fuzzy thinking?

[PuppetMaster]

How can someone accumlate a 13K stack at a 5/10 table? What is the buy-in? Talk about deep stacks

[Phat Mack]

Talk about deep stacks

Yeah. It sounds like a great game. Plus, limon is now in the catbird's seat. He cost crazy the pot, and crazy is going to come after him. Yum. Time to let some of the moths fly free from the wallet.

[Al_Capone_Junior]

I'm sure this also has to do with how deep the money is. For example, in a typical online game you might not have deep enough stacks to really get hurt by a check-raise. If the money is really deep then you could get shut out of a pot that you would like to see another card on.

Of course the chance of taking it down right away should always be considered.

al

[tewall]

[ QUOTE ]
Hey tewall, I'm not trying to be argumentative here, but I just don't see this as correct.

[/ QUOTE ]

I'm not taking what you're saying as argumentative. Just an attempt to understand "the truth" better. The quest behind posting in general.

[ QUOTE ]
If limon is behind the old man, then there is no such value. I don't see how old man's hand can then be considered irrelevant, particularly given that this is the strategy that limon was pursuing.

[/ QUOTE ]

There is value in betting. If limon improves to a hand that beats the old man, then there is value in betting whenever limon has a better hand than crazy, improves to a better hand than crazy, or knocks crazy out of the pot when crazy would have beaten him.

[ QUOTE ]

Here are my problems with the fuzzy math:

1. I believe the 4/15 number is much higher. That is, when limon is behind (crazy has a pair or better, he will likely call).

2. There is no accounting for the times that limon gets check-raised.

3. There is no accounting for the loss when limon and crazy both improve but crazy was pushed out of the pot on the turn.

4. I think limon's read of being ahead of the old man is much less than 50/50 (seven players saw the flop, old man called a raise preflop, old man pushed on the flop -- with no tag line that old man was a crazy player).

5. And also... you haven't subtracted the amount that limon wins even if he doesn't bet and it gets checked down from your second number! The first number represents the negative costs of betting; the second number should represent the positive costs of betting (i.e., the amount *over* what limon would win if limon had checked it down).

Given those modifications to your math, the numbers against betting wouldn't be close.


[/ QUOTE ]

1. What do you think the 4/15 number should be?

Implicit in my thinking the bet was good is trusting limon's ability to read the situation accurately. His read was that he could knock crazy out of the pot. If that assumption is wrong, and the reality is that crazy is likely to call him holding a better hand, then limon may have made a bad bet. But I don't know why you would assume you could guess what crazy would do better than limon can when limon was there playing the guy.

You're certainly entitled to your opinion, and you may well be right. But the reason you'd be right would be that limon's assumption that the guy would fold was off base.

2. That was accounted for by Ulysses in the side thread with limon. I made reference to that in the follow-up post to myself. This is a very important point, which I'll return to.

3. The isn't worth accounting for. It's too small a consideration. The point of the equation is to come up with a ball-park figure.

4. O.K. limon, who was there, thought he had a good chance of being ahead. 50/50 is saying he's only half likely to be right. You're saying he's more likely to be wrong than right. That's certainly possible.

5. If limon bets and crazy folds, then limon wins a pot that he otherwise wouldn't have won whenever he has the old man beat or improves to a hand that beats the old man. That's what this represents.

You assert that given the modifications you suggest the numbers wouldn't be close. I don't think that's true, unless limon's way off on his read. You could put in your own numbers for the E.V. calculation and see what it comes to.

If limon is off-base in thinking he's ahead of the old man, or that crazy will fold a better hand, then I agree with you that the bet may be slightly bad. I think limon would have to be way, way off to get it to be "not particularly close".

Where the bet would be bad would be if he didn't intend to call a check-raise, or if calling a check-raise would be bad. Against a non-crazy opponent, betting here would be a risky thing to do, and I believe your reluctance to bet here may be based on a healthy built-in defence mechanism against betting in this situation, which I believe is normally the right frame of mind to have.

[jen]

"crazy is 4/15 (about a fourth, but adjusted for happy math) likely to call... The plus side is P * (p1 + p2) * 11/15 where p1 is the probability that he is ahead (and stays ahead) of the old man and p2 the probability that he improves to a hand that beats the old man."

"5. If limon bets and crazy folds, then limon wins a pot that he otherwise wouldn't have won whenever he has the old man beat or improves to a hand that beats the old man."

That is *not* what this number represents. This number represents the amount that limon wins whenever crazy folds. But limon might have won the pot anyway had he not bet the turn and regardless of whether crazy folded. So you need to multiply by some factor in order to have a meaningful comparison (which reduces the EV of betting the turn considerably):

P * (p1 + p2) * [11/15 * (p3 + p4)]

where p3 = probability that crazy is ahead and stays ahead
p4 = probability that crazy improves to beat limon

[J.A.Sucker]

Hey Limon,

How much consideration did you give to raising preflop (pot), with the hope of the crazy Armenian (isn't this redundant?) coming over the top and running it with him?

Also, on the flop, I would raise, once again hoping to either win it there, or to run it with the crazyman. Of course, this decision depends somewhat on the stacks behind you. If they are short, a straight draw or flush draw may call anyway, since they are "priced in".

[tewall]

OK. I see what you're getting at. If limon's ahead of old man, then he gains considerably by knocking crazy out. If he's not ahead of old man, then he gains less because when he improved to beat old man, most of the time he would have won anyway, regardless of whether crazy was knocked out or not.

I see your point, but don't see how that changes things much. The big gain is still knocking crazy out when crazy is ahead of him (limon) and limon is ahead of old man. So we're back to the bet gains if either limon is ahead of crazy or crazy will fold to him if he's not. It loses when crazy will actually call him with a better hand (regardless of whether limon is ahead of old man). So the crux of the matter is how likely crazy is to be holding a better hand and will call with that hand.

[jen]

Ok, I hate thinking about poker this way, but here's a revision of your approximate math model to illustrate why it's -EV to bet the turn:

Let:
p1 = probability limon ahead of crazy = 0.5
p2 = probability crazy calls if crazy ahead = 0.8
p3 = probability crazy calls if limon ahead = 0.2
P = size of pot (also size of turn bet)

Assume: whoever's ahead stays ahead, no checkraising.

Cost of betting the turn = P * 0.5 (0.8 - 0.2) = 0.3P

Let:
p4 = probability limon ahead & stays of old man + probability limon catches up to old man = 0.5

Profit resulting from betting the turn = P * probability limon ends up ahead of old * probability that limon is behind crazy but gets crazy to fold a turn bet

Profit = P * p4 * (1-p1) * (1-p2)
= P * 0.5 * 0.5 * 0.2
= 0.05P

So:

Overall cost of betting turn = cost - profit
= 0.3P - 0.05P
= 0.25P

This, in addition to the fact that limon could get check-raised on the turn.

Therefore, limon should check.

[Ulysses]

[ QUOTE ]
probability crazy calls if crazy ahead = 0.8

[/ QUOTE ]

While there has been some good discussion around this hand, this illustrates why sometimes it's pretty futile to have these discussions, especially with the goal of determining the "right" decision.

limon said "I decide there is a good chance crazy guy has a small pair" and that he felt there was a good chance the crazy guy would lay down to his bet w/ that hand. Given that, is it really reasonable to use 80% above? I haven't been following the discussion between you and tewall, so I have no idea about the models/numbers in general.

Anyway, I think this point is similar in nature to my earlier point about you putting old guy on only two possible hands that limon could beat. Whereas, based on his description of the game, I gave that a much higher probability.

My main point here is simply that I think there's value in the discussion of hands like this, but it's somewhat futile to try and determine the "best" play in a situation like this without being there.

[jen]

"While there has been some good discussion around this hand, this illustrates why sometimes it's pretty futile to have these discussions, especially with the goal of determining the 'right' decision."

I agree that there are a lot of player-dependent situations in NL that are hard to consider on message forums. However, *in my opinion only*, I think in this instance, this was clearly a bad turn bet.

> probability crazy calls if crazy ahead = 0.8

>> limon said "I decide there is a good chance crazy guy
>> has a small pair" and that he felt there was a good
>> chance the crazy guy would lay down to his bet w/ that
>> hand. Given that, is it really reasonable to use 80%
>> above?

I think the numbers used are pretty neutral if not biased to help tewall's argument. For example, I give it a 50% chance that he's ahead of old man (which I think is very off). I also give it a 20% that crazy will call with a worse hand (which is obviously high since he folded the best possible worse hand).

The 80% only indicates that someone is playing the cards and not some random number generator. If you say he's only 50% to call if ahead and 50% to fold if behind, then you might as well say that crazy is playing blindly with no bias as to what he's holding.

The main idea here is that the cost of betting the turn is an order of magnitude greater than the gain resulting from betting the turn because of the necessary conditions that must pre-exist in order for the bet to be profitable.