I have to agree that KJs is a pretty sweet hand, either to open, 3-bet, or call a 3-bet with in position, and it may even be more profitable than A5s overall. I was more concerned with showing that A5s does better as a 3-bet against the hands that typically call than suited connectors like 76s do. I don't think the latter makes enough value-hands for it to be more profitable as a 3-bet than A5s.
Just generally speaking, I think players way overestimate how many flushes, straights and two pairs/trips they make (and how often they get paid) with small suited connectors.

Having some suited connectors in our 3betting range is usually good for our range and protects our range from some exploitative 4bet sizings.

So you're not really wrong that the 3-bettor is 3-betting some pretty weak hands because the original raiser will sometimes fold, but you have to look at the overall strength of each range rather than an individual hand.

Say I open in the button and you 3-bet me in the big blind to 9BB. I 4-bet you to 21BB and you call.

Whose range is weaker here? The guy who hasn't stopped raising (me) or the guy who said "Hmmm.... well, I'm risking 12BB to win 30.5BB, so I guess I'll call." Your range will be a lot weaker than mine here and you'll be OOP, so you probably don't need to defend enough on many flops to prevent me from being able to profitably bet any two cards.

Now, what you said earlier is more of an argument for why I am 4-betting you with some Ac5s (or whatever you prefer to 4-bet "bluff") stuff. I'm thinking to myself "Well, maybe this Arjen dude will fold, and even if he does call my Ac5s has a lot of equity against his 99/KQs stuff, and maybe he'll call me with AQ and we'll just both miss and I can make him fold on the flop." So your arguments are more of why I should 4-bet "bluff" some hands, but despite the fact that I 4-bet bluff some hands it doesn't change the fact that my 4-betting range is overall a lot stronger than your 4-bet calling range.

Most of it is on CardRunners in videos or just in posts I've made here or on CardRunners throughout the last two years. It's not all assembled in one place.

I like 3-betting both when the situation calls for it, but A5s is almost always better than 65s.

A5s makes better flushes and better pairs as well as has a bit of a removal effect (though not that big of deal). 65s makes 3 more straight combos than A5s (A5s can still make 1).

Pairs are wayyyyyyyyy more likely to happen than straights and I'd wayyyyyyyyyyy rather have top pair than a pair of sixes. Also two pair with A5s seems better to me than two pair with 65s as now the opponent may have a AQ-AT hand that called our 3-bet and we'll get paid. And last but not least the nut flush vs 6 high flush is pretty relevant a lot of the time.

The problem you likely have is 65s is a lot easier and more comfortable to play than A5s and if it's a lot easier to remember the times you flopped top pair and didn't know what to do than it is to remember the times you flopped top pair and comfortably won a small or medium sized pot. In general I think there's a huge bias towards wanting to play hands that are difficult to play (i.e. have a non-robust equity component, like A5s does when it makes top pair) even if they are the more +EV hands. Suited connectors are just easier to play and everyone likes it when they get to make the opponent make hard decisions rather than make them themselves, even if playing hands that require hard decisions are often the more +EV hands.

Preflop:

The book is least likely to semibluff 3b oop with dominated ace high hands and it or i add a5s to that. I wont 3b small pairs either as i have nothing after the flop. A dominated ace high oop is a bit less likely to play or 4b if it would do that, than when in position, but not sure i care about that as much, though calling 3bs is generally loose in these theories and need to be adjusted of course like if the utg opens like the button or the button opens like the cutoff, and so on.

To me it is somewhat too loose to 3b vs. early open raises w a5s, so i dont do those, though i will consider with aj oop also but i am generally adjusting to opponent then. Plo has made stuffs like these extra clear for me and i see no reason to do otherwise in nlh either, a dominated ace, suited or not, is just a dominated ace.

Building the pot 200bb deep is known with some excellent implied hands in any game, but ur free to do so in nlh or even in lhe, but i pass with a5s that other than the flush draw is just an ordinary suited connector in value and to me it is like jt offsuited, and sure, i can hope i flop two pair vs. top pair top kicker, or a strong draw, and jt was once considered a similar hand in nlh, for a moment.

A5s is generally not even cold called preflop, that obviously isnt right deep enough like on the button. I view it just like any suited connector with reasonable power, no matter what the game is. Omaha8 has once considered some a45x suited even as a strong 3b hand, but it is similarly just an open raise and a possible cold call and potentially a 3b in these best of cases, with a good x.

I picked to 3b axs in nlh only when it is just a step below cold call, meaning it is like a9s, and on the button vs. cutoff case it is a7s and a5s, as it is pretty much in the small blind vs. the button open raise where i use about the same hands as it is out of position, just rate to 3b them all like the current recommendations. In the big blind i might do the same but maybe not vs. small blind open raise that i prefer to float with small and medium suited aces and will exploit with the unsuited ones if necessary and possible as they never flop anything of value and usually miss, and in the small blind vs. a loose stealer the just below normal standards are of course good enough, including medium unsuited aces though oop.

The sb opener has been also too loose in many cases, over 50%, as open raise as well as a limp. And after making these especially exploitive raises and 3bs, no need to wait but okay flop to cbet in these cases, like in some other cases, especially in position and when exploiting. Well known poker strategy, though not gto vs. gto but looks gto enough to me vs. weakness.

High cards like aj, at, kq, kj in the book work better in position as 3b bluff, though i never liked them as they are dominated, but i changed my mind though no autoplays, and i add the suited similars also, and will of course just cold call with standard suited conectors in position, mostly adding 2 gapper or anything if i think he folds too often. Though no 3b any of them if it brings no good, and so selectively the high cards here work well as semibluffs when they can be even legit cold calls vs. looser raisers, almost like when in the blinds, and no one disagrees with that even in plo in position, though it plo is a bit too tight oop as fashion is against it, the stealer not folding easily, where i also use all these when fit, but only then or i cold call them as extra hands, that in similar cases and reasons could be done in nlh also, as i do there also when exploiting.

If kq 3b oop and the opponent floats more often, as rates the case, kq has some similar semi barreling value as suited connectors have and can beat a medium pair with one pair. Maybe one can pick it to balance ones barreling with though i have picked just missed draws for it normally, but not sure i have enough of them in tight situations until i have enough suited connectors, as i usually have.

Preflop kq etc has some removal value even over weak ax. Aj having a problem with the ace and the jack is lower than a queen, but jack has the nut kicker, but a reasonable idea to mainly exploit with it and At oop vs. late or comparable open raises and prefer the best suited connectors one doesnt as much like calling oop with, and blowing out hands that dominate them with a kicker. Though not that i am a big fan for such plays with weak hand and will more or less select when i do it and with what hands, oop being more a spot for playing gap concept and implied hands in small pots, agree with it or not, and it took some time for me to accept even the hands like qj doing 3b vs. button open raise in the bb but figured the better kickers are not the only hands calling it and if he calls the 3b or not, because he has, when he has, such a loose range like 40%.

I read the preflop talks are done, but just wanted to add to this a5s case.

Postflop:

I also have this system for postflop play pretty much and understood in my memory, just with minor simplified it, using just missed draws or and semidraws to balance on the river, and some selected pairs to raise bluff the flop, the turn when they get more as i prefer some showdown value especially in position. All balanced, and added the river bluffs to cases where not in this book, like when it is raised on the flop or turn.

Less than a pair or a real draw on the flop i raise and check raise only when i see or feel it to be good, sometimes even 3 ways, and gto vs. gto they would always be the case, hu, but i have them as an option only, eg. if i see the opponent seems to be on steal on the flop, i need just something to resteal it and hu that can be the case often till he adjusts.

Not too sure oop if he sticks to overcard or two, ace high, any pair, so it is pretty easy to see why i am so selective and that there is little point to balance in many cases. The other extreme next to never cbets, so i am not going to raise his nuts with the gto type range. Some of them delay their miss cbets till turn. Some of them are too strong preflop already. Some of them get too tricky for my raise hand. Some stick around. It is sometimes no good even to cbet, or one needs to bet the turn also and they can be like gto balanced for it, but i suppose it is gto then until he always floats in case i will always barrel and trap in cases.

I didnt like the flop raises with one pair, it being not usual in other forms, but changed my mind after i picked to bet some of them, call some of them, some as stop calls only, similar to plo in much of these, and pick to raise and check raise select pairs, when it seems better than a call or a fold. From plo i am familiar with that and with all the relatively minor changes i made to this system, so i can feel comfortable, remember and understand it and adjust per player and situation.

The draws are not off as flop raises in this book but mainly ace high flush draws or double draws with no pair, and i pretty much agree with that, or there was some small flush draws with an inside straight draw that also raised while elsewhere they didnt and a double draw didnt check raise but bet somewhere, the opposite.

I will play a draw and any pair selectively more aggro, with a raise, if i think or feel i better do so, and i added some select straight draw also in rainbow flops but only some, like with the pairs, maybe one combo each rather than 3 or 4 combos, that i also, some or all, might exploit in some situations but not usually. Poor flops this book plays raise passively until many draws can be there in case it might be more or less mixed strategy.

This book gave me the system. Had such for plo and limit holdem also and for some other games though less in those cases as no gto known until one adds there some more. I am using a system as a point from where i balance my play this way and that way based on the opponent, and that is the other half of poker, but i am happy to my nlh technique now as before this it was based more on raw power and adjusting, the old way, that though still is just as important multiway and when exploiting.

My changes also made it more easily balanced when the ranges are looser, and there are some sample hands that have eg. q7s, a medium kicker. Remembering what to balance with can be difficult if using to it too many different kind of hands, them being more for special cases or vary rather than remember them. Though those hands used in this books system are mostly not without a pattern but many are. But whatever, i just made some changes for myself, and additionally i will now focus adjusting to the opponent, that still needs more study and experience.

I have a question that is not related to the theory, but more how to use it in practical play.

Example:

You raise in MP with ATo and the button calls you.

flop comes J95 with 2 spades


now you would have to know:

ok so ATo is now in the top x% of my range, and I will do X action with it.


on the flop this seems doable with a lot of experience, because your starting range is pretty well defined (from preflop action)


But on the turn & river it seems much more complicated, because you only continue with a % of your flop (and turn) range.

Do you actually think at each point which combos are in your range... or do you just have a "feel" for it from lot of experience ?

For me, the turn and river are usually pretty easy to play if money has gone in the pot on earlier streets, because ranges get smaller as the pot gets bigger. (Playing the flop in BvB is much harder, for example, because there are so many more combos that need to be put in the right bucket).
With study and experience, your whole plan can be made right on the flop. With some hands, it's pretty easy to think things like:
"I'm going for three streets of value, because I have a monster", or
"I'll be one and done with this airball, because if villain calls on this flop, I basically have no equity", or
"I'm probably triple-barrelling this gutshot, because it's never winning at showdown and I need some bluffs to balance the value hands I have on this flop", or
"My hand is mid-strength on this board and will likely stay that way on most runouts, so I'm gonna bluff-catch 2 streets, and possibly three, depending on villain's bet-sizes".

It's only really on the dynamic boards (e.g. low two-tone flops) where your range strength or equity will radically change on the turn/river and you have to rapidly calculate which range benefits the most. On most boards, your range strength should be fairly clear on the flop, so if you get your frequencies correct there, (and you had a solid pre-flop range to start with) you shouldn't have too many problems later on.
One tip is to always think about the turn before you make your flop decision. e.g. If you're thinking of c-betting a weak draw, ask yourself "Are there many turn cards that I'll be able to barrel because they increase my outs, hit my range or are scary for villain?" If there aren't many good turn cards for you, then betting the flop at a high frequency often isn't a good idea.

To use your AT on J95 example, you might be more inclined to c-bet if you have the backdoor NFD to go with your backdoor OESD, since you could barrel all spades, eights and queens, but you might also decide that you have few value-bets in this spot, so you should balance with fewer bluffs (which would typically be better draws like FDs and OESDs), because villain's range is quite strong. This could ultimately mean that ATo goes into your check-folding range, so that your continuance range has more robust equity. It's a lot easier to play the turn with ace high if you folded it on the flop.

Equilibria in sequential games like poker can be calculated with a Linear Program (LP). OpenOffice has a linear solver which can only address small problems, but is enough for solving a river situation with short ranges as in Sample Hand No.2: UTG vs BT.

I was using this hand as an example to learn about the information we can get with these solutions and I got to something I found really interesting. Definite bet sizes are not indicated in the book, so I supposed that after the preflop open raise, flop bet and turn bet the pot is 41,5bb. We have a 30bb bet and a 80bb raise for the river. I suppose there are commercial poker programs that can solve this problems, but I don't use them.

Solving the LP we get that the Game Value in said situation is: +17,04 bb for the UTG player and +24,46 bb for the BT. The strategy the book proposes for the UTG player has +15,77 bb of EV against the calculated equilibrium BT strategy.

Sorry for getting too mathematical in this post, but I need it to get to the point, if someone is interested we can discuss further the math. To know which actions are 'mistakes' against an equilibrium strategy we can check the non-zero 'slack variables' and I found there is only one action I can consider a mistake and it is check-raising bluff with AQ. This resulted in a 6.7 bb mistake. When changing this action to check-fold the proposed strategy gets the expected +17,04 bb EV.

So I wondered which hands we could use as bluffs here, but every single hand except TT (the effective nuts) has a negative expectation by check-raising! As I could not understand this I checked the strategy for the BT player and I got that it never bet-folds after UTG checks.

The point I find interesting is that there are non-obvious situations where the optimal play never folds so that balancing our range with bluff raises is a costly mistake. I wish I were able to find those spots without having to solve a LP!

My thought process would be more of this:

"Ok, if I bet, what do I want it to accomplish? I want to #1) Deny equity and #2) Make the pot bigger in case I win at showdown. Does betting ATo fit that bill? Well, if I have the As, it probably does. Now I can runner-runner the nut flush or runner-runner a straight, both of which make very good hands. Ok, so I guess I'll bet. If I have no spade, my hand seems too weak to bet because the button called pre-flop and position is quite valubale here"

I probably don't run a single number through my head. If I wanted to get fancy and check-raise bluff then sure, I'll count how many "value combos" I could/would check-raise with.

Trying to "rank" this hand on the flop doesn't really seem possible or all that useful. It is useful to have an overall sense of how strong your range is though, and that will come in practice.

Cool, thanks for posting this. I definitely wouldn't recommend trying to "balance" your range without using computer software now as there's a lot of useful software now available.

Hi Matthew,

I understand the concept of using different bet sizes (with different ranges for each) post-flop depending on who has a polarized range, more equity of the flop, capped range etc.

ie. J96ssh flop, villian x/c's oop bb vs co, turn comes 2c.

Clearly here with our weaker values we should bet one size (say 2/3-3/4) and with our stronger values have an overbetting range (as most likely ahead of nearly all of villians x/c flop range and won't be raised).

My question is: In practice, what's the best way to distribute our bluffs/draws into these two sizing ranges (ie. overbet with draws which retain their equity more vs villians stronger hands and likely stronger draws) ?

Do you yourself have a method for this when playing live.

(+ applying this on the river as well in a spot with a normal size bet + an overbet range) (would guess that we should bluff overbet with our best blocker hands ??, except that our overbet bluff needs a smaller % of folds to be profitable)

Thanks

Interesting to note:

Poker snowie's website:

Design weaknesses


Allowing only one bet size for the whole range of cards in a specific situation

Basically all professional high stakes players only use one bet size in a given situation.

(multiple bet sizings)...approach is highly complex and not practical for PokerSnowie nor humans.

Overbet bluffs that have equity to make an valuebet vs the continuing range against OB and also blockerbluffs that are good vs tight continuing range.

I like PokerSnowie a lot, but most people I know (some play high stakes but of course most don't) do use multiple bet sizing. Also, being restricted to only 0.5 PSB or 1 PSB or 2 PSB is pretty limiting.

But still, it's a good program for what it is and I like it a lot.

You have to put draws in both ranges. I know that's a lame response but poker has tons and tons of mixed strats and I think if I write more it will just be confusing and not helpful at all.

This is a good way to look at it too, but I think in reality all "bluffs" in this spot will likely have some really robust equity (most will be flush draws or have some sort of straight draw).

So in practice, just put a bluff hand (when say an overbet spot occurs on the turn) in whichever range (for sizing) you feel like, going by feel to remember to use one or the other sizing(s) a proportion of the time?

Hi,

In your book you state that once you call a PFR in position and it gets checked to you, you can make a profitable bet with ATC. How come you didn't use your standard method of calculating the mdf for defending vs probe bets, and just automatically assumed that as the PFR, we can let villain show an auto profit.

You stated that we "risk" a lot when we call IP and (might get squeezed, PFR might c-bet), and that the EV of these probe bets are very low, only 1ish BB. While I understand that the EV is much lower than what we put in preflop and thus doesn't allow us to call a wide range pre, there is also the fact that our IP calling range is designed to play well vs the PFR's range and we have squeeze protection if the PFR is in EP, so how did you just come to qualify it as enough "risk"?

Your other examples, such as 3bet/bet/bet/check river are much more obvious because
1. you actually do put a lot of money in the pot
2. after taking that line we'll have a very polarized range a lot of the time (on rivers that don't greatly change the texture) so it's correct in theory to have a very weak check back range of turn bluffs, which isn't problematic if villain can auto-profit since he'll want to check back his condensed range anyway

For calling IP though, is there any sort of evidence that it's enough risk?

Hi All,

Apologies to Matt for posting another pre-flop question, but I hope there'll be lots of people who can answer this quickly.

4-bet bluffing range UTG vs IP 3-bet is AA, 98s, 87s, 76s. Thats 6 value hands and 12 bluff hands. Yet we need to call 50-60% of 5 bets and we only have 33% value hands. Is this just a simple mistake or hold we really be making some hairy calls with 98s?

I would think about the money instead

It depends on how good you are. That's kind of like saying "So basically, we should just raise our good hands on the flop when we're worried we might get outdrawn on the turn?"

Yeah, that response pretty much works for me, but if you're really, really good you might care more about small edges. For the most part when I'm playing you'll see me overbet bluff all types of hands in all types of spots where I think it's good to overbet. I'll rarely bluff a total garbage hand though, I usually at least need some equity to bluff on the flop or turn.

Snowie check-folds a lot as the PFR in many spots, enough that the cold caller can profitably bet ATC. You're right though I was mostly just guessing when I wrote Applications though, but it turned out the guess for the most part still looks pretty good (optimal poker tends to involve checking a lot out of position and check-folding enough to allow the player IP to profitably bet ATC). But again, you can't solve for this stuff and it really just comes down to the fact that the OOP player's checking range is pretty weak relative to the IP betting range.

I also don't know what a "probe" bet is, but hopefully that answered your question.

AA has so much equity that you can bluff lots of hands for every bluff. Unlike that poser of a "value 3-bet" AK.

This is kind of the problem with modeling pre-flop and why in retrospect I don't think it worked all that well. There's just too many little nuisances that it's overall not that helpful to have ratios for pre-flop (whereas post flop I think they're useful in a lot of spots when a betting or raising range is very polarized).

Yep, thanks!

Also, probe bet = betting vs missed cbet.

Hey there,

Is the front page updated with all the mistakes? I notice it hasn't been updated in a while. I'm just wondering if there's a single spot with all the specific errors/updates and any general updates or do we have to read through the whole thread?

Cheers