looking at the recommended hand chart - the BTN can profitably 5bet 100% of his 3betting range (KK+, AKs, AJo, KQo, A5s-A4s) against UTG's 4bet (AA, 98s, 87s, 76s). can you explain why that's not a big problem for us as the UTG player?

Sauce?

I'd recommend reading the other posts in this thread if you haven't already, but it's absolutely impossible to solve pre-flop. That hand chart was just me trying to show you how could begin to design ranges that you continuously tweak. There's no way to get everything to balance perfectly, especially since pre-flop will use a ton of mixed strategies. So if you are finding the BTN can profitably 5-bet any hands he's 3-bet, then UTG should 4-bet a bit less aggressively.

Also as previously mentioned I'd now emphasize 4-bet "bluffing" high equity hands (AJo, KQo, etc) rather than hands with robust equity, and those hands also have a favorable removal effect.

So, what you´re basically saying is that you still should be defending 50% of your range when the 3rd flush card on the river hits and your opponent bets pot?

No, because I don't know each player's range and who is in position. If my opponent's range is much stronger than mine then it's very unlikely I'm defending 50%. You can use previous action to determine whose range is stronger and you of course should do this as every good player does, but ultimately only his range and my range right now (as well as a few other factors, like stack depth and position) determine how aggressively I should call when facing a bet.

I talk a lot in the book about how if someone continuation bets the flop OOP then checks the turn, he's probably not defending enough of his turn checks to prevent the player in position from making an immediate profit on his bet. One way to look at this is "Well, the player in position had to risk money on the flop to maybe get to make a profitable turn bet, so just because he can profitably bet the turn doesn't mean the out of position player is getting exploited." That's good, and you should look at it this way. But you should also ultimately realize the reason the OOP player is check-folding so much once he checks on the turn on certain board textures is because his range is quite weak and he doesn't think calling is +EV with enough hands to prevent the IP player from profitably betting ATC.

Is this book FR or 6 max?

@ Spladle

The EV of Hero's BTN range is a function of the EV of the Villain's range. If Villain can 3bet 100% of his hands profitably, then obviously we ought not open any hand from the BTN that will not either call or re-raise a 3bet.

Obv, I understand them are linked, but what I don't understand is the relationship.



This is wrong. Even if 3-betting will show a profit, it may be the case that calling will show a greater profit. And even if Villain wins by 3betting less than 68%, it's still possible for 3betting to be better than folding.

We are talking only about preflop, def you are right if you consider postflop



What do the numbers/variables here represent?

-2,5 = profit when we open and fold to a 3bet
x = % of the times we fold to a 3bet
12 = "profit" if we defend, well this is not a profit, maybe I'm missing this point
1-x = % of times we call a 3bet



@ Matthew

First of all poker is a zero sum game?
Because, I'm pretty sure I'm missing this point.

Let's take a look at pg 69.
When we open to 2.5 big blinds and our opponent 3-bets from the big blind to 9.5 big blinds, he risks 8.5 big blinds to win4 big blinds. And as we’ve already shown, this means the bigblind’s 3-bet cannot be allowed to succeed more than 68 percentof the time.

4 (x) - 8,5 (1-x)
x= 0,68

so if villain wins more than 68%, he's showing a profit (assuming Hero is folding to a 3bet)
this percentage is useful to Hero? Knowing Villain will show a profit if Hero is folding more than 68% what implies? 1-0,68 = times that hero should defend?

Hero's EV on the BTN is the inverse of the blinds' EV.

I'm confident that you are confused about something, but I'm not entirely sure what it is. Language barrier may be getting in the way here. Let me backtrack a bit and try to respond again.

This is wrong. Simply knowing how often the BTN will fold to a 3bet does not allow us to calculate the EV of a 3bet, nor is it provable that there exists any hand which is indifferent between 3betting and folding.

If we profit 12bb by defending, this is a strong sign that someone is doing something very wrong. Notice that if you lower the assumed EV of defending, you're forced to defend more often.

Knowing that villain will be able to profitably 3bet with two pieces of toilet paper if we fold more than 68% of our opens to a 3bet demonstrates that we ought not fold more than 68% of our opens to a 3bet.

All the hand examples are 6-max, all the concepts should work for full ring.

Thanks Spladle, EmptyPromises, and anyone else who is answering the questions of others in the thread. Most of the responses have been great and it's helpful when people get to hear things explained from multiple different people.

Hey Matt,

I've seen a ton of chatter lately about pokersnowie, a program that claims to have a good approximation to GTO play, and offers a training product based on that. I'd love to know if you've tried it yet and if you think the claims are legit, ie, do its recommendations pass the smell taste of something that is actually playing close to GTO?

In before censorship.

why would that be censored?

PM´ed.

No, the claims are not legit; its recommendations do not pass the smell test.

Yeah ive been playing around with there preflop advisor at work and when facing a 3bet from BB after opening SB it recommends 4betting T7o 72% of the time? Doesn't recommend any 5bet bluffing when facing a 4bet unless you would consider jamming TT, JJ a bluff.

i'm starting to think it is legit, actually. i listened to an interview with the CEO, it is indeed the same company behind backgammon snowie, and he does not come across like a scammer or a liar. since every very strong limit bot has made some very surprising plays, i don't think you can use isolated instances of "bad" advice to dismiss it, since those plays could be correct but unintuitive (ie, against the standard advice of human experts, but not provably incorrect); or incorrect but rare.

Perhaps we have different understandings of what "legit" means in this context. I didn't mean to insinuate that the company's CEO is a scammer or a liar. He's probably just using a very loose definition of "a good approximation to GTO play."

True, but that's not what's happening here; many of its recommendations are simply wrong/bad. These mistakes are an inevitable artifact of the game not having actually been solved. They've done something interesting/worthwhile, and further progress on this front will surely continue for years to come, but they haven't made any sort of breakthrough or revolutionary insight. Humans are still > computers at NLH.

don't get me wrong, i am still reserving skepticism

as for the above, how can you be sure? a lot of people said the same about sonia and polaris. until a high stakes experts plays the thing over 50k or more hands (or is beating it massively before that number is reached), how do you know for certain that it's claims aren't valid?

If I gave the impression that I was certain about anything, I apologize. I'm not.

If PokerSnowie's creator(s) thought it could beat a talented human at NLH, it would be in their financial interest to prove it. That they haven't sought to arrange such a match is strong evidence that they know the outcome would be unfavorable to them.

well given that humans are still better than snowie at backgammon which has perfect information than i don't think I think we can induce that humans will be better at a more complicated game.

i was under the impression that was not the case with bg snowie (ie, i thought it was stronger than the best humans), but only because i'd heard that mentioned in a few threads so not exactly the most reliable.

as to your more general point, most lhe experts now agree that the best hu bots are stronger than the best humans, and that is ofc a game of imperfect information as well.

On P 47-48, chap "The EV of 4-betting of folding", it's written :
"Note the following tables show how much we expect to win or lose overall for the hand when we 4-bet. So since folding to a 3-bet results in us losing 3.5 bb, it's still more profitable to 4-bet with the intention of calling a 5-bet provided our total expected value is greater than -3.5bb .

I don't understand :
Folding to a 3-bet results an EV equal to zero, right ? I understand that i'm loosing 3.5bb but since my 3.5bb are in the pot, i don't care about it.
So, folding to a 3-bet (EV=0) is more profitable than 4-betting and calling a 5-bet that provide an EV > 0 (and not > -3.5bb).

Please, tell me what's wrong ! Thx !

I would say : So, since folding to a 3-bet results EV=0 is more profitable 4-betting and calling a 5-bet that provide an EV > 0 (and not > -3.5bb).

Please, tell me what's wrong ! Thx !

There is actually nothing wrong. It just depends on the choice of what is zero ev. This is an arbitrary choice when all you care about is ev difference.

Matt is considering total ev so he chose 0 to be the start of the hand. You chose zero to be after raising.

Just getting started in the book. I ran across a sentence on page 8, and I'd like to have a better understanding of the idea.

Discussing 89s and A9o. "...when they don't make strong hands and instead are used as bluffs, they will almost always make our opponent fold better hands so little equity is wasted."

"Wasting equity" I'm not appreciating the meaning here. Can someone explain further?

Thanks for any help.