Game Theory

[doubleup]

Anyone use this?

I'm trying to incorporate it into my game.  Here's a bit of an artificial situation to hopefully start a bit of discussion.

You have

The board is




Your opponent has a pair and bets 1/2 pot and you call.  He is fairly certain you are on a flush draw...   You are 4/1 to hit so it would appear to be a break even situation as the flush will be obvious if you bet the pot when you hit and he will fold.  So what % of the time should you bet on a heart river to make this situation profitable?   

[matt674]

Too many unknown factors to answer the question.

Was there any betting on the flop?

If there was and you called then he wont be putting you on a heart flush draw unless you have maybe AK or KQ of hearts

What is your opponents perception of you?

If you have been loose the entire tourney and regularly bluff at pots then there is a strong chance you may get called, if you have been playing tight solid poker then your bluff is more likely to succeed.

A bluff will only work if all the pieces of the jigsaw fit in your opponents eyes - if you suddenly bluff at a low rainbow board when you've been playing really tight poker isn't going to work.

What is your perception of your opponent?

If he's a calling station who will call with any pair then there's no point but if you know he is capable of laying down a hand like KQ or KJ in a situation like this then it may be profitable in the long run.

What stage of the tourney are you at?

If its early and it is still cheap you may get called for "information purposes" to see the kind of hands you play. If your nearing the bubble or another leap in the prizepool and your opponent is wanting to step up another rung in the prizepool there may be more chance of him folding.

[SupaMonkey]

Anyone use this?

I'm trying to incorporate it into my game.  Here's a bit of an artificial situation to hopefully start a bit of discussion.

You have

The board is




Your opponent has a pair and bets 1/2 pot and you call.  He is fairly certain you are on a flush draw...   You are 4/1 to hit so it would appear to be a break even situation as the flush will be obvious if you bet the pot when you hit and he will fold.  So what % of the time should you bet on a heart river to make this situation profitable?  

Everytime if you think he will fold if he believes you have it.

I believe this type of bluffing works better in stud than omaha/holdem and you should be able to beat the return you would get from optimal bluffing theory using good reads etc.

[thetank]

I think the whole idea is to simplify it down to a situation where...

50% of the time, his opponents hand will be good
50% of the time, his opponents hand will be bad
His opponent knows this.
his opponent will be able, over a period of time, to guage the % of time you bluff. If you always bet in this spot, he will get wise to this, and always call. You will gain nothing in the long run, and his call will be very profitable for him, ergo bad for you. (Game theory principles only work against very good players, as I understand it)

When you bluff at the pot 100% of the time, your opponent course of action is clear, always call.
When you bluff 0% of the time, your opponents course of action is still clear, always fold.

I think the idea behind all this is that there is an optimum % of the time to bet, and your savvy opponent will have to make a tough descion still.

[doubleup]

Too many unknown factors to answer the question.

Was there any betting on the flop?

If there was and you called then he wont be putting you on a heart flush draw unless you have maybe AK or KQ of hearts

What is your opponents perception of you?

If you have been loose the entire tourney and regularly bluff at pots then there is a strong chance you may get called, if you have been playing tight solid poker then your bluff is more likely to succeed.

A bluff will only work if all the pieces of the jigsaw fit in your opponents eyes - if you suddenly bluff at a low rainbow board when you've been playing really tight poker isn't going to work.

What is your perception of your opponent?

If he's a calling station who will call with any pair then there's no point but if you know he is capable of laying down a hand like KQ or KJ in a situation like this then it may be profitable in the long run.

What stage of the tourney are you at?

If its early and it is still cheap you may get called for "information purposes" to see the kind of hands you play. If your nearing the bubble or another leap in the prizepool and your opponent is wanting to step up another rung in the prizepool there may be more chance of him folding.

Forget about all that.

Its a straight cash game situation.  You are a loose player you could have either flush draw as far as your opponent knows

If you hit your flush you are going to bet the pot.  

When the other flush hits there is an optimal bluffing % that makes this scenario profitable.

 

[matt674]

Forget about all that.

Its a straight cash game situation.  

I'm an MTT monkey, i dont play cash games - hence why my answer was based around tournaments, your original post didnt say that this was a specific cash game situation.

[Bongo]

I think the whole idea is to simplify it down to a situation where...

50% of the time, his opponents hand will be good
50% of the time, his opponents hand will be bad
His opponent knows this.
his opponent will be able, over a period of time, to guage the % of time you bluff. If you always bet in this spot, he will get wise to this, and always call. You will gain nothing in the long run, and his call will be very profitable for him, ergo bad for you. (Game theory principles only work against very good players, as I understand it)

Surely if you bluff X% of the time and he calls you X% it's breakeven, as he will also call X% when you have the flush?

[doubleup]

Forget about all that.

Its a straight cash game situation.  

I'm an MTT monkey, i dont play cash games - hence why my answer was based around tournaments, your original post didnt say that this was a specific cash game situation.

The question would still apply early in a tourny when stacks were deep.  I was just trying to point out that it is only the maths of the situation that are important and hopefully get some ideas about where this concept can be practically applied rather than in my rather contrived example.

[matt674]

you programming a pokerbot?

[doubleup]

you programming a pokerbot?

No I am a pokerbot - resistance is futile...

[matt674]

 

[doubleup]

ok a clue

If you bet the pot any time a flush hits and he always calls you, do u win or lose?

If you only bet pot when you hit and he folds, do you win or lose?

If you bet pot when you hit the flush and half the time the other flush hits.  Do you win if he calls all the time?  Do you win if he folds all the time?

[SupaMonkey]

In answer to your original question.

The number of times you bluff purely depends on the bet size you make. If you bet half the pot then for every two times you bet with the flush, you should bluff once. To achieve this you should decide before the river card which heart cards you will bet on (i.e 2-8 hearts) and which you won't (i.e. 9-A of hearts). In this situation you should bet on any heart because you are going to pot the river.

However, you should be able to get a higher EV than this by accurately gauging your oppo's and figuring out which ones are more likely to call pot sized river bets and which ones won't.

If you want to get really complicated you could look at building the pot on the turn to setup your river bluff.


Edit: actually i made a mistake. When you bet half of the pot you should bet on 4.5 hearts and when you pot it you should bet on 9 hearts.

[thetank]

I've always thought the practical application of game theory is a tough un.

It involves playing very good players over and over again, and part of poker is avoiding these good players.

I think I only really apply it one scenario. Not even sure if you could call it game theory, but it's inspired by game theory at least.

Sit n Gos, some tough opposition I run into again and again, it's ok, so long as there are lots of fish on the table. We'll call this tough oppositon MR.P

Situations frequently come up where it is arguably correct to push with any two. If it's MR.P's blind, he can guage that we are in said situation, and make a desicion whether to call or not, based on our large range of any two cards.
The maths of the situation is now easy for him. He can do it based on how well his two do against a random hand, he can suss out his % chances of winning the pot, and go from there.

So I'll occasionaly fold. He wil notice this, and his desicion becomes much tougher. One round I may fold 79o, and the next round, with the exact same stacks and dynamic, raise with 69o. A bit of random in it means things are slightly tougher on him. The end result is he tightens up his requirements slightly, and this si good for me as the less calls I get the better.

Often I'll like Mr.P on my immediate left (even though he's "got position"* on me) so that these situations occur frequently. (He'll be in the BB when I'm in the SB)


*Doesn't matter so much when 95% of your game is all-in or fold.

The key in this example is, you need to be confident that the other player recognizes when you have a +EV situation (or whatever the sit n go computer mathematicians call it) to push with any two. It's also applicable in situations where it isn't quite an any two push too.

[thetank]

I think the whole idea is to simplify it down to a situation where...

50% of the time, his opponents hand will be good
50% of the time, his opponents hand will be bad
His opponent knows this.
his opponent will be able, over a period of time, to guage the % of time you bluff. If you always bet in this spot, he will get wise to this, and always call. You will gain nothing in the long run, and his call will be very profitable for him, ergo bad for you. (Game theory principles only work against very good players, as I understand it)

Surely if you bluff X% of the time and he calls you X% it's breakeven, as he will also call X% when you have the flush?

FTOP thing.

he gains massively from being able to either confidently call or fold all the time.

With a different %, not only does he have a much tougher job finding his optimal % to call you, but it won't gain as much (if it gains at all) than when he can either always call or always fold.