Game theory and bluffing

[cnfuzzd]

One of the factors mentioned in TOP about using game theory to determine optimum bluffing frequency was also the you must be playing against a much stronger opponent. Utilizing game theory simply replaces the "intuition" that some poker players have with a pre-determined ratio of value bets to bluffs. Noted, sklansky also says that a combination of the two would probably work best.

peace

john nickle

[DcifrThs]

[ QUOTE ]
I've been going more on feel and estimation of whether my particular opponent is capable of laying down a better hand than I have.

[/ QUOTE ]

My friends, we have here a perfect example of how i used to think and how evolved poker thinking goes.

the example is that you didn't consider THE MOST important piece of information (to be sure i did check to entire post twice to see if you mentioned it) THE SIZE OF THE POT relative to the size of your bet. if the pot is very small you need to be very sure you're bluff will work as you're only getting like 2:1 so it needs to work at least 33% of the time. this will most likely occur and is the theory behind the flop check sb bet out bluff (4 limpers sb completes bb checks. they all check the flop and the texture suggests nothing is out but people are scared. sb bets out on the turn no matter what and gets 3:1 on his bluff which will work more than 1 in 4 times). as the pot gets bigger your bluff needs to work fewer and fewer times to be correct. but to counter that, the opponents calling frequency SHOULD increase with your betting frequency and in very large pots they are usually very close. this is why weak tight opponents are best: they don't adjust their calling frequency to the size of the pot relative to the size of your bet.

also notice that the size of the pot drives the T9 flopped trips bet (poker essays or sklansky on poker)on the AsTsTx3xYs bet and you believe its quite likely the player was on a flush draw but aren't exactly sure. you should bet because you're protected by the pair from a raise and the pot is big enough where his calling frequency is such that he'll call with an ace but won't bet it so you should bet your own underdog hand yourself.

hope you understand this, if not, i'll try to do a better job explaining.
-Barron

[PairTheBoard]

That's a good observation that the pot/bet ratio is important and that a low ratio means your bluff needs to succeed more often and that a high ratio means you will be called more but the rewards are greater. But you fail to mention the most important thing. Should your bluffing frequency be higher or lower with a large pot/bet ratio?

PairTheBoard

[BradleyT]

[ QUOTE ]
there are 4 or 5 guys in my regular game that are pretty tough.

[/ QUOTE ]

Then why are you playing in it?

[DcifrThs]

Before i answer that question i'd like to rephrase the initial explanation because i didn't think it was straightforward enough:

somebody can be "capable of laying down a better hand" yet bluffing into them can still be wrong. If they lay down a better hand than you 1 in 15 times and there's only 10 bets in the pot and 1 bet to call he won't be laying down the better hand enough times for your bet to be profitable.

i think that does a much better job of summing it up.

now, as the pot gets larger your overall betting frequency gets larger and really gets pretty close to 1 if the pot is absolutely massive and you don't fear a raise. of course you always have to scale the betting frequency back just a touch because the calculations all assume no raising and if you fear a raise you would not want to bet at the calculated frequency...just a hair less often.

but as the pot gets larger and your betting frequency increases, you're value bet:bluff ratio should also increase IMO (i don't remember reading this anywhere so i may be out on a limb here and i hope somebody will confirm or disprove this- maybe bigpooch, fossilman, george rice, or somebody similarly qualified). overall you're bluffing should be a little less relative to your value bets even though you'd rather the opponent fold no matter what in this large pot.

your question was "should your bluffing frequency be higher or lower with a large pot/bet ratio?" this means that if there are 30 bets in the pot and 1 bet for your opponent to call you want to know whether you should be bluffing more times or fewer times in general. you should be bluffing more times RELATIVE to a small pot but your value bet to bluff ratio will be HIGHER relative to a small pot so in essence, you're doing less RELATIVE bluffing with respect to value betting i think. but more bluffing overall!! and certainly waaayyyyyyyy more betting. its just that the vast majority of these bets will be value bets and a minority will be bluffs whereas its a closer relationship in a small pot.

i really hope somebody confirms/disproves this here because i've certainly been thinking about it alot and may be thinking myself into a circle because of the general nature of the question.

the opponent's tendencies are extremely important and a feel for the game and the way in which the hand was played all come into play because that is a testament to your opponent's most likely hands. an example is if you knew your opponent and knew for a fact that he will never lay anything down on the river and you have a hand that can only beat a bluff you can't bluff yourself even though its your only chance to win the pot because he'll call 100% of the time and no matter how big the pot is here the only thing you can do is check and fold to a bet if you know your opponent even though the pot is huge.

but my thoughts and feelings after reading and studying and learning as much as i have (which isn't very much relatively speaking) are, for the most part, enumerated above.

lets just hope somebody more qualified than i steps in here.
-Barron

[PairTheBoard]

It seems clear that with a high Pot/Bet ratio you will want to value bet more. But I don't think it's so clear that your bluffing frequency would also be higher - even if not as much higher as the value betting as you say. From BolliTrader's explanation of the Game Theory on this it sounds like with a Low Pot/Bet ratio your bluffing frequency should be higher and visa versa. In practice I think we see this all the time, for example when the SB limps and goes heads up against the BB. Some players will bet the flop 100% of the time in that situation - which means that their total junk hands are getting bluffed with 100% frequency.

PairTheBoard

[DcifrThs]

very true which is why its hard to come to general conclusions.

as you pointed out, i may have missed an inflection point as the pot gets smaller.
-Barron

[PairTheBoard]

There are Game Theory Experts here who must know the answer to this. How does your optimum bluffing fequency vary according to the Pot/Bet Ratio? I'm with you Dc. I'd like to hear what they have to say as well.

PairTheBoard

[BolliTrader]

[ QUOTE ]
Quote:
--------------------------------------------------------------------------------

there are 4 or 5 guys in my regular game that are pretty tough.


--------------------------------------------------------------------------------



Then why are you playing in it?


[/ QUOTE ]

There are quite a few reasons to continue to play a game with 4 or 5 tough opponents.

1. I guess the best one I have is I make money, I was just looking to see if there was a way to make a little more.
2. To get better.
3. Live competition... I live in Pittsburgh and finding a (serious) regular game isn't that easy. I mean sure I have alot of options if I want to play dealers choice, but I don't like guts / screw your neighbor / and whatever other variation of luck / poker games they want to throw out.
4. A few of the players are friends, so it as much about a night out w/ the boys as poker for me.

I'm sure there are more, but that is what pops into my head when asked that question.

deke

[turnipmonster]

I'm not a game theory expert by any stretch of the imagination, but I do have a academic background and interest in it.

in order to calculate bluffing frequency with no cards to come based on the pot size, you first have to make a lot of assumptions about you and your opponent. once you figure out what you both have, how often you can improve, how often your opponent improves you can derive a formula.

there was a great thread on rgp that had sklansky and lederer discussing this. here it is:

http://groups.google.com/groups?hl=e...TF-8%26hl%3Den

--turnipmonster