Put it this way: if you set up a game where SB opens some range (any range) and is forced to c-bet 100% and everyone plays perfect GTO poker under the constraints of the game, the solution ends up being that BB never 3-bets preflop. And the BB edge in this game is way bigger than in a standard SB vs BB game.

equilibrium is actually 100% cbet from sb on all board textures vs a gto bb.

soooo if bb flats everything, the new solution for sb is to start donk checking. <3

SB raises, bb calls 654r sb still fires 100% if bb plays correctly preflop. Surprised me first time i saw it wasn't donk checking, I used to missplay this spot often vs good players by donk checking flop too much.

I much prefer players who either
1) fold too much in the bb (lol ever) or
2) don't 3bet preflop

It allows me to use all that hard work I put into developing balanced donk xr15/xc65/xf20 to be a fker in all spots =D


Side Note: I have played against many people who's BB calling ranges are VASTLY ahead of my opening range.. lol. Only defending 30-50% and never 3betting. Cbet 0% ftw. They always seem confused when I donk xr30% hu oop on the flop and never cbet. Have I mentioned I love strategy? =D Actually finishing up my model for NLHE atm, onward to PLO soon

I really need to work on my blind play. Good stuff in this thread guys. This tagfish appreciates it!

I'm pretty sure you can prove this false with fairly basic logic. The gto SB strategy shouldn't be exploitable by BB calling 100% preflop.

I believe we disagree on the use of the word exploitable. Any deviation from fundamental gto range play is, by definition, exploitative. A BB deviation away from the GTO 20% 3bet to flatting 100% would absolutely cause a change in strategy from the SB opener on diff flop textures. Because of the BB's (small) mistake, SB's new model for the game will cause less cbets.

If SB c-bets 100% and doesn't adjust then BB destroys SB by flatting 100% compared to having a 3-betting range. So from a GTO perspective, either SB's 100% c-bet is wrong or BB's >0% 3-bet is wrong. And I'm pretty sure it's the 100% c-bet.

I like the post about calling with some very strong hands pf. In 20 years of playing limit holdem and being the most consistent winner in every regular game I played I definitely play sb more passively than most and try to play bb contrary to my general table image when heads up. I don't mind if players rob my small blind all day long unless im playing short handed obv. I will almost always get extra bets in the hands i win and come out well ahead. If I have one criticism of people who believe limit holdem is "solved" and that there is always an opimal play (and i hv more than one criticism) it is that the absolute best poker players are aware of their image how their opponents view them and adapt so much better than those convinced there is a formula that is always optimal.

is there like a standard link to post for people who haven't heard of game theory or otherwise think john nash is a moron?

That's a pretty strong statement. The top 3-handed LHE bots in the most recent AI competition (2014 was last year with LHE) do not cbet 100% from small blind. But I'm not sure that's even that informative. Recently, two U of A guys described publicly how far they are from an equilibrium solution to multi handed poker. They don't even really have an efficient algorithm for computing it and a multiway equilibrium doesn't necessarily have the same useful properties as a HU equilibrium:

http://www.thinkingpoker.net/2017/04...ustin-morrill/

Discussion of multiway equilibriums starts around 31:35, but I found the whole interview very interesting.

You seem pretty harsh lately..
Everything's good ?

U of A could easily solve SB vs BB equilibrium for 3-handed poker, provided that we can make assumptions about card removal effects, using the same methods they used for cepheus. They may not yet have dedicated the same computational power to it as they did for cepheus, but I'm sure they have a close approximation (it might even be public, i don't know). Obviously 3-handed poker is a massively different problem.

Go look at my posts for the last 13 years... I'm much nicer now

You guys are talking past each other. What avoid is saying is IF we know the big blinds preflop strategy includes a 20% 3 bet range than the optimal solution (maximally exploitative) is to c bet 100%. And if we don't, you must donk check.

The confusion is coming from the fact that avoid often mentions something that is a GTO principle but will then throw something in which is a Nash equilibrium solution given a specific range. Basically he doesn't want to just ignore useful data so his play sorta combines ideas. Please correct me as needed

My interpretation of his posts was that the SB vs BB equilibrium if both play GTO poker on all streets involves a BB ~20% 3-bet and SB c-betting flop 100% when BB calls preflop. I don't think that can be correct.

Feel free to clarify if I've misunderstood

^ I understood the same.

I also think he meant that. I don't know whether it is correct or not. It seems...unlikely

The short answer is when you flat 100% you reduce SB's options (SB can't 4-bet preflop or donk flop and he "has to" c-bet) and BB gets to act with more information, ie. knowing the flop cards, while also giving away less information. So conversely, if BB has a 3-betting range he's giving SB more information and more options without actually gaining anything since SB has to put that third bet in on the flop anyway.

What is being referred to here by "it" c-betting 100%? I know people talk in other threads about what HU bots do, but I have never heard of one that simulates BvB in a non-HU game (where BB is in position).

"It" refers to a bot that was part of a 3 handed bot play study.

Good posts! I have all of the UofA hand histories. It's very true 3h poker is no where near solved. However, HU LHE is! This where most of my study has gone (90%+).

The 3h data is by no means useless, especially if you grasp the fundamentals from the HU play.

Long story short, flatting your bb is totally fine... lol. It's just a different situation you create for the dynamic between yourself and the SB opener. The most profitable line is almost always going to be vs which range the opponent plays worse. Vs passive types, clearly we should have a 3b range. Vs aggromonkeys, maybe we flat everything... but maybe if we 3bet they will 4bet. Always consider the opportunity cost of a line before deciding.

- Multiple lines are viable preflop. Consider opp tendencies + your image to decide which metagame to develop.
- GTO is useful, inclusive of multiway play. When you have working models for HU version of a game, fundamentals can be derived. These transfer nicely to MW range construction + balance considerations.
- a preflop BB 3betting range is not polarized. It should be linear + including TONS of suited hands 65s, T7s Q8s etc. It will play well on all board textures + SB "gaining information" is not a strong reason to avoid 3betting preflop at all imo

perfect =D pure gto gives up massive consideration for opponents ranges. much prefer to adopt a balanced approach with a primary respect to the RvR EQ in a spot.... this allows us to exploit on the margins VERY well

Well said.

@stinkypete, the above is all I was trying to say with my earlier posts.

The point wasn't that those are all strong reasons not to 3-bet individually, but that literally every single EV difference between flatting and 3-betting against a 100% c-bet favors flatting in a gto solution. Things change when neither player is playing close to gto, but if BB plays good and SB isn't going to tilt his balls off it's hard to argue that having a 3bet range is a good idea.

@rodeo doesn't matter what you were trying to say, you misinterpreted every one of my posts

Sent from my Nexus 6P using Tapatalk

When I look at a 20% range it all makes me feel comfortable. It just looks like a 3 betting range. A8o, K8s, 55 are at the bottom. The problem I have is that there are many hands not captured in that 20%. Your 56s and other such hands for instance aren't in the top twenty. How do you decide which hands to remove from the TOP 20% in order to create the RIGHT 20%, assuming you're a human being?

years of experience with range construction w\ a focus on board coverage =P