Quote:
Originally Posted by FindNameHere
I havent looked much into the GTO stuff, seems like you have spent a great deal of time studying it, and not that im trying to rob you of all your hours spent learning, but would you be able or willing to give me a few short outlines of spots where GTO comes up OFTEN
Well, there's a theoretically optimal way to play your range
in every spot, whether we're talking about which hands to open UTG, or which hands make the most sense as check-raise bluffs on the river. Most regs are pretty solid pre-flop, but they can be quite unbalanced/exploitable in various spots post-flop. (Not that being exploitable/exploitative post-flop is a huge problem against weak opposition. Indeed, you make more money by exploiting their "leaks".)
Quote:
Originally Posted by FindNameHere
Like single raised pot in position, with KK and flop with Axx , or same but we 3bet CO from BTN with KK and flops Axx, is there a standard line for spots that come up often ?
One of the most curious things about optimal play is that mixed strategies are extremely common, by which I mean that - if you play your range optimally - it's actually correct to sometimes bet and sometimes check a particular combo. It's
usually the case that KK is a clear check on Axx, but in some spots (where you're betting a large proportion of your range, mostly for a small size), it's apparently "GTO" to bet some hands that don't seem to make sense as value-bets or bluffs. It's easy on a human level to divide your range into "obvious value-bets", "obvious bluffs" and "obvious checks", but true GTO is much more complicated, because to prevent exploitation you need to "randomly" put some hands into two categories at once. (Quick example: If BTN opens and you have AQo in the BB, it's fine to 3-bet and it's fine to call. The GTO strat would 3-bet at X% frequency and call at Y% and a GTO bot would get those frequencies perfect, such that the 3-betting range and the calling range are both maximally profitable. As a human, it's easier to just say "I always 3-bet AQ here", or "I have that in my flatting range 100% of the time", as you don't need to "mix up your play" with every combo that requires a mixed strategy in theory, as an opponent will never get a large enough sample size to be able to attack the way you play one specific hand.
FWIW, to this day, I don't think I've ever c-bet KK on Axx, but it certainly crosses my mind to do so sometimes, because it's definitely (according to solvers) part of the optimal strategy at some frequency in some spots.
A more human way to approximate a GTO style post-flop is to categorize your hands into groups like I mentioned above. e.g. If you open on the BTN and get flatted by the BB, then on the flop you can divide your range into:
1. Strong value hands that want to bet multiple streets.
2. Mid-strength hands that want to keep the pot small.
3. Weak made hands that could benefit from protection.
4. Hands that make sense as (semi-)bluffs because they have robust equity even against villain's continuance range.
5. Total air that has very little chance of winning, no matter what action you take.
When you've got your hands categorized like that, it's fairly simple to think "I'll bet the value hands for value, the weak made hands for protection, and the semi-bluffs as bluffs, and I'll check the mid-strength hands for their SDV or bluff-catching ability, and I'll check the air and plan to fold unless I get a decent opportunity on a later street."
A more important idea is that whatever action you take needs to be roughly balanced, such that it's not completely obvious when you're strong or weak. So when you're value-betting, you need to have the right amount of bluffs in the same spot, and when you're checking your air you also need to check some decent bluffcatchers, so that you can defend against villain's turn bluffs. (This is why you should nearly always have some top pairs in your check back range.). You also need to check back some of your draws, just so you can have flushes/straights in your turn range whether you c-bet the flop or checked.
When you're OOP, it's more complicated because now you have check-raising as an additional option. (i.e. If you're in position with the nuts, you almost always bet it. That's simple. When you're OOP you can bet or check, and it can be hard to work out which line is more profitable, because checking might lead to missing a whole street of value/protection when villain checks back, but it also gives you the potential to make the pot much bigger if villain stabs IP.)
What people used to do is think "C-betting is instantly profitable, so I'll c-bet 90% of my range". Nowadays, they are thinking things like "This combo makes more money as a c-bet, that hand does slightly better by checking back" or "I'm gonna get check-raised a lot on this board, so I'll bet at a lower frequency" or "Villain doesn't have many strong hands on this board, so I'm gonna bet huge and put him in a cage."
I could go into the reasons why some hands that seem "natural" c-bets to beginners would actually make more money as checks, but this is BQ. There's loads of discussion about GTO-style play in the theory forum and also in the books/publications forum threads about books by Janda et al.
Quote:
Originally Posted by FindNameHere
Your example about c/r to get more money in with Value hands since GTO says we on average get only 50bb into the pot (assuming 100bb) then I understand wanting more money in the pot but assuming he has a hand that will call a bet or a c/r but never continue past a turn that crushes his calling range, does that not turn our c/r into I hope I take the pot down now, or is it always with the intent to get it in on the flop?
Again it's situational, but taking down the pot (and making villain fold all his equity) often has a higher EV than getting called. e.g. If you have top set OOP and villain has random air, he's never calling if you c-bet, but if you check and he takes a stab, it doesn't matter that he folds to a check-raise. You still got more money out of him than you would by betting. Janda's written a lot about the concept of "preventing equity realization" and "not wanting action" in his books. Obviously with the stone cold nuts you want to play for stacks, but with hands like 2 pairs, you'd frequently prefer to win the pot before villain sucks out with a straight or backdoor flush, or a set or whatever, so you'll try to pick the line that achieves it best. (Pro-tip: 2 pairs is most commonly a bet, not a check-raise, since it is more vulnerable to a free card than a set is, since a set can "easily" re-suck out with a boat vs a flush/straight, whereas 2 pairs has very few outs once villain overtakes you on the turn).
Quote:
Originally Posted by FindNameHere
how close were we? eg, 2+2er mostly advocated TAG/LAG , 3bet or fold from the blinds, iso limpers wide etc, do they come up in GTO , or does GTO say something completely different ?
It's hard to say how close to GTO today's players are. Players are much much better today than they were 5 years ago, especially pre-flop (people have learned to fold pre, and they also 3-bet/4-bet more often, just like GTO suggests) but in the micros especially everyone is still pretty terrible post-flop. e.g. They don't bluff anywhere near as often as GTO does, they are overly value-heavy when aggressive, they don't check-raise out of the BB at anything near GTO frequencies, they c-bet OOP way too often and for sizes that are too large etc.
Depending on which site and stake level and format you play, the games might not have changed much in the past few years, so you don't really need to think much about what the theoretically optimal play is, and can just stick with ABC value-heavy poker. Zoom cashgame players on Stars are much more solid than random low stakes tourney players on Unibet for instance, so should be treated somewhat differently if you want to crush them.
Last edited by ArtyMcFly; 11-23-2017 at 08:39 AM.
Reason: typos, additions