Do you use game theory?

[POKhER]

Posted here as i assume some of you apply game theory & metagame and all rest of it.

Anyhow i wondered if anyone had hands where they'd like to post their chain of thought or if/where/how they applied game theory.

I just read TOP last night and am trying to figure out how we'd apply such a concept when we don't have complete information.

For example we don't know how often an opponent bluffs... etc.

Thanks.

[Digs]

I thought I was using game theory every time I varied my play?

[Gildwulf]

Yea, you use game theory every time you do stuff like:

-in blind battles 3-bet KK or AA 80% of the time from BB and smoothcall 20%
-check-raise a flop into a PFR with any part of the board and 10% of the time as a bluff

[POKhER]

Hmmm, yeh fair point. I jsut discovered it last night upon reading TOP.

Thanks for applying it to hold'em gilwulf.

So do any of you use it in the sense sklanksy has described, such as bluffing?

Hmm i find this hard to apply, i guess its more of a random thing we apply to our game to "Mix things up" rather than conciously apply mid hand vs an opponent?

I'm struggerling to understand how we can ever use it to call potential bluffs. I'll never have a read on a player that says "he bluff 20% of the time approximatly".

On to the next subject was about making people whom bluff lots, bluff more and those who bluff rarely... make them bluff less.

So by check calling etc, dam i need to re-read this book and try see if i can think up times where i'd apply these deeper concepts.

Poker just turned a whole new corner and revealed a heap of things to learn [img]/images/graemlins/smile.gif[/img]

If you are intereseted there was an article about game theory in the Jan Ed. of the magazine.

[Gildwulf]

Also Pokher,

Put in very simple terms you are looking for a 'best reply' (Nash Equilibrium) to someone else's strategy. It's not just "mixing it up" i.e. 'hey, I check-raised the flop with this hand last time so I'll turn-raise it this time'. It's a theory of strategic interaction.

I.e. if you are playing against a PFR who bets every flop no matter what but rarely 3-bets a flop, the best reply may be to check-raise more often. The amount you bluff-checkraise depends on the amount he is willing to lay down a hand to a check-raise.

If the amount of information you have changes, the optimal strategy changes. For instance, if you know a PFRer likes to 3-bet lightly on flops with overcards, you may want to checkraise less and raise your hands for value more.

Hope that clarifies a bit of the practical use of game theory.

[POKhER]

Thanks for link pharity.

Gildwulf - Yeah, i was just finding it strange when it came to the bluffing %. Sklanksy gives us alot of information to explain the concept and this information is usually unknown or assumed by a rough read.

Definatly an interesting concept, do you yourself apply it in game?

It seems its not something like... Free card play that you learn and apply at the table in the right circumstances.

I mean could you ever work out the perfect bluffing % vs an opponent at a 6max table after a few hands... or even if you logged 1000 over the next three days with him?

Maybe im attempting to learn something not worthwhile, but still nice to understand it fully.

[climber]

The people that are the best at using game theory dont often understand they are even using game theory.

The people that are the best at making smart sounding posts about game theory tend to hang out in the poker theory/microlimits forum and aren't usually very good at poker.

One guy who at least on the surface seems to be good at poker and also knows how to talk about and incorporate game theory into his play is Andrew Prock

His blog often includes HH and discussions as to why he made certain moves.

He plays a lot of short-handed medium stakes.

[jpmack21]

Game theory is most definitely not about applying judgement, i.e. varying your play based on a perception of history, reads on your opponent, etc.

It is the opposite of this. It is randomizing your action such that an opponent with potentially superior judgement loses his/her advantage.

For example (paraphrased from TOP), if a bet on the river would give your opponent 6-1 odds, randomizing your bets so that you end up bluffing 1 in 6 times in this spot would be an application of game theory. This bluffing frequency is accomplished NOT through (potentially inferior) judgement (i.e. I did this in the past and I think he knows that so I'll do something different this time) but rather through randomization (I'll bluff if the river card is lower than 5 and red, or the second hand on my watch is between the 5 and the 7). In this way, game theory allows you to vary your play without having to rely on your judgement skills (which may be inferior to your opponent's).

The randomization of action in a competitive context is one of the principal drivers of the attention paid to poker in military/intelligence circles. Eliminiating disadvantages in judgement when faced with incomplete information is a goal shared by poker players and field generals alike...

J

[Gildwulf]

Randomizing your play is definitely a major aspect of game theory; this is the same as using a mixed strategy (I will raise p% and call 1-p%). It's only one method though.

Game theory is fundamentally about strategic interaction- action BASED on your opponents actions in an effort to maximize your returns. If the payoff for one action changes, your strategy changes.