Checking behind.

Many people check behind low coordinated boards these days but does anyone check behind boards like these?

The bot Sonia when playing against herself for a prolonged time would only c-bet this flop ~70% of the time. Maybe this could be interpreted that GTO would suggest us to c-bet less then 100% of the time. Now what hands (range) would you chose to check behind with against a capable opponent?

Perhaps a discussion starter.

10/20 Limit Holdem
2 players

Pre-flop: (1.5 SB, 2 players) Hero is BTN
Hero raises, BB calls

Flop: 7 K 2 (4.0 SB, 2 players)
BB checks, Hero?

Vs non-3betting light-checkraisers I check back this flop like it's my job.

I would be very surprised if it was GTO to autobet any flop texture 100%. It would be interesting to see which textures Sonia would check back against herself more or less often. How did you figure out what she did on that board?

If we cbet 100% here then our opponent can CR ATC and show a profit (I suspect)
It seems to me that the conventional wisdom says that vs. opponents who will do that we should be willing to call and fight back with hands that cannot win a showdown.

Sonia seems to guard from this by not cbetting all hands that would have to fold to a CR, so that our opponent can not show immediate prfoit from Cring the flop.

So some part of the checking back range should be hands that cannot a CR but do have equity, 59 with 1 spade comes to mind.
If we are only expecting a bet to be very slightly profitable, it can be more profitable to check back and see if we pick up a pair or a draw.

How to balance it with high card hands so that our opponent cannot show a profit on a turn bet I don't know, I suspect we will be checking some A/Q-high hands here intending to call down alot.

Yes, she seems to fight back somewhat less against herself and just not put herself in that position.


I´ve actually thought she did it the other way around. Checking hands that would have to call a flop c/r but can´t call down. Because a hand that can fold to a flop ch/r can always make a profit on this board just looking for folds. But a hand that have to call the flop c/r will be losing that fold-profit whenever it gets ch/raised.

Apart from that I think it´s obvious we should do it with some WA/WB hands like A high. But would anyone also do it with 2x and Qx (even Jx?) hands?

Yeah, but that´s a very specific opponent type. What about against a player who does 3B pre flop, who check/raises second pair+, draws and some air on this board and is all-round capable?

Not really, in that if they are doing that we can adjust by floating some (but not all) of their flop check/raises intending to rebluff on the turn (or just outright rebluff on the flop), as well as lower our calldown range. The problem is more that our opponent can check/raise fairly frequently as a bluff (as well, of course, as check/raising for value) and we won't be able to do much about it, precisely because they won't be check/raising any two so their bluffs will be much more successful.

The check back range will want to be balanced to include hands that will fold on the turn unimproved to a bet as well as hands that will call the turn and river bets unimproved. I am not sure how many hands should check back intending to call turn and then fold the river if the opponent bets again, but I suspect that it will be a small (but nonzero) amount. It's also not clear how many (if any) strong hands will check back intending to raise some (even all?) turn / river cards if bet into.

Checking both hand types obviously have merit depending on what we expect our opponent to do, and are both probably part of a balanced checking back range.

There was a hand posted on another board where Sonia though polaris would fold 23% but she still checked back 74 with backdoor flush most likely expecting to get slightly more value from picking off bluffs if she hits a pair or makes the draw.

If the EV of a bet is only fractionally better than break even, then the EV of a check is probably running very close, specially when we have draw equity we don't want to give up.

Right right, I didn't mean ATC as in always, just as in we cannot really stop him from profiting from a bluff some fraction of the time here, regardless of what hand he bluffs with.
And it is extremely hard to use the only information we have (frequency of CR) to judge if he is just picking up hands or picking on us) .. and if we play back at him we risk just shoveling more money into the pot the times he is picking up hands, so it is a dangerous adjustment.

Are you sure this wasn´t because sonia thought villain would check/raise some and that she would then call a flop raise because of backdoordraw?

Also, 23% is not that much at all. Some percents might also be hands worse then 74.

Can you elaborate on this for me plz

Do you mean vs solid opponents who rarely 3bet pf you will chk back? What type of hands?

I assume the non 3betting pf strategy is best used vs solid opponents who will play well in big pots in position

Thanks

Yes.

If his strategy is to never 3-bet preflop he will naturally c/r a lot more on the flop simply because his distribution is stronger and also because his range is so wide that he can very credibly represent hitting a lot of dry flops.

So if I have a hand that has some showdown value but doesn't like calling down a flop c/r I will check it back a fair percentage of the time. I will also check back a few hands that I would normally bet/fold the flop with intending to call the turn if I improve and give up otherwise. Then rarely I'll check back a strong flopped hand, intending to raise the turn. This I will usually only do if I have been checking back a lot in the match.

Pretty much. If they cap light but correctly for value, show down everything with showdown value and also extract the maximum from their marginal and stronger made hands then 3betting them light is pointless and you can't really just 3bet your strong hands cause that would be too face up.

So the solution is to never 3bet.

Sonia 3B around 20-25% against herself and she also donks some flops so SB shouldn´t be that weak IP against OOP (as a non 3B type) and she still checks back a lot. But I guess it´s way more easy for a bot to play perfect after checking the flop behind then for a human.

Well, it also depends on how often she was peeling the flop. I'm guessing a bot won't be check/folding many flops.

Cbetting a very high frequency is a good strategy vs weaker players that give up too many flops and don't check/raise enough either.

The value in a c-bet is that you get folds or get to see the turn and the river for the cost of 1 small bet vs the weak players and you mask your range vs stronger players.

The disadvantage is that if your opponent isn't giving up often enough then you really won't get enough folds to justify the investment in an automatic bluff and you give up the chance to get a free look at the turn if your hand has some potential to improve.

Even then, if the opponent doesn't get any ideas here, and more so if he peels a lot or little, a cbet must still be the best move, this being a poor enough flop.

as far as i know she uses weight tables and calculates positive and negative potential (the chance that you will win although you are behind atm and the other way round) and so on and her checks may be based on these and similar indicators. But im sure she will adapt and change her equilibrium to exploit the opponents (for example bet more against passive non-tricky players).

From what I´ve picked up she seems more prone to check behind with hands that have more negative potential then positive. Mostly it seems to be hands with slight SD-value but sometimes it´s strong hands or crap.

then it might be:

low equity + high negative potential -> give up unless there is a lucky turn catch

marginal equity + high negative potential -> way ahead/behind + induce bluffs

I guess a function of equity and negative potential should correspond with bad flops (for example draw heavy board -> high negative potential).

+ some sort of bucketing for similar flop boards, so it is "basically" a function of a.) flop texture b.) equity c.) negative potential d.) opponent model

Yes, something like that.

I think the strong hands she checks behind are hands that have low negative potential... hands that seldom gets outdrawn by giving a freecard. There is probably more to it but that is one factor it would seem.

Oh, and Sonias "handstrength" got nothing to do with equity. It´s just how often she has best hand vs opponents percieved range. So if you use both equity and "+/- potential" you´re kind of double counting.

So, so far we got:

low strength + high negative potential -> give up unless there is a lucky turn catch

marginal strength + high negative potential -> way ahead/behind + induce bluffs

high strentgh + low negative potential -> slowplay

What hands would that be on this board do you reckon?

perhaps check and give up 64 (with or without a spade?)
check and induce Ax, Qx, bottom pair type hands?
slowplay K7? or maybe KQ since K7 want a lot of action?

That's the thing though. A non-3betting opponent will almost always "get some ideas" on such a dry flop. This is exactly the kind of flop that his strategy tries to abuse. And if he's really tough he might even c/r this flop for value super-thin with some bigger A-hi's and stuff.

So you'd better have a solid plan for how to deal with a c/r. If your hand is the sort of wishy-washy in-the-middle equity wise hand that we'll often have then taking a free card and hoping to improve can be a very nice move.

Then you just need to be careful that he doesn't run over you by firing the turn every time. This you can counter by raising thin for value on the turn, or by slowplaying a few flops, or even by bluff/raising the turn although that should be used very carefully cause it's extremely expensive (risking 2BB to win 3BB).