Poker Therom?
Dave,
I really enjoy your posts and your thought process. Maybe we can get a few of the folks on this forum and create the Krazy Kanucks and go after the Crew (that's another post).
I've got a theory which still isn't fully formed. However, I think I'd like to get your thoughts to help.
My theory suggests that Slansky's Fundemental Theorm of Poker, as it's commenly interprested does not apply in NL tournament play. When you talk about "fold equity", it somewhat explains what I'm thinking. Sometimes (actually many times) it's better to get your opponents to fold at a certain spot rather than continuing on with the hand. Taking a pot of a certain size is more beneficial than the risk of losing the pot even if the odds suggest you should continue.
So, I think the commonly accepted approach to "get the money in with the best of it" doesn't allways apply in NL tournament play. I'm not advocating getting the money in with the worst of it, but instead we should s somtimes get enough money in with the best of it, to get your opponent to fold. And sometimes it's better to fold with the best of it, if the odds are so thin and you cannot make your opponent fold.
The problem I'm having is how does one create a model to represent this thinking. Is it pointless, and more situational. I really think there's model there.
You need to factor in the risk-tolerance of your opponent, the pot-size (what would be worth winning 100% vs. continuing), and you're current needs at this particular junction of the tournament. Stack size would factor in the risk-tolerance of your opponent, as the stack grows shorter, the higher the risk-tolerance and higher chance they will continue with the hand.
So, I'm not sure where to head with this. Just some thoughts and ask for your comments.
Cheers
Magi
I really enjoy your posts and your thought process. Maybe we can get a few of the folks on this forum and create the Krazy Kanucks and go after the Crew (that's another post).
I've got a theory which still isn't fully formed. However, I think I'd like to get your thoughts to help.
My theory suggests that Slansky's Fundemental Theorm of Poker, as it's commenly interprested does not apply in NL tournament play. When you talk about "fold equity", it somewhat explains what I'm thinking. Sometimes (actually many times) it's better to get your opponents to fold at a certain spot rather than continuing on with the hand. Taking a pot of a certain size is more beneficial than the risk of losing the pot even if the odds suggest you should continue.
So, I think the commonly accepted approach to "get the money in with the best of it" doesn't allways apply in NL tournament play. I'm not advocating getting the money in with the worst of it, but instead we should s somtimes get enough money in with the best of it, to get your opponent to fold. And sometimes it's better to fold with the best of it, if the odds are so thin and you cannot make your opponent fold.
The problem I'm having is how does one create a model to represent this thinking. Is it pointless, and more situational. I really think there's model there.
You need to factor in the risk-tolerance of your opponent, the pot-size (what would be worth winning 100% vs. continuing), and you're current needs at this particular junction of the tournament. Stack size would factor in the risk-tolerance of your opponent, as the stack grows shorter, the higher the risk-tolerance and higher chance they will continue with the hand.
So, I'm not sure where to head with this. Just some thoughts and ask for your comments.
Cheers
Magi
Comments
You work to end pots earlier and take your small wins, the only problem is that when you have opponents who don't care about variance and are swinging for the fences as it were, then you sometimes can't play that game
Wish I could draw a graph. I'm not saying win the small pots. Rather, there's a point when the pot becomes large enough that winning any more from the pot won't radically change your situation, but losing it will.
For instance it's midway through a tournament, and you're about 25% above average stack with T12,000 and the blinds are 200/400 with antes of 50. You're in a three way pot that's 6,000. It's the turn and you're ahead and have a 40% chance of winning the pot, and your opponents have 30% chance respectively. Here it may be better to make the bet big enough to get one or both players to fold, than to make a bet that they can call. You have postive EV here, but they collectively are ahead. So, IMO, it's better to get them to fold here if possible. If you can get one of them to fold, they you're odds of winning the pot go up so you could now be 60% favourite to win the pot. However, getting one player to fold may be a smaller EV play in the long run -- playing the hand say 1000 times. But, it's better in the short run as it gives you much better odds to win the pot, and if they both fold you get a decent size pot which does make a difference in how you move ahead.
Looking at this in the worst situation -- you go all-in with the best of it and get called by both opponents. It's a positive EV play in the long run. However, if you win you'll have T36,000 vs T16,000 if you got them to fold at the turn. With T36,000 at this point, you likely won't be a chip leader and it's not that huge a difference between the two stacks. A good hand will double up T16,000 -- unfortunatley with T36,000 it's extremely difficult to double up, as your opponents won't have enough chips to double you up. IMO, there's not that much difference in the stacks at this point in the tourney. Given that you would have a 60% chance of busting out, it's not worth the risk to get the T36,000.
As you mention, you always need to be aware of which opponent will be swinging for the fences. I describe that as risk-tolerance, and figure that must fit into the model. You need to assess the risk-tolerance of the opponent before trying to get them to fold, when you actually want them to fold.
Cheers
Magi
i try to cut out my risk when playing tournys and take down pots whenever i see the chance, unless i feel extremely confident my hand will be the nuts at the end of it
Unless you have a monster hand, I think there is value in taking the pots quickly when you believe either you are heavily favoured or your opponents are likely to fold. Once the pot gets to a decent size, a significant bet will scare out most conservative players, even those with better hands. I'm sure lots of players steal big pots from me with lousy hands. :banghead:
As an alternative, you could slow-play a strong hand to see if your opponents will build the pot for you. Your hope is to still have the best hand on showdown but you run a risk that an opponent can outdraw you OR make a big bet then force you to decide if you want to risk your chips. I don't like either of those scenarios. Fortune favours the brave!
I'd like to hear Dave's opinion on this subject, too.
The pros answer to this question is situational. I haven't seen a model that shows the point at which winning the pot is better than taking the risk to win any incremental amount that goes into the pot, with you as the favourite. If you could see all the cards, except the ones to come, what is the best decision?
I think of ring games as a Casino. You just need to get enough hands in, and you have an extremely good chance of comming out ahead. You don't even need that much of an edge. Think craps, or black jack. Volume with a small edge. With the proper bankroll, you can do this in a ring game.
I think of NL tournament play as a bookie. Yes, the bookie uses the same principles as the Casino, but they always need to protect themselves against ruin -- a big long shot that bankrupts them. Yes the Casino has to worry about this as well, but it's not as much of a factor. The bookie needs to hedge some bets, even though they are the favourite. They are willing to sacrifice some of their EV in return for some security that they won't go bankrupt. Kind of like the Insurance business -- they reinsure many risks, so they protect against ruin.
I'm looking for a Model or a Therom that can show the point at which you sacrifice EV for the security that you don't go bust.
Cheers
Magi
You're right, there is a model there, and it is much more complicated than the model for ring games. In ring games, your goal is simple: win (chips/money) in the long run. This is acheived by making the play with the maximum EV (including such factors into your EV estimate as whether your opponent will fold to a raise, how it will affect your table image, etc.), assuming your bankroll is large enough to handle the swings.
In a tournament, your bankroll is NEVER large enough to handle the swings - that's the nature of tournaments. Winning money in tournaments is no longer about maximizing your EV*, it's about balancing a high EV* with a low variance. Sklansky's tournament book gives an excellent treatment of this, with some theory and examples, along with some tips about how to achieve it. There is definitely room for development of this model, though.
* Actually, you're still trying to maximize your EV in terms of money, which is no longer the same thing as maximizing your EV in terms of chips.
It is especially tough because any model you have to figure out will also involve the sliding pay scale of a tournament. And then it gets REALLY tricky to estimate what the various outcomes are...
So, for instance, you are in a hand and there are twp options:
(1) Win with 60% cerainty and double up or go broke 40%.
or
(2) Increase your chip stack by 10% at no risk.
Which will pay you better in teh long run? The model will have to include a lot of things -- how far away are you now? Fir instance, if it is the first hand of the tournament I will pick (2) but if there are 11 players left and option (1) double you back to average chip status then I will pick (1).
I am not clever enough to build a math model.
Having said that, I do not think one can EVER pass a 60-40 edge. If you KNOW you are 60-40 then take it. That is a BIG edge. Cross your fingers.
Many thanks for your response. I often think they journey is more important than the destination -- this case in particular. So, while I may never get to a model, how it forces me to think through the problem will help my thought process.
I really like how you're factoring in the sliding pay scale into the thinking. I hadn't thought too much about that. And, I like Axiom's thought on maximizing money EV as opposed to chip EV.
Well, I guess I'll let it percolate in the subconcious for a while. When it resurfaces as a more refined thought I'll post again.
I'd be happy to hear anyone other thoughts as well.
Cheers
Magi