Archive for Tagpoker Forum A friendly place for winning poker players to share their strategy to making money playing poker.
 


       Tagpoker Forum Forum Index -> Heads Up Poker Strategy
N86xps

order of decision making

Hello everyone! Just had a thought about the order in whinch we consider different options while making a decision. Say we have a typical poker decision to make whether to call, fold or raise.

Suppose that we have some kind of a weakish draw on the flop and face 2/3pot cbet. We consider our options:

1. fold
We think about it and come up with a conclusion, that it is too weak.

2. raise
we think about it and decide that we do not have a good plan if we get reraised and do not want to sacrifice our equity (it is just an example of thought process, I am NOT claiming that this reasoning is correct).

3.call
BINGO! This is our only option since the other two are bad!

Now consider another order:

1.call
We calculate our pot odds + take into account, that our opponent second barrels a lot + we are OOP... so calling is out of question.

2.fold
We have equity + our opponent does cbet a lot with complete air + we do not want to make life too easy for him folding too many flops. So we do not like it.

3. raise
Bingo! Our fold equity (which we believe we have) compensates for the lack of pot odds, it is a great spot to semi bluff.


This was just a thought. What do you guys think, is it good to always consider the options in a specific order? To me it weems that call - fold - raise is quite a good framework. It's like that: our hand is not good anough to call, but it is too good to fold, thenwe should (probably) raise.
ChrisB

Nonono, never ever ever ever think like example 1!

Every option should be weighted individually and the order really doesn't make any sense.

When you evaluate the villains range and your perceived range you should be evaluating the pros and cons of every decision and then come up with the result that is the best of all three, or said in different way, the least bad decision!

In your first example calling isn't good because folding or raising is bad, this just makes no sense.

I'd think about it like this, with your draw:

There are a LOT of factors to consider, has to be said. Position, Stacksizes, table image, his stats and your history together are among the big ones.

First I'd evaluate his hand range and evaluate my equity, then I'd compare that to the current pot odds I'm being offered. It's much easier to call with draws if you are getting immediate odds to draw it them, do consider however that your equity calculation does not include any turn bets made by villain, whilst your actual estimated equity takes into account that you get to see turn and river. So it's not enough to say, well im getting the required odds, I call.

Anyway, with that said we have to evaluate whether or not we can actually get him to fold anything better if we raise or get him to call with anything worse.

I'd really prefer a specific hand history instead of "we have a weak draw" there's just not enough information.

Your second thought process is slightly better only again I would not assume raising is the end-all-be-all just because you evaluated the other two options as bad. I'm much more likely to raise weaker draws than strong draws and so semibluffing is a pretty strong move, but that's not to say that calling is neccesarily bad.

If villain genuinely barrels every time we call flop then we definitely have implied odds, but we rarely hit so I agree with you that calling against very aggressive villains with a weak draw without showdown value (notice that calling with A2o on 34J is much better than calling with 56 on 34J just because the straight is super concealed, an ace is likely good and we actually have the best hand very often.

I got distracted whilst writing this and I have no idea if my message came through, sorry about the confusion Very Happy
N86xps

ChrisB wrote:
Nonono, never ever ever ever think like example 1!

Every option should be weighted individually and the order really doesn't make any sense.

When you evaluate the villains range and your perceived range you should be evaluating the pros and cons of every decision and then come up with the result that is the best of all three, or said in different way, the least bad decision!

In your first example calling isn't good because folding or raising is bad, this just makes no sense.

I'd think about it like this, with your draw:

There are a LOT of factors to consider, has to be said. Position, Stacksizes, table image, his stats and your history together are among the big ones.

First I'd evaluate his hand range and evaluate my equity, then I'd compare that to the current pot odds I'm being offered. It's much easier to call with draws if you are getting immediate odds to draw it them, do consider however that your equity calculation does not include any turn bets made by villain, whilst your actual estimated equity takes into account that you get to see turn and river. So it's not enough to say, well im getting the required odds, I call.

Anyway, with that said we have to evaluate whether or not we can actually get him to fold anything better if we raise or get him to call with anything worse.

I'd really prefer a specific hand history instead of "we have a weak draw" there's just not enough information.

Your second thought process is slightly better only again I would not assume raising is the end-all-be-all just because you evaluated the other two options as bad. I'm much more likely to raise weaker draws than strong draws and so semibluffing is a pretty strong move, but that's not to say that calling is neccesarily bad.

If villain genuinely barrels every time we call flop then we definitely have implied odds, but we rarely hit so I agree with you that calling against very aggressive villains with a weak draw without showdown value (notice that calling with A2o on 34J is much better than calling with 56 on 34J just because the straight is super concealed, an ace is likely good and we actually have the best hand very often.

I got distracted whilst writing this and I have no idea if my message came through, sorry about the confusion Very Happy


I know, that example 1 is bad thinking:). And I know, that example 2 is a bit better. We obvioulsy need to weigh all options, and consider them individually. But I still think, that the particular order in which you consider could at least save you some time. And maybe affetct your decision especially if you do not have enough time at the table.

Say if you eliminate calling, then your decision is down to folding or raising. If you do not find a reason for calling o raising to be +EV then it is an auto fold, while folding is (usually) Zero EV.

So, the framework:

1. Consider calling, and decide whether it is profitable.
2. Consider raising.
                     IF raising is -EV, THEN
                                    IF calling is -EV, then fold
                                    ELSE call                      
                     ELSE
                                    IF calling is -EV, then raise
                                    ELSE weigh rasing against calling and decide
3. ........ there is no "3"Smile.

It looks like pseudocode, though if I was to write a poker playing computer program it would probably have lines similar to these.

Note, that this order does in fact save some time. For example you do not have to consider folding, just fold when you have already considered calling and raising and both are -EV. Also, if you have found a +EV solution, then why consider folding?

If some of you are familiar with alpha-beta pruning (search method used in chess programming and similar tree search tasks), then you probably know how imprtant it is to, search the tree in some specific order.
ChrisB

No i still dont agree, just because calling and raising is bad does not mean that folding is the right option.

And I think you need to be very careful with making your game so mechanical, your thought process should be free flowing and dynamic instead of limited by this framework
chesslw

Yeah- in an ideal world if you know everything about an opponent- analysing plays like this is the only way to go. Working through decisions in this way is still very useful don't get me wrong, but tbh when I play (and I think most people here also), we find most plays to be pretty much instant/standard. This is because of how we have learnt/played poker and comes only from experience.

I don't know if you play backgammon- now there is a game that the computers dominate. The (maybe former) world number one player "felafel" learnt his game on the streets of New York, and constantly reviews his play. And when people analysed his game against that of a computer, it was discovered that they played very similarly. He is a lot better than many other players who sit and analyse/copy computer plays all day. What does this tell us? The human brain's way of learning is through experience I'm afraid.

When I first started, after reading books like the mathematics of poker I tried to "solve poker", or at last think about it methodically (trying to play unexploitively etc). Now I realise just how impossible this task is. Just 15bb plays is already difficult enough/unsolvable.

I'm not sure what you want out of poker, but I would suggest that playing tons of volume and posting hands here for review is the best way to improve.

I suspect that you are similar to me and enjoy poker as a mathematical puzzle in itself. But we have to admit, that approach may get us somewhere, but it doesn't beat playing games, making notes, and reviewing hands.
N86xps

chesslw wrote:
Yeah- in an ideal world if you know everything about an opponent- analysing plays like this is the only way to go. Working through decisions in this way is still very useful don't get me wrong, but tbh when I play (and I think most people here also), we find most plays to be pretty much instant/standard. This is because of how we have learnt/played poker and comes only from experience.

I don't know if you play backgammon- now there is a game that the computers dominate. The (maybe former) world number one player "felafel" learnt his game on the streets of New York, and constantly reviews his play. And when people analysed his game against that of a computer, it was discovered that they played very similarly. He is a lot better than many other players who sit and analyse/copy computer plays all day. What does this tell us? The human brain's way of learning is through experience I'm afraid.

When I first started, after reading books like the mathematics of poker I tried to "solve poker", or at last think about it methodically (trying to play unexploitively etc). Now I realise just how impossible this task is. Just 15bb plays is already difficult enough/unsolvable.

I'm not sure what you want out of poker, but I would suggest that playing tons of volume and posting hands here for review is the best way to improve.

I suspect that you are similar to me and enjoy poker as a mathematical puzzle in itself. But we have to admit, that approach may get us somewhere, but it doesn't beat playing games, making notes, and reviewing hands.


It is a very difficult puzzle:). Yet mathematics might give us some clue of how to go around different aspects of the game.

Just something to note: mathematics itself is progressing ONLY because things, that seemed complex are constantly being simplified and generalized. This could be the way to go around poker as well. If you simplify some particular part of your thought process, then you can concentrate on other more complex things.
chesslw

From when I last thought/explored the concept- I think the actual mathematical tools needed to solve nlhe (HU of course- multiplayer game is not solvable) are already there. The only problem is only complexity. Hold'em is a pretty "simple" game mathematically. Complexity is the same reason that chess/go isn't solved- and (deepstacked) poker is more complicated than both of those... (Though nash equilibrium exists for under 8bb play and is optimal).

I've been informed that quantum computing is the way forward- if they manage to make that work, then poker being solved is the least of our worries (all present encryption methods become useless- bank cards, password to the nuclear deterrents throughout the world etc).
AxeMage

hm.. interesting convo..

i feel like chesslw is right, in the sense that it's a game where you need to adapt a lot and having a fixed way of looking at the game is gonna limit you on how many opponents you can beat.. just like rock paper scissors, every style can be beat by another style and also different thought processes can be beat by another type of thought process, and if you have a fixed thought process, it can also be read, making it easy for your opponent to figure out how you think and outplay you
chesslw

I suppose when you are making any decision fundamentally you are trying to make the most +EV decision. Although most good players won't think about it that way- and will make the decision they think is right (maybe the highest stakes players do idk). Like sometimes you can say that this hand is not good enough to call with, but you have enough equity to make it +EV to raise with it rather than fold...

To be honest though- like counting combos (in the other thread)- I rarely think like this when playing poker at up to 20 dollar stakes hu... Although you need to think a bit more mathematically for sngs considering the mathematical nature of icm. I also haven't played anyone good enough at hucash to think like this either.
N86xps

chesslw wrote:
From when I last thought/explored the concept- I think the actual mathematical tools needed to solve nlhe (HU of course- multiplayer game is not solvable) are already there. The only problem is only complexity. Hold'em is a pretty "simple" game mathematically. Complexity is the same reason that chess/go isn't solved- and (deepstacked) poker is more complicated than both of those... (Though nash equilibrium exists for under 8bb play and is optimal).

I've been informed that quantum computing is the way forward- if they manage to make that work, then poker being solved is the least of our worries (all present encryption methods become useless- bank cards, password to the nuclear deterrents throughout the world etc).


Of course poker could be solved 60 years ago if they had the computing powers.

Yet finding the equilibrium solution of poker is not the only thing, that math can do for poker. There could be so many interesting mathematical concepts, that make exploitative (as opposed to optimal) play easier and more +EV. As well as in chess, you can just search the whole tree of variations, OR you can take some certain position and solve it logically/mathematically. A simple example is the square rule for a pawn against opponents king. Also it does not take an extensive search to show, that a Queen and King always checkmate a lone King.
chesslw

Yep- I know what you mean (In chess actually all 6 piece and 5 piece endings have been solved- i.e. all legal positions with 6 pieces on the board are solved). I also agree that maths is very useful (having studied it), although I would argue that it isn't really maths we are talking about here (actually solving a game such as poker), it is "engineering".

I agree that taking some parts of the game and solving it is very useful as a learning tool for any game, especially if you can draw some rules of thumb and see patterns (like your chess example). So far though, the only (practically) useful mathematical concept in poker is nash push fold (though I'm taking certain concepts like fold-equity, semi-bluffing for granted)... And I really don't see the usefulness of the games solved in "The mathematics of poker" except to illustrate some concepts (card removal, unexploitable betting frequencies, what it means to solve a game etc).

I used to try to play unexploitively- e.g. on rivers I would decide that sometimes I have to pay someone off (depending on their bet size). I used maths for this- e.g. if they bet small enough such that with the hands I would arrive at the river, I have to call with the top x% otherwise I am being exploited (by someone always taking this one and betting with 100% of his range on the river). This worked in the same way as nash, although is in practice quite difficult to carry out especially if it is borderline. And in poker, sometimes you are supposed to be "exploited" (e.g. nash 4bb is advantageous for the sb).

To be honest, I found it much more +EV to use reads and experience in such poker situations... I still contend that volume is the best way to improve your game. Looking back at some of the hands I posted (here and in other forums), I now laugh at my old self- some of those decisions seem trivial to me now.
       Tagpoker Forum Forum Index -> Heads Up Poker Strategy
Page 1 of 1