blitzing on third and long and the problem with poker forum advice

[KurtSF]

Great post. So so true. I've been noticing this more and more recently, which is why I long since stopped coming here to learn the right play and started coming to learn how to think about the possible plays.

[RoundTower]

I occasionally write something like "I check about 2/3 of the time here against most players". but I don't do that very much because a) I can't really justify arriving at those numbers, other than it seems like an appropriate way to mix up my play and b) these forums are great and all, but that is a lot of information to give away for free.

[RoundTower]

also I don't use a die or the suits of my cards or anything to randomise those decisions. I just guess, and that's not really random.

[BluffTHIS!]

pete,

I suspect your post might have been inspired by Chen & Ankemann's new book, either from reading it or just contemplating the topic. But maybe you shouldn't read such stories before bedtime or you'll have nightmares about the exploiter man coming to get you [img]/images/graemlins/blush.gif[/img].

However, both you and GoG imply that there is or can be theoretically derived, a complete theory of n-player (where n>2) poker games, which really isn't true, despite the book mentioned above, and not just for plo, but for almost any common variant of limit or big bet poker. That book, and other game theory books and studies can indeed give significant insights that can lead to better strategies in the play of hands. But there are just too many variables to come up with a a complete strategy.

But of course you can get close in many cases. However you only need to do so with other players capable of using same. In reality though you aren't often dealing with tables full of mathematical all-stars, and if you are, then you are most likely excercising very poor game selection without the addition of at least one significant fish.


All,

There is something really more important that has been clear to me about this forum for a very long time, and that is that many regular posters, even ones who play modestly high stakes (compared to the minimum ones), haven't even mastered the first 2 levels of expert play (see Sklansky's essay in Sklansky on Poker). Everyone here *should* be able to do elementary probability calculations, even though using the combin function on excel is fine for same. And everyone here *should have* read and reread fundamental texts like TOP and C/R, the latter of which has many very concise points where an essay needs to be studied paragraph by paragraph to be sure to get all the derivative implications.

Many such posters are really just lazy, and seek (like posters in other strat forums), reliable "standard" plays to use when in a zombie-like trance 8-tabling online. But the fact is, and what separates the best and most knowledgeable players in any form or stakes from those who are merely good, is those players know that many if not most of the so-called "standard" plays actually have sub-cases. These sub-cases can be by position, by opponent tendencies (2nd level counter-strategies), by stack size, the number of players in the pot, etc.

Mason said in one of his posts or essays somewhere, that each 2+2 book starts out as pages of notes, which is expanded into a more explanative and functional written form. And that the difference between experts and those who are not, is that they have more pages of notes. This is where those sub-cases come in.

Also in regard to PLO specifically, many here are lacking in fundamentals about common situations, and for example think that they are destined to go broke in more situations than is actually the case for max 100xbb stacks. Two such situations I see here repeatedly are board sets in small or unraised pots out of postion, and situations where you make a straight on the turn with someone else and no redraws, and think that although you would certainly fold same on the flop, you just have to let yourself get free-rolled on when you make it on the turn for half your stack just because you play it passively then (exceptions are when you can bluff a board pair assuming you have FE versus a certain opponent).



A poster above said:

"which is why I long since stopped coming here to learn the right play and started coming to learn how to think about the possible plays".

How many here or in other strat forums really seek to drain all that can be learned from each hand/situation posted? I mean by thinking about all the possible alternatives, determining what the most often used should be ("standard"), but also which might be used if some of the variables were changed, like with the type of opponent, the position and the stack sizes (and one of the most useful types of analysis is to analyze how villain's hand should be played and not just hero's). The answer is only a handful. And when someone does post such a type of analysis, it often is either not appreciated/discussed, or made fun of because it is more complicated than "standard".



GoG said above:

"Much of the reason for the uselessness of the HSNL forum is that there's little discussion of balance and adaptation, which are essentially the entire game at high stakes and especially in short-handed games. I understand why this is true: a long time ago I stopped wanting to tell my opponents how I was adapting to their play."


I agree with that 100%, and it goes to what I said above. There are a couple reasons for this case of affairs in HSNL and in this forum:

1) Many posters don't have the mathematical aptitude or willingness if they do to use it, to do these types of analyses.

Thus there really isn't much incentive for those who do to do so, when the worker to leach ratio is so high (and further proven by all the serious lurkers who never or rarely post). And when occasionally a serious such topic does get started, like the one posted by Johnny#5 as a result of the infamous plo debates between myself and others in NVG and HSNL threads as to what portion of one's stack one should be willing to call with pre versus presumed AA hands, only one player, SilentA, was willing to do any such analysis. I could have, but I said from the beginning there I wasn't going to spoon-feed lazy posters. I have spreadsheets full of such simulations and probalistic analyses, but why should I or other similar posters share that when they won't get much if anything in return? I pointed out the variables that needed to be considered, but only SilentA was willing to do any math work.

2) Many if not most of the handful of posters capable of such analysis, don't see it any EV in such discussions, as the games get tougher at higher stakes, and especially now post-legislation, where the merely good players are the new fish. Especially when as listed in reason #1, they won't get much in return, and will just be feeding free strategy to lazy datamining stat-whores who use hyper multi-tabling trawling to maximize their earn.



The bottom line to me is that you don't have to be a game theory expert, or even worry about true optimal play. But you do have to have a solid mathematical foundation and ability to do some basic analysis past plugging hands in a sim. And you have to also be adept at the 2nd level of expert play, counter-strategies, or your own play becomes too predictable and either allows you to be exploited, or prevents you from fully exploiting the weaknesses of others who are incapable of adapting to same. If you haven't mastered levels 1 and 2 of expert play, it's foolish and unproductive to try to just to #3 (though as David pointed out in his essay they wrap around and come full circle).



Returning to what GoG said, a lot of expert play revolves around range analysis ("shania" to long time readers of the HSNL forum), both for yourself and your opponent, and adjusting for stack sizes. But again, most posters aren't willing to put the time in with same, and then share it. I've done it a few times in HSNL, as have some others, but it doesn't seem to encourage others to follow suit and really analyze some situations thorougly past simple "standard" plays and basic pot control & stack size adjustments. If you can read someone's range well enough, you can react appropriately with your own in response and come out a big winner. But if you can't, and your own range is very easily defined or played in such a way that all an opponent's decisions are pot-odds driven, then you will get screwed by better players and never realize why.

I will close by giving an illustrative example of same from NL, which is simpler to make points than PLO. There are some very tight players in FR games especially, and even in 6-max, whom I know can NEVER have 22/33 in early position for either a limp or a raise. So when the board is J 8 3r, and I have 88 and such a player is willing to push in deep stack situations, then I know for certain that since he wouldn't overplay a an overpair like that but would exercise pot control, and can't have bottom set, that I am in a set over set situation, and can make a rare online fold of a set.

Now just take that kind of analysis, figure out the equities for all possible sub-components of your and your opponent's range, weight them properly, and you will be able to make the best plays (and come close to optimal play). But if you are too damn lazy to do that, you won't cut it for long with better players who do.


(Yeah I know, another long BluffTHIS! post has mercifully come to an end.)

[pete fabrizio]

[ QUOTE ]
pete,

I suspect your post might have been inspired by Chen & Ankemann's new book, either from reading it or just contemplating the topic.

[/ QUOTE ]

I haven't read it. My post was only inspired by the tidal wave of vapid analysis I found while catching up on the last couple week's postings on this forum.

As for the rest of your post, I think "tl; dr" applies.

[piiop]

That is one long post

[pete fabrizio]

[ QUOTE ]
My only quibble with your post is that in a game as complicated as PLO no one knows what the optimal mixed strategy is, and that in practice what you're playing is your best guess at the optimal mixed strategy against your specific opponents. This is often a strategy that you know is exploitable but that you guess your opponents don't have the information, knowledge or courage to exploit. In a big bet game where showdowns are rare, everyone is groping in the dark.

Taking your example a little further, imagine you bluff too often with the dry ace and people start noticing. This drives up your profit when you actually have the nut flush and allows you to start value betting big non-nut flushes harder since now you're getting called by small flushes or even two pair. Maybe you make more money even though in isolation your dry ace bluffs are unprofitable. Like a loss leader in a supermarket.

I don't think I'm telling you anything you don't already know, but talking about these things keeps the game interesting to me.

[/ QUOTE ]

Thanks Gab. I agree with you completely that the necessity of considering mixed strategies isn't really negated by the extreme difficulty of figuring out the game-theoretically perfect solution. Sometimes it can be quite simple -- e.g., people ask over and over again whether they should call a particular river bet. If their opponents are complete automatons, then this is simply a matter of analyzing the opps hand ranges and figuring out which move is right. But if their opponents are smart and adapt well, then my standard answer is "call enough so they can't run over you, and fold enough so they can't value-bet you to death."

Also, I think another reason that people's eyes glaze over when they think of "mixed strategy" is that they think of really boring defensive poker -- i.e., the perfect strategy that can't be exploited but which won't necessarily win you the most money. But mixed strategies obviously don't have to be defensive -- part of the reason you mix up your play is to get your opponents to make "adjustments" that put them in a worse position than they were to begin with. To take the lone-ace bluff example again, the optimal % of the time that you should run it depends a lot on how much your opponent will react. Against some opponents you only have to run it every once in a while in order to get paid off more often with your high and/or nut flushes, while against others you would have to run it so often that you lose money on the bluffs themselves. But figuring out the optimal % isn't just some dry game-theory question about when your opponent should be indifferent between calling and folding, it's still a question of which strategy you should employ to make the most money.

[Troll_Inc]

[ QUOTE ]
also I don't use a die or the suits of my cards or anything to randomise those decisions. I just guess, and that's not really random.

[/ QUOTE ]

I personally use Mickey's big hand for such purposes.

[CallYNotRaise06]

this is the best plo thread ive seen in a looooooong time.

[PokrLikeItsProse]

This reminds me of Matt Matros, a math-y guy for sure, writing about how some "game theory" players are just not as good as they could be because they can't accurately put their opponents on a range. I'm also reminded a bit of Daniel Negreanu's pro-smallball arguments that have the NVG folder up in arms.