has philosophical implications in poker. Here's a situation. If I have pocket fives and (for various reasons too long to get into) I fold to a raise, and a five comes up on the flop: is it necessary that the same five would have come up if I had stayed in the hand? In other words, is the state of the flop "set" after the cards are shuffled? Or is it possible that had I stayed in the hand, the five would not have come up? This may seem a bit out there, but I feel that my future in cards hinges on the lessons of physics.

posted by jimmydare to science & nature (27 comments total)

No. It implies that there could be a universe where the card was a five, but here in this one, that card was there the whole time.
posted by TheOnlyCoolTim at 9:47 PM on April 14

Yeah - that's a bit out there. I've fallen victim to that "logic" before, but I gotta say (reluctantly) that the card on top is the card on top, no matter what you're holding. Any other expectation is what they call "magical thinking" and trust me, you don't want to go there.

Sorry....
posted by The Light Fantastic at 10:11 PM on April 14

Not exactly what you're looking for, but somewhat related is the Many-worlds interpretation (of reality, I assume).

Basically, all possibilities (given the laws of reality) exist in otherwise similar parallel universes. So in one universe everything is the same, except the five did come up. Another a seven came up. In another, you didn't go to Atlantic City at all, ad infinitum.

Starting from there might lead you somewhere closer.
posted by Nelsormensch at 10:31 PM on April 14

As I understand QM and the Everett many-worlds interpretation, if the state of the deck is a superposition of two states (one with the five on top and one without), then the universe will continue to be in a superposition of states— one in which the top card was a five and one in which it wasn't— and those states will continue to evolve independently (so later the universe will be in a superposition of a state where you won the hand with a state where you lost, and so on).

The more traditional Copenhagen interpretation would have it that the deck is in a superposition of states, but that at some point— say, when it's observed by a human, because we all know humans are extra-special in the cosmos— the deck's state will collapse into a randomly-chosen one of its superposed states, and the universe will continue along that path, and that path only.

In neither case does your deciding to stay in the hand theoretically affect the deck's state.

The interesting thing is that these two scenarios are observationally indistinguishable. That's why they're called "interpretations"; they don't affect the nuts-and-bolts of the theory at all. If you build an instrument to detect whether Schrödinger's cat is alive, you can't tell the difference between it deciding "CAT ACTUALLY ALIVE"/"CAT ACTUALLY DEAD", or it entering a superposition of CAT ALIVE/DEAD and you, in turn, entering a superposition of having to explain things to the SPCA or not.

Disclaimer: I am not a quantum mechanic.
posted by hattifattener at 10:50 PM on April 14

Actually, I don't think you can use those quantum mechanic interpretations in a purely mechanical situation. Schrodinger's cat paradox and its ilk are always dependent on a particle measurement where the uncertainty principle holds. With super-attenuated vision, or a particularly tactile dealer, the order of the cards can be known by a human before the cards are turned over. The uncertaintly principle applies to areas where the state cannot be known before a measurement.
posted by FuManchu at 11:07 PM on April 14 [2 favorites]

No "theory" of the "multiverse" has anything to do with real life. We can take the idea seriously for a moment and get an idea why. First guess at the number of alternate universes where an alternate you with alternate pocket fives doesn't fold and gets the three-of-a-kind. Let's call that number N, and assume it's very large but finite. Now, by the same process, estimate the number of universes where you don't go to Atlantic City at all, where you were never born, where the laws of physics are inhospitable to life, and so on. Call that number M. Which is larger, N or M? Which is more likely, that the deck was dealt differently, or that you don't exist at all?

There was a recent NYT science article about a calculation reporting on some dialogue on this subject in the physics literature: a calculation that, in a multiverse, disembodied brains with solipsist delusions are thermodynamically much more likely than stable, "real" worlds.

I can't stop you from imagining that the cards would have dealt differently if you'd played differently. But I can warn you that the idea comes with some surprising and unpleasant baggage.
posted by fantabulous timewaster at 11:28 PM on April 14 [1 favorite]

Anything that has yet-unobserved outcomes cannot be said to have a definite outcome until that outcome is observed. The top card in a freshly-shuffled deck COULD be ANY card in the pack, and IS one of them. You just can't say WHICH one until you turn it over and look. Until then, it exists as a cloud or field of the various possible outcomes - it is every card and none of them at once. Once you turn it over and observe which card it is, that field of possibilities collapses to become one single outcome. It has always been a card, and it has always been that card, but the collapse of probablilities is the divide between the possible cards it COULD be and the card it is obeserved to be.
You don't have any influence on which card it IS, other than your observation of it as being one card or another - directing your personal quantum vibrational influence on the outcome, also known as "wishing" wouldn't have made it the five-spot.

This kind of thinking is what happens when exotic math/physics gets projected upward or downward to concrete things in the popular mind. (a cartoon electron's orbit around a cartoon atom exists as the sphere of all possible locations for the electron - it's "moving so fast that it's a blur" - you can only say where it is, or even that its existence can be treated as a discrete particle rather than a field in your calculations, until you take a "flash photograph" of it, and locate it in one point in space at one point in time and say Aha! it is there rather than here or somewhere else!. But you also can't say where it's going to be next - you can only take another photograph at another moment and say that it was there, then - in between it goes back to being an indefinite blur of its possibile states.

The 'Many Worlds" hypothesis comes from an even more exotic principle, trying to do something like the conservation of mass and energy, but with probabilities: if that top card existed at one time as "all possible cards", then when it collapses into one single card, those probabilities have to "go somewhere", they can't just cease to exist. Thus the idea that every possibility DOES concretely exist, but each outcome is a different branch of reality - when an outcome is determined, the observer and the outcome can be said to have followed (or existed all along) in one branch over the other. There are 52 universes, yours just happens to be the one where it's a three of clubs as opposed to the five of diamonds.
Which inspires lots of interesting fictions - ("I found a dime that had Wilson on it instead of Roosevelt- proof of an alternate universe!") - but again, your own influence on which universe you're in or branch your're following is nonexistent. It's fun for reading, but thinking you can make it happen will give you ulcers and empty your wallet.
posted by bartleby at 11:36 PM on April 14

Objects the size of cards behave classically, i.e. they behave like everyday objects obeying the laws of classical mechanics (Newtonian physics, basically). We don't observe quantum effects until we look at scales much smaller than the size of a card. (To be technically correct here, I'll say that objects the size of cards behave very, very, very nearly classically). We can build bridges, drive cars, play cards, and so on, because all these objects behave in a classical manner. They are subject to cause and effect and they behave deterministically. A card is either on top of a stack of cards, or it isn't.

It turns out that when we do experiments on very small scales, weird things start to happen. On this very small scale, things behave according to the rules of quantum mechanics, which are probabilistic. An electron may or may not be in a certain place; as far as we can determine it is neither there or not there until we try to measure its position (this is a gross simplification). In short, things don't behave how we expect them to.

Why is this? What is the nature of energy, matter, space or what have you such that these strange (to us) laws can hold? We don't know. Different physicists have different ideas about the answers to this question. These different ideas are called interpretations. The many worlds interpretation, which I assume you are referring to, is but one of these interpretations. No one here can tell you if it is true or even if it is possible to determine if it is true.

I'd suggest that for your purposes, you stick with the classical mechanics that have worked so well for everything card sized for many years and assume that the order of the cards is set once the shuffle is complete. If these sorts of issues interest you, go ahead and learn more about them, but you'll find it will be much easier if you ignore questions about cards and deal with the questions on their own terms.
posted by ssg at 11:49 PM on April 14

Even if there was a quantum-event card shuffler back there somewhere, if the top of the card has collapsed (and it has, you can see it) then surely anything connected to it in the classical sense has also collapsed? Card backs don't swap faces on the sly.
posted by Leon at 1:22 AM on April 15

Many in the thread are waving off your question by saying that the card is too classical (big) to be subject to QM but that's more or less a cop out. If QM applies to small scales and we can tie the state of something macroscopic to something small, then it applies to big scales too. While it is hard to do experiments that would verify your card is in a superposition of states prior to you looking at it (e.g., how does one look at interference patterns of playing cards?), wavefunction collapse (the card picking a state when you look at it) is not something that is well understand physically or philosophically. Also you really mean the Many-worlds interpretation of QM (one particular multiverse theory). Other multiverse theories don't have much to do with the card stuff.
posted by njgo at 2:56 AM on April 15

The multiverse theory (or the many-worlds theory, or any non-collapse theory of QM) says that from your perspective the top card of the deck is not determined until you see it.

The main point is that in these theories there is not one deck of cards, one casino, one Las Vegas, and one you, but many many versions, all slightly different. And over all these worlds, every possible permutation of the deck of cards exists many times over. In out of every nine worlds there'll be a 5 on the flop, and a corresponding you to win or lose.

However, your behavior has no effect on the order of the unseen cards, or the likelihood that you will see a 5 or not.

MWI says some remarkable things to say about the nature of reality, but it does not predict you can alter probabilities with your actions. So keep folding your fives against the overpairs unless you're getting the implied odds to call.
posted by justkevin at 6:36 AM on April 15

Thanks for the considered responses everyone!

I should mention that my curiosity about this subject is not borne out of any desire to CHANGE the state of the card. That's just plain crazy.

But I'm curious if there's any consolation in science to the poker adage that a good decision is the one that is based on all the available information at the table (spiced with your desire to gamble) and that any information that emerges after that decision (i.e. the five comes up) doesn't matter. If I could find any reason in physics to believe that there was a chance that the five would not have come up if I'd stayed in the hand it helps salve the regret of making a good decision - and that philosophy (as far as I can tell) would have no negative impact on my poker play. (although poker players, feel free to disagree)
posted by jimmydare at 7:52 AM on April 15

Assume you are back in that same situation, and its now time for the flop. But instead of turning over the flop, the dealer instead shuffles the deck, and then prepares the flop.
Do you fold or not?
posted by forforf at 10:52 AM on April 15

I have to disagree with njgo.

The short answer, that the card is too big and thus acts classically, is the right one, and I don't think it's a cop out at all. Yes, wavefunction collapse isn't well understood. However, "difficult to do an experiment to see the quantum properties" is misleading; it's bloody impossible. It's thus a question for philosophy, not physics, whether the card is in any small way ever in a superposition of states. (I think it's very silly to think it is, but that's just my opinion).

(By bloody impossible here, I mean just as hard as walking into a room and finding all its air molecules have decided to congregate in one tiny corner. It's just not going to happen.)

Which brings us back to the point for jimmydare: of course it's quite possible that the five would not have come up- if the dealer had shuffled differently, if the cut had come differently, etc etc. You don't need quantum mechanics for that. The point is that's information you don't have; you can't run the classical mechanics that more or less determines which card comes up next, all you know is there are say 45 possible cards that could have come up, all with roughly equal probability.

What you're doing if you're playing poker right over the long run (right=earning money) is statistics. If you made the exact same "right" play with the same information (same cards showing and in your hand) 10,000 times, the five will come up such a small amount the time that overall, you'd win money.

Best way to salve the regret: know that your decision was right, play enough that the statistics work out eventually for you making the same decision (or same kind of decision), and feel warm and cozy with your extra dollars (or other currency of choice).
posted by nat at 10:57 AM on April 15

Seems that if your guess/theory were correct, jimmydare, Chrith Angel'th job would be a hell of a lot more difficult. Sorry I don't have a real answer, but seriously, "magic" would be a different kettle of fish all together if things were that random. Amiwrong?
posted by heyho at 11:16 AM on April 15

I'm sort of playing devel's advocate here nat, but there are very deep issues here that are easy to gloss over (and are in most elementary quantum texts, classes, etc...) My point was really that wavefunction collapse "determines" reality in QM in a very serious way. Without it, reality is just a sort of abstract information evolving by Shrodinger's equation. What isn't understood is who or what is capable of collapsing wavefunctions, and what collapse means (beyond just positing some non-unitary operator that throws the system into an eigenstate, a highly unsatisfying theory). And so the question is "who is doing the measuring" with the cards? At some point in there past, a shuffler may have moved their hand in a way subtly dependent on some quantum state in their brain, thus yielding multiple possible eigenstates for the deck). So who picks the state? The brain? The deck? The observer of the deck? The line between not knowing something (and being unable to know, to be precise) and there being no answer is hazy in QM.
posted by njgo at 11:37 AM on April 15

While it is hard to do experiments that would verify your card is in a superposition of states prior to you looking at it (e.g., how does one look at interference patterns of playing cards?), wavefunction collapse (the card picking a state when you look at it) is not something that is well understand physically or philosophically.

The physical behaviour of cards is very well understood and that behaviour has nothing at all to do with superposition of states or wavefunction collapse. Wavefunctions are useful ways to mathematically represent the behaviour of electrons, etc. but they are of no help in representing the behaviour of cards.
posted by ssg at 12:02 PM on April 15

njgo: your phrasing of the question is what I'm trying to get at, and much more clearly cast than I'm making it.

Yes, the cards have predictable behavior - but since it's all just information, does my folding (and in doing so, adding additional information to the system) have any impact on that original shuffle? Totally mad, I completely agree, however imagining that this is the case should make for better poker play, not worse, and fewer regrets when a bad beat comes down. (which is more than one can say for most crazy beliefs)

And again, I am not suggesting that my behavior has any positive or negative impact on the cards in terms of the ODDs; just that the same hands don't necessarily come out every time in the timestream.

forforf: your alternate scenario helps illuminate the problem nicely. As far as good poker play goes, nothing changes no matter how long the dealer shuffles the cards, or when in the pre-flop process they shuffle the cards. The only thing that matters are the odds I make with the information I have. What I wonder is whether my actions might drive the results. Obviously, this is totally counterintuitive, but since it doesn't hurt the gameplay, it's a useful cast of mind.

Oddly, poker players are incredibly prickly about what happens with the deck even if it doesn't change anything. They seize on any error on the shuffle, the order the cards are dealt face down to the players, when the shuffle happens, burn cards, etc, which should have no impact at all on the information of the game (there's no difference to the odds if the dealer shuffles continuously thought the game, unless someone has seen the cards)

Also - I'm perfectly willing to concede that I'm applying the ideal of the multiverse (or alternate worlds, if that's what my pidgin physics is really referring to ) incorrectly in this situation if that's the consensus.

But I would argue (and I hope you wouldn't think ill of me, or peg me as some sot of dement ed I.D.er) that science has not told us enough about time and ... well ... fate, to indicate that it's always true... since it isn't actually verifiable.

But I'm more interested to know if how muddled my understanding of the physics i - and it sounds like it is pretty muddy unless njgo has a background in it... in which case - woot - one in x physicists agree with me which would be enough for me to adopt this as faith.

Of course, then I start getting into vexing problems of free will - which I console myself by watching the mario multiverse video on loop.

Seriously, thanks all for the considered responses.
posted by jimmydare at 2:48 PM on April 15

njgo, I guess my point was that wavefunction collapse has most definitely occured by the time jimmydare is deciding whether or not to fold.

Wavefunction collapse, if it is the proper view, has to occur somewhere in mesoscopic physics; that is, between the scale of atoms and that of cards. You can't really prepare an entangled state of as many atoms as are in a card; another way of saying this is that macroscopic objects can't really be put in a superposition of states. Whether this is for practical or theoretical reasons doesn't really matter; you just can't do it, and it won't be happening in jimmydare's poker hands any time soon (read: ever). Cards are like mixed states; they are not in superpositions.

(for the distinction between "mixed state" and "superposition", see sections 2.4 and 2.5 in this Wikipedia article.)

In short, I concur very strongly with ssg. Cards are classical, and to pretend quantum physics gives us insight into how they act is disingenuous.
posted by nat at 6:04 PM on April 15

I'm not pretending anything of the sort, and do I understand the difference between density matrices and superposition. The dynamics of the cards are well approximated by classical laws, obviously. I said in my first post that one could not practically construct an experiment to determine whether or not the cards were in a superposition of states in my first post. But I'm sorry, you're still coping out when you say "somewhere in mesoscopic physics". That's just not an answer.
But from a philosophical point of view, either QM is right or its wrong, it doesn't simply cease to be true at the micron scale.
For a phenomenal discussion of some of these issues, Sidney Coleman gave a lecture many years back that's finally made it only google video:
http://video.google.com/videoplay?docid=-4674461198051839963
I can't recommend it strongly enough.
posted by njgo at 6:25 PM on April 15

"Whether this is for practical or theoretical reasons doesn't really matter; you just can't do it, and it won't be happening in jimmydare's poker hands any time soon..."

Nat and ssg are, unfortunately, wrong. It is difficult, but not impossible, for macrocosmic objects to be in superposed states. This theoretical possibility is the entire point of the Schrödinger's Cat, gedankenexperiment.

Well, in fact, the point was to illustrate the problems with QM by showing that the supposed tidy division between the micro and macro with regard to the weirdness of QM was not so tidy; and thus to suggest that there might be a problem with QM if we don't want to accept macrocosmic quantum weirdness.

The Wikipedia page on Schrödinger's Cat has a nice discussion on this, as well as a pretty good quote from Weinberg.

Over the years, there've been a number of other thought-experiments discussed which move superposition into the macrocosmic realm. The bottom line on this is that A) like njgo says, in the most rigorous, theoretical sense there's no basis for eliminating superposition from the macrocosmic realm; while, B) nevertheless, in the world as we typically experience it, wavefunction collapse occurs at the microscopic scale.

On a more general note, it is very difficult to find reliable information on these deeper questions implicit in QM and the related philosophical issues. There's an enormous amount of crap and downright wrong pop-science that has a lot to say about the philosophical implications of QM. Almost all of it can't be trusted. My recommendation for a long time has been to only read books on QM by actual particle physicists.

But the problem with that is that these implications and issues are pretty much completely elided in contemporary particle physics. After the intense interest by and discussion among physicists in the early days of QM, it's as if physics collectively shrugged its shoulders and decided that it's a bunch of hand-waving. I once corresponded with a theoretical particle physicist who spends some of his time on "the philosophical implications of QM" and I asked him if he gets some negative feedback from colleagues because of it. He said, yes, he does, a little, but that he was tenured and senior in his department and so he gets away with it.
posted by Dances with Werewolves at 1:22 AM on April 16

Well, gedankenexperiments are not the same thing as actual experiments.

I say call me when you get the cards in a pure quantum state, and then we can have a discussion. Yes, saying "things happen in mesoscopic physics" isn't the full answer; I'm not satisfied with it, but I'll leave the question of where and how exactly to those who actually study that realm (I don't).

The point is there's no reason to believe quantum mechanics gives a useful description when discussing macroscopic objects; I suppose there's no reason to believe it wouldn't, if we could prepare a pure state, but for the most part you can't make such a state macroscopically (and I certainly would be extremely suprised if you ever did it with cards).

(As a string theorist, I feel a little pot calling kettle black here, but I think it's important to keep in mind what is physics and what is speculation inspired by physics. Just don't ask me where that line falls in my own field.)

I actually did run into a bunch of physicists at the Perimeter Institute last year who do study "the philosophical implications of QM". I think what they do is interesting, even if I don't agree; some of you might be interested too: link.
posted by nat at 10:58 AM on April 16

Over the years, there've been a number of other thought-experiments discussed which move superposition into the macrocosmic realm.

Thought experiments are not at all equivalent to physical experiments. You can posit as many imaginary cats in imaginary boxes as you like, but you can't prove anything at all.

It is difficult, but not impossible, for macrocosmic objects to be in superposed states.

You should say: "According to the interpretation that I subscribe to, which has, of course, no experimental basis, it is difficult, but not impossible..."
posted by ssg at 11:10 AM on April 16

I disagree, Schrodinger's cat is not a matter of pure sophistry. It points to the disturbing reality that it isn't at all difficult to set up an interaction hamiltonian between a superposition state and a macroscopic system (for example a Stern-Gerlach device). Yes, our everyday intuition tells us that the possibility that this device could be in a superposition of states is absurd, and so we have to posit wavefunction collapse.
But, near the end of the lecture by Sidney Coleman I posted, he asks us to consider for a moment the possibility that there is no wavefunction collapse. Say we're all quantum systems, and when we look at the results of a Stern-Gerlach experiment, we too become a superposition of states. What would that look like to us? He argues that due to the linearity of quantum mechanics, each part of our wave function would be convinced that they had a definite answer, and thus our entire wavefunction be "certain" it had seen wavefunction collapse with probability 1. That is, the world would look very much to us like it does now. This picture derives the classical world from the quantum world, rather than trying to force quantum mechanics to obey our classical intuitions by this ad hoc wavepacket reduction hypothesis.
posted by njgo at 5:27 PM on April 16

The more formal version of this argument is made here: http://www.univer.omsk.su/omsk/Sci/Everett/paper1957.html
posted by njgo at 5:37 PM on April 16

It points to the disturbing reality that it isn't at all difficult to set up an interaction hamiltonian between a superposition state and a macroscopic system (for example a Stern-Gerlach device). Yes, our everyday intuition tells us that the possibility that this device could be in a superposition of states is absurd, and so we have to posit wavefunction collapse.

I don't find a Stern-Gerlach device at all disturbing and I'm not sure why you do, nor do I understand why you would think that a Stern-Gerlach device itself is in a superposition of states.
posted by ssg at 6:45 PM on April 16

"You should say: 'According to the interpretation that I subscribe to, which has, of course, no experimental basis, it is difficult, but not impossible...'"

No, because this is not a matter of "interpretation". There is no interpretation required. The theory doesn't exclude superposition of macrocosmic objects—the whole point is that theory is silent with regard to the this division. The burden of proof is on those who assert, without theoretical basis, that macrocosmic objects cannot be in a superposition.

That assertion is a much different thing than simply saying that, in practice, there seems to be a pretty profound division between the macrocosmic world and the microcosmic world in this regard and we neither experience macrocosmic objects in superposition, nor has anyone managed to conduct an experiment where this is accomplished. Therefore, an ad hoc explanation for this is that whatever it is that causes the wavefunction to collapse can reliably be assumed to take place always at the microcosmic level.

But just telling people that macrocosmic objects can't be in superposition, as if theory demands that this be the case, is at best an exageration and at worst a serious misrepresentation.

I understand that given all the really bad QM pop-science books out that there which use sensationalist accounts that assume macrocosmic quantum weirdness that anyone who really knows anything about this stuff, especially particle physicists themselves, will naturally want to nip these sorts of speculations in the bud. But I think that flat overstatements like yours and nat's end up doing more harm than good when people go on to read/here comments like that from Weinberg or Coleman that contradict it. It's much better to say something like, well, it's complicated and while we can't rule out macrocosmic superposition, there's lots of good reasons to suspect that the wave function always collapses at the microcosmic scale.
posted by Dances with Werewolves at 7:22 PM on April 16