Ideal stats for Micro-/Smallstakes

Well it's more expensive to continue but by the same logic it's also more expensive to bluff on those streets so there's definitely truth behind that arguement, although the maths does change slightly becasue now they're essentially betting 1.5BB into 3.25BB (About 30% rather than the previous 18%)

I agree that you can't fold that much more than 1-street alpha and folding something like over 30% fold on the flop is going to be a horrific leak...

*checks stats quickly*

Phew

Simlar arguement in that even through exploitation you shouldn't be far away from GTO.

That said I think raising can have an effect on tendency to bluff as most flop bluffs unless the hand is bottom 25% of the range are going to be complled to call until the turn which does increase the expenditure on bluffing somewhat.

Obviously Folding (hand over) and Calling (plays exactly to alpha) don't affect the calcuation but I think Raising espeically when you take into account we have to continue with about 75% of our range (cost 2 into total pot of 7.5 to bluff) will have something on an effect.

But as you say at most it's going to shift it by probably around 2-4% at most from alpha.

NOW LOOK WHAT YOU STARTED!!!

@Boomer: You wanted a tired answer, you got a tired answer...

Simlar arguement in that even through exploitation you shouldn't be far away from GTO.

I don't agree with this statement, GTO play is probably way off general exploitative, or even the so-called balanced game. In GTO you take into consideration millions of different strategies that will never occur in real life (like limping T9s, raising T2o and 32s, and folding the rest). That's why I think we can use a GTO-like strategy that weighs the opponent's potential strategies by their likelihood to occur (that's what I call balanced strategy). Of course it's kind of exploitable, because we have to weigh our options, but still, even if we just take out billions of impossible strategies, this style will be much more effective.

Obviously Folding (hand over) and Calling (plays exactly to alpha) don't affect the calcuation but I think Raising espeically when you take into account we have to continue with about 75% of our range (cost 2 into total pot of 7.5 to bluff) will have something on an effect.

I see your point here, but I kind of don't agree with it. With some of his bluffs one's forced to continue, but he needs to keep some bluffs in his range after being c/r-ed, so that he could potentially rebluff the turn or river (and of course fold some, if the board comes that way - so no valuehands should be folded). Therefore I think it's even beneficial for us. (Imagine we'd fold all of our bluff on the flop, and then a good turn card comes, and you can't raise cause you have no bluffs.) Of course that's not a proof that it shouldn't change our numbers, I probably wouldn't be able to prove it, and you may be right. However having enough bluffs on future streets is really important and valuable.
On the top of that, certain situations might exist, where I call the flop with "too much" bluff, and fold "too much" valuehand, cause it won't effect my later street play that much, but I have more equity with a bluff.

NOW LOOK WHAT YOU STARTED!!!

It wasn't me, you started, or... or JLeitmotiv started it. Yes, he's the one to blame!♦

Ok My Turn ♦

Simlar arguement in that even through exploitation you shouldn't be far away from GTO.

I don't agree with this statement, GTO play is probably way off general exploitative, or even the so-called balanced game. In GTO you take into consideration millions of different strategies that will never occur in real life (like limping T9s, raising T2o and 32s, and folding the rest). That's why I think we can use a GTO-like strategy that weighs the opponent's potential strategies by their likelihood to occur (that's what I call balanced strategy). Of course it's kind of exploitable, because we have to weigh our options, but still, even if we just take out billions of impossible strategies, this style will be much more effective.

The basic strategy behind what I was sayign there was that, unless you've playing a truckton of hands with someone making extreme shifts to your game from balance based on a read (unless your opponent is legitimately horrible) is always taking a risk that you're wrong therefore unless there's something close to mathematical certainty based on previous experience allowing you to make a radical shift.

The decisions we make should always have a basis on what we believe to lie in the balanced strategy rather than just making a huge shift to react to 1 spew

With extra data we can move further from a balanced strategy so I definitely see your point there...unfortunately I have to work in soundbites for vids

Obviously Folding (hand over) and Calling (plays exactly to alpha) don't affect the calcuation but I think Raising espeically when you take into account we have to continue with about 75% of our range (cost 2 into total pot of 7.5 to bluff) will have something on an effect.

I see your point here, but I kind of don't agree with it. With some of his bluffs one's forced to continue, but he needs to keep some bluffs in his range after being c/r-ed, so that he could potentially rebluff the turn or river (and of course fold some, if the board comes that way - so no valuehands should be folded). Therefore I think it's even beneficial for us. (Imagine we'd fold all of our bluff on the flop, and then a good turn card comes, and you can't raise cause you have no bluffs.) Of course that's not a proof that it shouldn't change our numbers, I probably wouldn't be able to prove it, and you may be right. However having enough bluffs on future streets is really important and valuable.
On the top of that, certain situations might exist, where I call the flop with "too much" bluff, and fold "too much" valuehand, cause it won't effect my later street play that much, but I have more equity with a bluff.

Unfortuantely as we go deeper into this particular Jungle I think we're going to keep hitting counter-arguements until we hit the cap

This would involve taking into account the effective cost of those bluffs on the turn and how, in theory that should be mitigated by a balanced bluffing range on the turn and that there's more streets after that and....

ffs :f_confused:

This might be something where we literally have to walk through a hand and see what, if any, massive differences would be caused by 1-street alpha vs an attempt to balance an entire hand.

I'm up for it if you are

You're definitely correct though that some of our peeling range is going to be bluffs which we can win the pot on the turn with as well as the obvious value hands and hands which, in position especially, we're probably going to spike our opponent with for more value.

Are we going to hit "close enough" at any point or are we going for a PS.com record thread

NOW LOOK WHAT YOU STARTED!!!

It wasn't me, you started, or... or JLeitmotiv started it. Yes, he's the one to blame!♦

Yeah, Blame Him!!

*Glares at JLeitmotiv*

Originally posted by madorjan

NOW LOOK WHAT YOU STARTED!!!

It wasn't me, you started, or... or JLeitmotiv started it. Yes, he's the one to blame!♦

Sure, now it's the bronze-status-microstakes-player's fault. Don't blame it on me guys

I love your discussion right now!

Originally posted by JLeitmotiv

I love your discussion right now!

@Boomer:

first point is valid, I misunderstood your statement, so next time I'm not gonna let me talked into posting tired.♦

Second point: we can hit close enough any time, but I think we're once again going elsewhere with this. That effective cost idea - I'm not sure I'm sold really. I get what you're talking about, and in a static game that might be an issue, but given that from street to street our range's construction changes that can be much less of a factor. Imagine this, we call the flop and the turn with an A hi, and get a 4flush river, where we would donk a lot, therefore donk our A hi as a bluff. However, that hand was well in our bluffcatching range until the river hit, therefore we didn't intentionally chose this line, we're just bluffing the bottom of our range. It's just that hand moved so down, it became one of our bluffing hands. But also on an A river, our hand becomes a valuehand, and we have to bluff or bluffcatch with other hands, etc.

I think the key lies in 2 things: trying to balance our range as well as we can for future cards, keeping in mind the probability that they will come (e.g. it doesn't make sense to be perfectly balanced on the Ad river, given we're gonna be imbalanced on all the others). The second thing is on the future streets you act balanced, even if we wasn't prepared for that exact card (meaning sometimes bluffing stronger hands or anything like that).

In poker - unlike in a [0;1] game - you have finite options, unless you use mixed strategies with some hands at least. That makes it even more difficult to try and find "perfect balance", and that's why I'm not suggesting it also. So much stuff depends on these, and I guess they're just tiny factors, I think we could say close enough this time. If we could work it out, great, but I think it's mostly a waste of time and energy, while we have a method that makes you play better than 99% of the players, I'm happy with it.

Unfortuantely as we go deeper into this particular Jungle

I lol'd soooo hard.♦
Oh, anybody who missed the show: balance is overrated, guys. (Well, in this particular topic it probably is...)

Are we going to hit "close enough" at any point or are we going for a PS.com record thread

Nah... you probably need either some stupid promotion or a "site is rigged" theory to do that in my experience.

@JLeitmotiv: Boomer - or the weird ship-captain looking singer guy - is right. I could copy-paste here pages... what pages, books of similar arguments and conversations we had on Skype.

Oh, and BTW Boomer, until I saw you the other day, I imagined you exactly like the Captain in the video you linked.♦ No kidding tho...

Originally posted by Boomer2k10
Well on average I play 37/29 and 15 3-bet recently when playing at micros for videos and madorjan thinks I'm a nit.

If I were to criticise your stats your VPIP and PFR are too far apart, this seems to suggest you're cold-calling when you should be 3-betting or you're open limping, either of those things, except under extreme circumstances, is a problem.

If you want to pot your stats I'll do a quick rundown on them

I am open limping only with the intention of limp/reraising when I have a maniac after me, but I like to cold call more at very loose tables with my lowest range and hope to hit a monster flop.

I thought more serious about poker only recently and I know I have a lot of leaks to fix and a lot of learning to do, but I think I may be on the right way

My stats for 58k hands are:

VPIP 38
PFR 21
3bet 11
WTSD% 49
W$SD% 45
Agg% 42
bb/100 6.26

Originally posted by decentplayer

Originally posted by Boomer2k10
Well on average I play 37/29 and 15 3-bet recently when playing at micros for videos and madorjan thinks I'm a nit.

If I were to criticise your stats your VPIP and PFR are too far apart, this seems to suggest you're cold-calling when you should be 3-betting or you're open limping, either of those things, except under extreme circumstances, is a problem.

If you want to pot your stats I'll do a quick rundown on them

I am open limping only with the intention of limp/reraising when I have a maniac after me, but I like to cold call more at very loose tables with my lowest range and hope to hit a monster flop.

I thought more serious about poker only recently and I know I have a lot of leaks to fix and a lot of learning to do, but I think I may be on the right way

My stats for 58k hands are:

VPIP 38
PFR 21
3bet 11
WTSD% 49
W$SD% 45
Agg% 42
bb/100 6.26

Limp re-rasing is a very old concept more based around FR live game and not really applicable to online SH games. If the guy behind you is a maniac it's actually just more effective to open raise (so you have a chance of stealing the blinds which in FLHE is huge) and then 4-betting him if he 3-bets you. (Or calling an x/r'ing him a ton on the flop, whatever floats your boat)

Cold calling looking to hit a monster flop is really a NL concept based around implied odds.

In FL IO and RIO play a much smaller part than the actual straight up value of your hand and the blinds. In multiway pots cold-calling with hands like pocket pairs and decent SC is fine but in HU pots 3-betting is almost always superior to cold calling.

Looking at your stats you are HYPER SD bound. In a vacuum this may not be a bad thing and you've got decent results to show for it but I would be surprised if at 0.25/0.5 or 0.5/1 the same stats would work. Admittedly my WTSD is mid-40's recently but I play in a lot more HU wide-range pots than I imagine happen at nano-stakes. Such a High WTSD at nano's usually means one of a few things;

1) You can't fold any type of made hand even in a MW pot
2) There is almost no river betting so you get to check down a lot.

You've certainly done well so far, 3BB/100 is nothing to be sniffed at. Looking forward to seeing your future results

Thanks Boomer for looking at my stats

I cold call especially in late position and almost never in a HU pot(which I don't see so much at the limit/tables I am playing). I know that I am SD bound and this is a huge leak(I still can't fold JJ on a AK board), in fact this is one of the things that stop me from moving up.

Hello!

I feel like being a bit provocative :evil:.

I don't think that GTO is anything to ever strive for in the long run. It might take a losing player to a break even player (even a winning one if he implements GTO in a robot so that unintentional deviations are disposed of). But hell, almost the same effect is achieved by using starting hand charts and some common sense.

The more experience I get playing at various limits, the stronger my convictions that good old fashioned poker thinking is worth a hell of a lot more on the later streets than GTO. If you play GTO, then, sure, you will never really screw up. And. you'll win. But I think that making the right deviations from GTO is what brings home the bacon to the really good player.

Since this thread has an underlying quality lurking there in just about every post, I should really come up with something more substantial than "GTO sucks" (it doesn't suck, it's just a bit boring).

Here is one argument to support my case: It's on the river. We are out of position. We have the options B/3-bet, B/C, B/F, C/R, C/C and C/F.

In this spot, the EV formulas are all simple. GTO is simple too, because it is the same thing. GTO will say "do X" here. Here, the choice of Y instead of X might be much more profitable. Why? It is because the EV formulas are very sensitive to the general behavior of the opponent. In GTO, the difference between, say, B/F and C/C may be very small. You have a marginal play X to make. But, throw in a good read and you may have a very solid play Y to make.

In the same vein, on the river, when it comes to bluffing (and, to a lesser extent, catching bluffs), part of the range thinking can probably be thrown out of the window. By this I mean that one shouldn't bluff because one is supposed to bluff here with the bottom of the range. One can first consider if a bluff is profitable with this board and this opponent. The actual holding, and it's ranking in the range (and perceived range) can then take the back seat. Sure, one might lose a little EV by bluffing with a showdownable hand, but this is regained by the effectiveness of the bluff part if one is right.

"Isn't all this exploitable?"
Of course it is, but you are the one doing the exploitation. If they play back, well, take it to the next higher (or lower) level.

"In order to deviate from GTO you need to know GTO."
Yes. GTO (like starting hand charts) is an ingredient, but it doesn't constitute the final delicious dish.

Your turn!

/Johan =

Ok, so from this post it's a 100% clear to me you don't fully understand what GTO is, and especially the way it's being used by balance-monkeys like me.

First off, no one can play GTO, or approximate it in any kind of way you present it here. If someone could, you probably would be right, it would be a pretty complicated starting hand chart, but thank god, no one can. You can derive some numbers / tendencies from GTO play, but given CRE and equity is not calculated into these numbers, even those numbers will differ from situation to situation.

The "balanced range building" method we discussed here (maybe mistakenly call it GTO-play instead of GTO-based play) is built on the fact that we get the set of all strategies possible, weigh each element in a certain way - mostly just by regular tendencies, but sometimes more exploitatively - and take the most profitable balanced (=/= GTO) line against the set of strategies. Of course this can hardly be calculated, so we use simplifications in the process, but all in all, that's the pure basis of our gameplay. (When I say we I mean Boomer and I.)

You're right when you say that the money doesn't come from playing GTO, but also no money is lost - so against tough regulars that's probably the best way to play. Also, by the so-beloved leveling war you get here in the long run too - why waste time and potentially money by trying to level your opponent when you could just skip to the end.

Also, playing this style does NOT exclude any exploitative adjustments you can make. Maybe most of us will adjust a tiny bit less, than the exploit-monkeys, but that doesn't mean we don't exploit the weaker players - we just know the deviation from the standard strategy we have to make.

But most of this argument is pointless, really, cause most regulars have absolutely no clue about GTO or any kind of balanced play in my experience, therefore they cannot build on it, like you suggest. I can easily say that 90+% of the players would be totally surprised if they could get their hands on that "GTO-SHC". If you don't believe me, look at some of the posted hands and evaluations, or some non-Boomer vids.;)

Originally posted by JLeitmotiv
And here will come Madorjan in his shiny horse to tell you that you shouldn't stat-fix your game. And I'll have to agree with him. Moreover, in that case, it will be better if you post here your own values.

Anyway, Sendra, if you want, you can add me on Skype (JLeitmotiv) and we can do a sweat session. I am no pro whatsoever (I'm basically the opposite, just a little fish!) but maybe we can discuss those spots where you are having difficulties

what if I fix my game vs my opponents ranges? Say I know the fellas calling 20% of time and I will adjust my range vs this: 22+,A2s+,K9s+,Q9s+,J9s+,T9s,A8o+,KJo+?

I didn't argue against what you and Boomer have been discussing :), not at all.

Edit: To be fair, I haven't yet read most of the posts in the thread yet. I have been saving it for a rainy day. So what I wrote above is not to be seen i relation to any previous post.

I know what GTO is. It is, as you say, best described as a very very complicated version of a starting hand chart. (You can throw in stats and assumed ranges of opponents if you want to, and you can throw in history versus the opponent, his behavior versus other opponents, their ranges, ad infinitum, but whether you do this or not is up to the definition of what you mean by GTO. If you put in opponent ranges, then you need to incorporate this into the rules of the game itself (or you would, by definition, exploit and be exploitable, hence not GTO). The chart will get longer, but is, in principle, a chart.

Edit: A simpler description (shorter but not better) is that of GTO play being not exploitable period, and take that as a definition. I reality, the GTO play will, as far as I can see, very rarely deviate substantially from basic EV calculations. Take that river situation I used in the previous post. I can't see any rationale for GTO to deviate much from ordinary EV calculations versus an average opponent here. True, in GTO, the goal is not to maximize EV for this bet , but to avoid being exploitable. My guess is that this doen't change the outcome (decision) very much.

What you cannot really throw in is making assumptions about adaptions on part of your opponents. That would really be stretching the terminology past a point of appropriateness. Then it would enter a territory where no GTO has been proved to exist (as far as I know) mathematically.

Apart from these remarks, from what you wrote in the last post, we probably think very much along the same lines. I'll read everything carefully when I have some time :).

An interesting truism about GTO: It is optimal if and only if the opponent plays GTO himself. Corollary: GTO is never optimal, since nobody can play GTO.

/Johan =

Originally posted by YohanN7

An interesting truism about GTO: It is optimal if and only if the opponent plays GTO himself. Corollary: GTO is never optimal, since nobody can play GTO.

/Johan =

I recently added some wisdom to my game and I try to apply different game style. It's called GTFO and its meant to work with my opponents and my pots. It sometimes backfires as I was trying it with AJ vs AQ and then tried the same.. They perhaps had same theory in relation to me and staying in the tournament. Hence GTFO ♠_cry:

Sorry, Yohan, to be a pain in the ass, but your post just proves my point even further.:)

GTO is a 100% static strategy, that makes you at least 0EV versus every possible counterstrategy. You cannot throw in stats or assumed ranges, cause GTO does not care about the opponent's strategy or range or whatever. GTO is an absolute static strategy, that will never lose against your opponent. The definition of GTO is not up to me, it's a really basic game theory definition that I'm not gonna go into. But anyway, the point is GTO is the same whether you play against a loose-passive fish or a maniac.

I reality, the GTO play will, as far as I can see, very rarely deviate substantially from basic EV calculations.

This is where you're awfully wrong. Any purely exploitative strategy will deviate from GTO or balanced plays a lot. Imagine a very simple 1 bet 1 street AKQ game. The GTO here is bet all your As, and alpha amount of Qs. The opponent's GTO-reply is to call with Ks a given amount of the time. However, if the opponent just bets a tiny bit more Qs, your maxEV (=/= optimal) response is to call with all of your Ks, and if the opponent bets a tiny bit less Qs, your maxEV response is to fold all Ks. Therefore that really really really small percentage in his strategy matters 100% in our strategy. So you can see how substantially GTO (that always call x% of Ks) differs from basic EV calculations.

What you cannot really throw in is making assumptions about adaptions on part of your opponents. That would really be stretching the terminology past a point of appropriateness. Then it would enter a territory where no GTO has been proved to exist (as far as I know) mathematically.

Once again, I don't get that part, or if I assume I get it, than you're wrong again. GTO is a strategy you can use in any circumstance coming up. There is no right adaptation against it, and since adaptation in itself is not a strategy, but a change in a strategy, there can be no GTO for that.

An interesting truism about GTO: It is optimal if and only if the opponent plays GTO himself. Corollary: GTO is never optimal, since nobody can play GTO.

Sadly, it's not a truism. It's maxEV only, if your opponent plays GTO himself. MaxEV is not optimal. Optimal means, that there's no succesfull counterstrategy against it. Also, it provides us with the great feature that against a completely unknown strategy, it's EV is >= than the EV of choosing another random strategy. (Not the EV of a random strategy, you might accidentally choose a better EV line, but if you choose randomly, the average EV will never be higher.)

If you wanna hear a more successful player's opinion on it, listen to one of the latest 2p2 podcast with sauce, he says awesome stuff in this matter.

But all in all, no one said you should play GTO, however the more I learn about poker the more I believe in that it is a necessity to know as much as possible about GTO, because that can help you shape your strategy in ways a pure exploitative player will never be able to. I see plenty enough evidence proving it at the tables, but if you look around the forums or videos you can get a feel for it too.

No.

But please see private message madorjan.

Edit: Can you imagine how hard it is not to write a lengthy reply here?

Coach madorjan is 100% correct about strategy and GTO. He is also 100% wrong about me being wrong about GTO because we say precisely, exactly, identically the same thing, only that he apparently doesn't really read my posts before firing away.

/Johan =

1) I'm not a coach anymore. At least not on this site.:)

2) Summarizing my answer to you in public: I agree we write mostly the same thing, I also partially agree that I may have misunderstood some of what you said, however my main point that - it looks like - you're missing throughout the thread still holds: the usefulness of GTO is not in playing vs GTO-bots, but playing vs unknown strategies or sets of strategies, where GTO, or a GTO-based strategy has to be applied, otherwise we're shooting in the dark.

All in all, our seemingly endless battle with Yohan continues privately, if he's up for it, short summaries will come soon.:)

I propose a 24 truce I really need to play a bit (after this post ).

... however my main point that - it looks like - you're missing throughout the thread still holds: the usefulness of GTO is not in playing vs GTO-bots, but playing vs unknown strategies or sets of strategies, where GTO, or a GTO-based strategy has to be applied, otherwise we're shooting in the dark.

Yes. I haven't even touched on that subject, because I think it is obvious. It follows trivially from the definition of GTO. If it was not true, then the GTO wold not be the proper GTO, i.e. A Nash equilibrium of a particular kind. Besides, one showdown later GTO is no longer optimal (only game theory optimal).

I understand what you mean though, and my formulation was less than precise. The following is still true:

An interesting truism about GTO: It is optimal if and only if the opponent plays GTO himself. Corollary: GTO is never optimal, since nobody can play GTO.

But, one needs to add that if the opponents strategy is unknown, then GTO will be the best guess - this not because it is particulary good versus any opponent (it still isn't), but because it is not bad versus any opponent (any other style will be bad versus some opponents). Once we know something about the opponent, GTO is no longer optimal versus him, except for the idealized case where he is a GTO robot

Let me point out a situation where GTO may be quite bad. Three handed. One fish, one GTO robot, and you. You, of course, exploit the fish. You, as well, exploit that the GTO robot is not adjusting properly to the fish (assuming here the simple version of GTO where ranges are a nono.) or to your adjustments to the fish. Suppose you have position on the fish. Now you have a gold mine thanks to the GTO player. If the cards fall off without surprises, then the GTO will get X% of the fish's money, and you will get Y%. X will be above zero (by definition of GTO), but just barely above zero if you play your cards well.

I don't know how to calculate GTO, but I'd still be very surprised if a GTO decision often differs much from simple EV calculations and common sense. Iv'e heard about J high calldowns and stuff like that, but the more mundane river spots, medium hands, medium pot versus a predictable opponent can't be all that surprising, can they?

As far as charts getting longer if we allow ranges in (modified FL) GTO. Probaly not longer, but the calculatiions will probably take a billion trillions times longer

A final word about terminology. I think it is best to write GTO or game theory optimal if that is what one means. We can't ruin the language by having optimal defaulting to game theory optimal. "Optimal" can very well have the undefined, yet intuitively correct linguistic meaning of being perfect in all aspects.

/Johan =

Damn, I can't disagree with anything in that post, please rewrite!

It's also an interesting point that we should treat an unknown opponent as a GTO-player. Will have to think about it a bit, but that sounds a good abstraction I haven't thought of.

So yeah, truce, and GL at the tables.

Originally posted by madorjan
Damn, I can't disagree with anything in that post, please rewrite!

It's also an interesting point that we should treat an unknown opponent as a GTO-player. Will have to think about it a bit, but that sounds a good abstraction I haven't thought of.

So yeah, truce, and GL at the tables.

Dunno man. I play lower stakes for about 2 months now. I don't count previous lenghty experience with poker as lot has changed.

Ppl play about the same so you can exactly figure out what they doing. Does it count as GTO as well if you learn how to exploit them little motherfathers? ♠_evil: