Game Theory + Bluffing + Hold'em

[Shandrax]

Hi there,

Q1: Let's say I got a flush draw. 9 outs on the flop give me a 37% chance to make the hand. With pot-odds of 2:1 I should bluff in 18,5% of the cases...right? (-> TOP p. 186)

Now how do I connect this percentage to bluff cards (the cards are out when I have to make the decision after all)? Since I can't roll a dice, how do I make my decision at the table? So far I used intuition, but I'd rather blame game theory than myself...

Q2: I more or less understand that game theory works well in limit games, because I know exactly what to expect and can base my decision on that, but what about NL? I could face a small bet on the flop and an all in on the turn. Can I even use game theory on bluffing in NL? I mean, can there be a "correct" bluff frequency in NL games?

[blackize]

Yes bluffing with accordance to game theory is profitable in no limit. There is an optimal bluff frequency in NL. The reason that game theory includes bluffing is so that your opponent is more likely to pay you off when you have a hand and thus there will always be an optimal frequency.

[diebitter]

make your decision based on your hole cards. In the case of a flush draw, make it based on your previous hands' hole cards.

For example, if it was 25%, only do it if your first card is a heart, say.

For 18+%, remove a few cards from the list to make it as near as. In this case, I think that means remove 2 or 3 cards, so say you decide 2. Remove 2 and 3 of hearts from your 'make the play' cards, and you've got a randomish method you can use at the table. Only go if the first hole card in the previous hand is Ah to 4h.

[PokrLikeItsProse]

One idea which I am borrowing from Paul Phillips is that, if you are in a spot where the math says you should bluff 25% of the time, you can look at your watch and bluff if the second hand is between 1 and 15. You are on your own in determining which way.

Of course, this makes looking at your watch a possible tell, but the whole point is that you make random decisions, not that you use key cards to make random decisions for you. Certainly, cards are useful, but they are not the only thing you can use. If you wanted, you could base it on the the letters in the next word out of someone's mouth at the table. You could use a generate a random string of numbers and memorize it and use the digits of that number to form the basis for your random decision making. Of course those would be more work than necessary, but if you are not imaginative enough to come up with your own method that doesn't require cards and can only play as you are told, then you arguably don't deserve to advance beyond a certain point in your poker abilities.

[Spikey]

There is still so much left open here. The correct bluffing frequency is also a function of how much you bet, and the 2:1 you are quoting assumes a pot limit bet from memory.
Also how many players still there at the flop is also important. With 2 or more players I actually bet for value hoping that they call because you are getting over 2:1 on a slightly better than 2:1 shot on your money. Invariably though I find I often only get one caller making the bet a bad one, but still there's still lot's that can happen on the turn and river to give implied odds and things.
The bluffing frequency is also of course, dependant on the opponent folding at the correct frequency. If he folds more then you can bluff more, if he always calls then you should not bluff at all.
This original problem is far to complicated to be solved simply on this thread I think.
BTW: I haven't been here for a long time and hello to anyone that maybe recognises me, although I don't think there would be many.
Good afternoon.

[senjitsu]

Are you talking about a bluff or a semi-bluff... either way, i think you're a bit confused about how pot odds relate to bluffing.

Assuming you're on an out and out bluff, youre bluff frequency should be determined by the size of the pot relative to your opponents likely calling frequency.

For example, assume there's 6BB's in a pot. I would spend 1BB to bluff at it (for simplicity's sake, assume i cannot win if im called), so I would break even if if my opponent called me 1 time 7 of the time (I would win $6 one time and lose $1 six times). So I would bluff here if I believed that my opponent would fold more than one time in seven.

NL bluffs are a little trickier, because you can adjust the size of your bluff to tweak your odds. For example, I know that if I bet the pot, my opponent has to fold half the time in order for me to show a profit, but if I bet half the pot, then my opponent only has to fold one time in three for the bet to be profitable.

As far as semi-bluffs go, you just add the possibility that you will be called an win anyway into the equation.

so say you have a flush draw on the turn (20% to hit). The pot is $20 and you decide to make a pot sized bluff (which will take you all in). Under normal circumstances a pot sized bluff would just about break even if your opponent called half the time. But as a semi bluff, here's what happens:

50% of the time, he folds and you win $20 (+10)
40% of the time, you lose $20 (-8)
10% of the time, you win $40(+4)
for a total of +6 (where normally the play would just break even)





[ QUOTE ]
Hi there,

Q1: Let's say I got a flush draw. 9 outs on the flop give me a 37% chance to make the hand. With pot-odds of 2:1 I should bluff in 18,5% of the cases...right? (-> TOP p. 186)

Now how do I connect this percentage to bluff cards (the cards are out when I have to make the decision after all)? Since I can't roll a dice, how do I make my decision at the table? So far I used intuition, but I'd rather blame game theory than myself...

Q2: I more or less understand that game theory works well in limit games, because I know exactly what to expect and can base my decision on that, but what about NL? I could face a small bet on the flop and an all in on the turn. Can I even use game theory on bluffing in NL? I mean, can there be a "correct" bluff frequency in NL games?

[/ QUOTE ]

[senjitsu]

I do it based on the first four letters of the next word someone at my table says (add their numerical value, then multiply by two) in addition to randomizing your play, it randomizes how quickly you act.

[ QUOTE ]
make your decision based on your hole cards. In the case of a flush draw, make it based on your previous hands' hole cards.

For example, if it was 25%, only do it if your first card is a heart, say.

For 18+%, remove a few cards from the list to make it as near as. In this case, I think that means remove 2 or 3 cards, so say you decide 2. Remove 2 and 3 of hearts from your 'make the play' cards, and you've got a randomish method you can use at the table. Only go if the first hole card in the previous hand is Ah to 4h.

[/ QUOTE ]

[ACPlayer]

Get a watch with a second hand. Glance at the second hand if you want a bluffing freq of 25percent, then bluff when the second hand is between 0 and 15 -- etc.