CBet OOP
Posted by Langerz
Posted by
Langerz
posted in
Low Stakes
CBet OOP
First an intro since it's my first post. I've been lurking for a month or so as I get back into poker (it's been several years), but have plenty to learn as things have changed some. When I left I would say the GTO/solver discussion was just getting started. I'm pretty new to using Snowie and GTO+ but while I was doing some review today I had a realization that the GTO+ never asks who the preflop aggressor is. I understand the reason is initiative isn't really a concept and the difference is really about ranges, but it led to some interesting things related to cbetting OOP.
There were a couple hands this came up and I included one below. I'm having trouble with the converter for some reason, but it's a simple hand so I'll just post the details and figure out the converter on a future post.
Effective Stacks 100bb
I open KK in the CO for 2.5x
BTN calls and everyone else folds
Flop = KcJs8s
I cbet for about 75% being a pretty wet board.
Snowie identified this as a blunder and suggests checking 100% of my range.
I built a tree in GTO+ to see if it as just snowie and it suggests the same thing.
This seems weird to me that I wouldn't want to bet KK on a board like this. I can obviously get value from all kinds of hands and there are plenty of combos GTO+ has in the BTN checking range I would rather not see the turn for free.
I can see one question being "if you weren't the preflop aggressor would you donk this flop" because from a GTO standpoint it's the same. The difference is I never have KK (or JJ and maybe not always 88) as the preflop caller here.
I'm curious if there is additional insight. I think at least at low stakes it would be a mistake to check 100% on that board. That may just be a non GTO adjustment to the pool, but I'm still curious on the solver output.
Loading 4 Comments...
First thing to check is what GTO+ and Snowie call with here preflop. I am very familiar with Snowie and I know it is not going to be calling wide preflop and is going to be 3betting more than it calls. I don't know what GTO+ ranges look like.
Run these ranges in Flopzilla and see the distribution. BTN hits the flop more with the ranges given by Snowie with more TP+ hands. Run the same ranges in PokerStove, BTN has more equity on this flop. This is only an indicator its enough to get started.
All of this is assuming that the BTN 3bets optimally with Ax hands and isnt defending here with T9s. Your CO opening range will have tons of Ax that miss here vs a BTN range that shouldn't have them at all, so even though you have KK here your range is behind on this board. Your missing at a high enough frequency and your equity gets dragged down lower than it first seems.
This is assuming vs good opponents. The reason you swing towards betting here at low stakes is because they are flatting with those hands that ought to be 3bet and calling with hands that ought to fold. Thus your range equalize/gets head on this board so now you can start firing.
My Snowie evulation isnt showing a blunder though, its close enough where it doesnt matter if you bet in practice. We certainly don't have the same stake/stack size settings so that is the reason. But for this spot, my opponent would have to be some kind of world class gosu for me to want to check here. Going to bet vs the pool of 99% instead of playing optimally vs the 1% for sure.
When you get to GTO land heuristics such as wet board protection or dont give free cards dont matter. You just don't put in more money when your range is behind even though your holdings are absolutely strong.
Awesome.
Thanks for the great reply. Initially makes sense. I need to think a bit to see if I have questions.
As a note GTO+ doesn't solve preflop. You enter ranges and start at the flop. But the ranges I used would apply to your comments.
well its good that if you put snowie preflop ranges into gto+ it comes to similar conclusions because now you can tweak ranges to fit your pool and see what exploit deviations are best knowing that you can somewhat trust the outputs
Be the first to add a comment