Can EV depend on life-span?

[Tommy Angelo]

Can the EV of a particular betting decision depend on the life span of the hero?

I know that seems like a silly question, but I'm quite serious. I am hazy as to what "EV" means and I cooked up an example to hopefully help clarify my confusion.

Think of this as a fairy tail with strange conditions and consequences, like the pumpkin thing in Cinderella.

Joe, our hero, never tilts or plays bad. That's because he has a system that works. Before each session, Joe chooses one hand to fold that day, specified to rank. He never picks a pocket pair to be his folded hand of the day, or a hand better than K-J, or worse than 10-5. To determine his folding hand, Joe deals pairs of cards, at home, until a qualifying hand appears, and that is his daily choice.

As long as Joe continues to fold his chosen folding hand for the day, as he has done for decades, then he will never tilt or play bad, for the rest of his life. But if he is dealt the forbidden hand at the poker table, and he chooses to see the flop with it, just one time, then he will go on total tilt full steam ahead for however long it takes until he is totally broke and destitute and friendless.

Joe prepared at home for a trip to the casino. He got out some playing cards and shuffled them and dealt the top two. Jack ten. That was his folding hand for the day.

Joe arrived at the casino and he got a seat in the world's softest hold'em game. Joe was on the button. Everyone limped. Joe looked at his hand. J-10. Suited even. Joe folded.

Would calling have been a higher EV choice than folding?

What if Joe died of a heart attack one minute later? Does his death retroactively affect the correctness or incorrectness of his fold?


Tommy

[dogmeat]

Tommy, your posts are getting stranger!

Obviously since you added additional information to the playing of hands other than winning or losing on that particular one, the EV of playing becomes hugely - . Or did I not understand this crazy story?

If he gets to play the hand out before he dies, yes, you made it a +EV again.

This can't really be what you are fishing for - what is it you want?
Dogmeat /images/graemlins/spade.gif

[Dov]

Hi Tommy,

Playing the JT, even in the softest of games has to be -EV in this case. That is because under the conditions that you set, Hero will go on to lose everything, his friends included, regardless of the outcome of the hand.

The fact that he destroyed his discipline destroyed his life in this case.

I don't think that it makes any difference if he dies later. He can only make the best decision(s) at the time, with the available information. Since he doesn't know that he is about to die, he cannot factor that information into his EV decisions.

Dov

[bisonbison]

dogmeat's right, either I'm missing something or you're asking the wrong question.

Let's say if he could play JT here 20,000 times, he'd end up earning X/hand, where X is positive. So he'll play it once, with an expectation of X, and after that, he burns through his entire bankroll at a rate of Y/hand, where Y is negative.

Unless he knows that he'll never play another hand (unless X is more positive than Y is negative, in which case he could play a couple of hands, I suppose blah blah blah), folding is better.

[Tommy Angelo]

hi dogmeat,

You were right in that I am fishing for something. Understanding.

I have a hard time with the idea of EV because on one hand, according to some people, the concept of EV takes into account every meta game consideration imaginable, even things that haven’t even happened yet, as with Joe. But on the other hand, according to how the term is usually used, the EV of a poker situation is often stated quite confidently as being plus or minus, as if it were clear for all to see, even when the essential EV-determining variables are not known.

That’s one thing. Another thing is the time duration required for EV determination.

Let’s say we have two players, Neil and Bob. Both of them reraise a hand with 8-7 and get there on the river and win a huge pot. Neil goes on to adjust perfectly to how his opponents adjusted to his 8-7, in ways that maximize the potential residual EV of his 8-7 play. Bob, however, does just the opposite. He adjusts to his opponent’s adjustments, but he does it poorly.

In this case, we cannot know the EV of the 8-7 play for either of these guys until hours later. Is that right?

So when does the statute of limitations run out, on judging the EV of a play, according to the standard definition of EV? At the end of the hand? At the end of the session? And if the EV of a particular play is said to take into account the later effect of image and adjustment and all that, then how can the EV of a play be rated at all, before the adjustments and image play out, or without knowing exactly what those adjustments were and how it went?

I’m thinking about players in general, but also me in particular. I am currently playing very few hands in a way that would be deemed correct by a 2+2 expert panel when taken in isolation. Usually when I post a hand, I am told that I made plays that were -EV. Yet, I do have a decent feel for this kind of thing at the table. I kinda know when I’m making -EV plays and when I’m making +EV plays, and I have never even come close to feeling so many +EV plays right in a row, spanning months, like I am now. It’s euphoric.

So, obviously, something gives here, with the definition of what EV really means, at least to me. I can’t be making +EV plays and -EV plays at the same time. That much I know. As a word guy, when this kind of conflict comes up, I turn first to the words themselves to see if that’s where the mixup is. I’m just pointing a microscope at “EV.” That’s all.


Tommy

[dogmeat]

I assumed you were headed in that direction. The term EV is thrown around here by many posters, and often given in the wrong context. For a player of your caliber I have to make another assumption, and that is that you will make plays based on situation - based on game conditons - based on the other player's understanding of your ability - and based on your ability to "move" the other players in the direction you want them to go.

Weak players never move beyond ABC, and never think beyond the first level of "I have this hand and I'm playing it". As you and many posters here know quite well, an experienced player thinks on three levels, and plays the players (the cards help, but often become secondary). In these cases, two cards in a weak player's hands are -EV, but in a strong players hands can often be +EV. While that expectation is usually given for a specific hand in a specific position, it is not given for real-money decisions based on actual game situations when they include all of the factors I listed.

This probably explains the heated arguments found in the WPT section about certain hands because we see only the cards themselves, not the thinking that was actually used by the players.

Dogmeat /images/graemlins/spade.gif

[Stork]

Tommy, it seems like you are trying to get at the idea that while an individual play can be -EV if taken alone, if that play helps to create a certain table dynamic or image for you that is overall +EV, then the play is +EV in overall, which is what really counts. If that's what you're trying to say, then I think you're right.

P.S. Isn't this like that Shaniah thingamajiggy?

[Ulysses]

As long as Joe continues to fold his chosen folding hand for the day, as he has done for decades, then he will never tilt or play bad, for the rest of his life. But if he is dealt the forbidden hand at the poker table, and he chooses to see the flop with it, just one time, then he will go on total tilt full steam ahead for however long it takes until he is totally broke and destitute and friendless.

Nobody should have any trouble with this question given those constraints.

[ QUOTE ]
Would calling have been a higher EV choice than folding?

[/ QUOTE ]

No. He will lose w/ this hand x% of the time, which leads to him going broke. He will not make enough the time he wins to make up for that. He should only call if his bankroll is so small that he is very likely to go broke anyway in the near future.

[ QUOTE ]
What if Joe died of a heart attack one minute later? Does his death retroactively affect the correctness or incorrectness of his fold?

[/ QUOTE ]

No, since he did not know he was going to die. If he knows this is the last hand he'll play before dying, the EV of the call is simply the EV of this individual hand (based on odds, opponents, etc.) without taking into account the bigger picture considerations, so knowing he's about to die will impact the correctness of the fold. Since he didn't know that, though, taking his heart attack into account is the same as saying his 72o UTG fold was bad because the flop was 222.

[Ulysses]

I raise a lot of hands on the button. However, that range of hands becomes much smaller when I happen to get good hands in the CO and CO-1. When my CO and CO-1 hands take down the pot without showdown, I'm gonna be much more likely to fold my button than I would have had I folded the last two hands.

It's possible to look at EV of a hand in two ways. One is for an individual hand, more or less in a vacuum, taking into account how that hand fares in that situation given all of our collective knowledge about holdem. This is most easy to calculate in all-in preflop situations since we just have to take into account our opponent's range of hands and the chance he folds pre-flop. No post-flop considerations to worry about.

Here's an example. In NL, we all know that if we're all-in preflop v. one opponent w/ AA, that's definitely +EV for that hand, since we have the best hand. On the other hand, if we're all-in w/ 44, that's far less likely to be +EV since our opponent probably has a bigger pair or two overcards (so our opponent is either a big favorite or slight dog).

However, about a year and a half ago, I saw a good friend of mine get all-in preflop w/ 44. The very loose, bad-playing button called w/ KJ. My friend had a very small stack at this point, early in the game. My friend lost a small pot.

Later in the game, my friend went all-in again. This time, he had a very big stack. Another bad player called his all-in. This time, my friend had AA and won a big pot.

His play w/ 44 clearly made it much more likely that he'd get paid off when he made a big hand. So, I think this is an example where the hand itself looked at in isolation was possibly -EV (in actuality, there was some fold equity here, so it was probably +EV against that opponent, but let's pretend he called all-in rather than re-raised all-in), but taking into account the additional money he might make later in this game due to showing down this hand, the "overall" EV of this hand was probably positive.

The tricky part is correctly taking into account the value these plays really do have on future hands. That's where the primary disagreement comes into play in a lot of these discussions.

Another example. You could decide to muck AA pre-flop face-up on the button every time you get it. People would definitely think you are crazy. Would this crazy factor make you enough money in future hands to make up for how much you're giving up by not playing AA there? If, for example, this meant that your opponents would fold every single time you raised, then this would be a good move.

[Monty Cantsin]

I think some of the confusion arises from the slippage that occurs when we start thinking about EV as a measurement of some attribute of a particular situation, like weight or temperature.

I think it is more accurate to think about EV as a property of our information about a situation.

In other words, you take some things you're certain about: (what cards you have, what cards are on the board, the number of bets in the pot) and some things you're uncertain, but have partial knowledge, about (your opponent's hand, what cards are going to out, the reverberating effects of your actions this hand on your opponent's behavior in future hands...), you put percentage weights against the partial information to the best of your abilities, and you arrange all of this into a forumla and crunch it and get a number: the EV.

In other words, EV is a way of organizing what you already know about a situation into an arrangement that helps you see certain properties of that information. It doesn't really add new data about the situation.

If someone happened to know exactly what cards were coming out next, their EV calculation would be much more useful than that of the guy who's using standard card probability for that part, and their EV calc. would end up being much closer to the actual results of the situation once it finishes playing out, but it doesn't make sense to describe it as being more correct. Because, in both cases, the EV is simply a property of the information both players have about the situation.

If you think about it this way, you can stop thinking "how can I get my hands on this elusive piece of mysterious data that is somewhere out there: the EV of this situation?" and instead think: "how can I organize all of the information I currently have in such a way that I can make the most rational estimate about the outcome of this situation?"

So, things that happen in the future that allow us to look back and fill in the partial info sections of our EV calculations don't retroactively change the EV of the situation.

And in many cases, the difference between the EV you come up with and the EV your critics come up with regarding the situations you describe is a factor of the different weights and values you assign to the partial info parts of the EV calculation. For instance, for most people the influence on opponents' future behavior is so ambiguous and unquantifiable that they just make that zero. But if you play against the same guys over and over and have a really good sense of these effects you might be able to plug a reasonable number in there at a fairly high degree of certainty and get a very valid different result.

/mc

[Zetack]

[ QUOTE ]
What if Joe died of a heart attack one minute later? Does his death retroactively affect the correctness or incorrectness of his fold?

[/ QUOTE ]
[ QUOTE ]

No, since he did not know he was going to die. If he knows this is the last hand he'll play before dying, the EV of the call is simply the EV of this individual hand (based on odds, opponents, etc.) without taking into account the bigger picture considerations, so knowing he's about to die will impact the correctness of the fold. Since he didn't know that, though, taking his heart attack into account is the same as saying his 72o UTG fold was bad because the flop was 222.

[/ QUOTE ]


I agree with your answer, but not with your reasoning. His death can not retroactively change the Ev as it existed at the time of the decision. However, the possibility of his immediate death should be factored into the Ev consideration. You'd have to give us details about his age, health etc, then we could consult a mortuary table and get his life expectancy and get a calculation from that on the odds of him dying immediately. I'm sure, however that the possibility of him dying immediately is so vanishingly small that it can not tip the playing the hand from -EV to plus Ev. the math, of course is beyond me but I'm sure Sklansky could manage it.

On the broader concern, I don't think there is a problem with considering Ev on both a micro and macro level. Generally we, or most of anyway, look at Ev in the context of individual hands because the macro factors such as how it will affect other hands in the future are so nebulous that most of us cannot make reasonable approximations of those values. Also, in a teaching context, one is usually concerned with how to play a particular hand in particular circumstances.

For an advanced player, macro considerations can certainly be valuable and can affect the ev of a given play. If you are good enough to make those determinations then by all means they should factor into your Ev considerations. I am both not that good, nor do I play against players who take note of my play, so macro considerations should rarely be factored into any given hand, even were I capable of doing so.

Oh, and on a truly meta level if our Hero were somehow to know he was about to drop dead, I would think playing or not playing that hand would be absolutely Ev neutral since I can conceive of no reason why a few BB's either way in your bankroll at the time of your death could make any difference to anybody.


--Zetack

[Zetack]

[ QUOTE ]
For instance, for most people the influence on opponents' future behavior is so ambiguous and unquantifiable that they just make that zero. But if you play against the same guys over and over and have a really good sense of these effects you might be able to plug a reasonable number in there at a fairly high degree of certainty and get a very valid different result.



[/ QUOTE ]

Oooo, hey that's what I tried to say.

Another thought, Ev is always a guess, not an actual number. It works real well in hypotheticals because you can make assumptions about the opponents hands and behavior. Opponent x will only raise with this range of hands, X will cold call with this range of hands, X will fold thirty five percent of the time if you do this... In the real world any estimates of the opponents hands or behaviors are much more nebulous. Valuable yes, but far from exact. Kind of like advanced economics.

Given that, it always bugs me when somebody says something along the lines of--always make the plus Ev play--as if that were always possible to even determine. Perhaps though, the "expected" part of Ev takes into account the potential inacuracies of any Ev calculation.

--Zetack

[Tommy Angelo]

[ QUOTE ]
I think some of the confusion arises from the slippage that occurs when we start thinking about EV as a measurement of some attribute of a particular situation, like weight or temperature.

I think it is more accurate to think about EV as a property of our information about a situation.



[/ QUOTE ]

Ahh. This resolves the issue for me. Thank you.

For example, if the available information is "you have 7-2 at limit hold'em," and that is all we know, then the best preflop EV option, given the limited givens, is to fold. But if we add extra info, such as "and it is folded to you on the button, and the small blind has indicated that he intends to fold, and the big blind will fold 85% of the time if you raise, and he will only reraise with AA and KK," then the best EV play with the 7-2 changes, with the new information, to raise.

This wonderfully resolves the problem of determining the EV of folding for one chip from the small blind after everyone limps. If that is the only info, and the books say that calling is right, then I agree that calling is right, in the same way that folding with 7-2 is right, by the book, in an information near-vacuum. But with additional information, then folding for one chip can be a better choice than calling, just as raising with 7-2 can be a better EV choice than folding.


Am I still aligned with your thoughts?


Tommy

[Monty Cantsin]

[ QUOTE ]

Am I still aligned with your thoughts?


[/ QUOTE ]

Precisely.

/mc

[Lawrence Ng]

[ QUOTE ]
Can the EV of a particular betting decision depend on the life span of the hero?

[/ QUOTE ]

If I knew I died tommorow as opposed to knowing I would live for another 100 years, yes I would say EV of a particular betting decision would highly depend on my life. Not much quantative EV per say, but the emotional EV would be different.

[ QUOTE ]
What if Joe died of a heart attack one minute later? Does his death retroactively affect the correctness or incorrectness of his fold?

[/ QUOTE ]

Yes, provided Jack knew he was going to die very soon. Who knows what's going through Jack's mind at that given point in time? The situation for him is very different as opposed to having a longer life span.

[Lawrence Ng]

[ QUOTE ]
I’m thinking about players in general, but also me in particular. I am currently playing very few hands in a way that would be deemed correct by a 2+2 expert panel when taken in isolation. Usually when I post a hand, I am told that I made plays that were -EV. Yet, I do have a decent feel for this kind of thing at the table. I kinda know when I’m making -EV plays and when I’m making +EV plays, and I have never even come close to feeling so many +EV plays right in a row, spanning months, like I am now. It’s euphoric.


[/ QUOTE ]

Well, there is no substitute for experience. Your experience has definitely turned -EV situations into +EV ones. Tommy, your getting into some mode of thought here that most of cannot relay upon. I think we've heard this mode of thought before. Most people call this thought the "it" factor.

When you put a Neil, Harry, and Bob out in your example, we have no clue who they are and how they play and what level of game they are at. It's very hard to judge and make a more defined objective statement based on the lack of information. With you, Tommy Angelo, we have a lot of information, and we know you are a respected poker authority.

[Doubling12]

I've got a good one for you -

Suppose a person invented a supercomputer that could analyze every hand played by all opponents, remember everything, and give you the perfect Bayesian / Game Theory / whatever play in all situations against your specific opponents, even adapting as players' tendencies shift (i.e. time of day, etc). This computer could be fit on a watch and brought to the casino, without the other players or casino knowing it exists.

Are the decisions made by the computer +EV? What if I told you that this player keeps his cards face up in front of him when he plays?

Of course that is ridiculous (and would cause a redeal), but the point is that EV can never be discussed when your actions affect the decision-making processes of others. In my example, the player's "action" is to explicity tell everyone what he has.

No one will ever know if, for example, folding 4 out of 5 blinds is right in your shorthanded game. It cannot ever be known. It is kind of a bugaboo actually to read people talking about EV. Short for "Expected Value", it is a defined term that requires all probabilities and payoffs to be known with absolute certainty. Only when the cards are face up does EV mean anything.

[Dov]

[ QUOTE ]
Short for "Expected Value", it is a defined term that requires all probabilities and payoffs to be known with absolute certainty. Only when the cards are face up does EV mean anything.

[/ QUOTE ]

This isn't true. Hence the term 'EXPECTED Value' and not Value.

The reason it is expected to happen is based on its probability of occurance over a significant sample size. How much it will net also depends on a significant sample size.

Since any individual trial can have a range of outcomes, we cannot know for certain what will happen, but we DO know for certain what will happen if we repeat this situation enough times.

This is what EV is telling us. It doesn't have a one to one relationship with the outcome of any single trial.

Dov

[driller]

Well, it's a fairy tale, so wtf? How about this. Anytime someone posts some statistics, someone else will say "sample size". How about this: all the 2+2 books say that everyone gets the same hands in the long run. But what if they don't? Is it so hard to imagine that over a few million hands someone might run lucky? I think it happens all the time. Look around.

[Doubling12]

Umm, I appreciate the reply, but your post does not make any sense. Using the terms "sample" and "trials" in a definition of Expected Value would send some frequentist statisticians into a mouth-foaming frenzy.

[Tommy Angelo]

"No one will ever know if, for example, folding 4 out of 5 blinds is right in your shorthanded game."

Agreed. I think I have a good example here to clarify why.

What if I fold four out of five for two consecutive session, but on one day I fold AT when I should have played it against a total hyper bluffing super payoffer who openraised from the button. And on that same day, I stay in with AT against a super rock non-payoffer who raised from the small blind after the others folded, and I know he most surely has a bigger ace or a pair higher than tens to make that raise.

And on the next day, I get it right, and I play the AT when I should, and I fold the AT when I should.

On both days, my percentage folded was 4 out of 5.

Does the 4-out-of-5 stat, on it's own, tell us anything about the correctness of my blind play? Hardly.

Tommy