Is this never not Aces?

Is this never not Aces?


Can I just exploitatively fold the river (or turn even) here?

) 1 View 1
22 February 2025 at 08:07 AM
Reply...

11 Replies



I would suggest never folding the river here since you called the all-in on the turn so might as well table your cards. haha jk :P

This is an interesting hand, definitely feels like it's going to be AA a lot so I ran the hand through my predictive model to see.

What stakes is this at? I don't think it's mentioned anywhere, so I just assumed it was 500NL and ran it through the model as that, but let me know if it's different stakes.

I was actually really surprised at the results. The model predicts that villain will have AA here literally 69% of the time! 69%! That's crazy lol, I'm not sure if I've seen many hands where the model was that confident in one specific hand.

It seems like players at 500NL are just not bluff jamming 124bb in this spot very much lol, who would have guessed.

So the bad news is that yes, it is almost always going to be AA here, almost 3 out of 4 times when you call.

The model predicts the top hands in villains range are:

AA: 69%
QQ: 9.4%
KK: 5.6%
AK: 3.9%
JJ: 2.6%
TT: 1.9%
AQ: 1.3%
99: 0.7%
A5: 0.5%
AJ: 0.5%

So villain will have one of those ten hands roughly ~96% of the time! Funny to see A5 in there too lol, I guess some players at 500NL like their A5 5bet bluffs.

On the turn, we need to call 124bb to win a 455bb pot (ignoring rake), which means we need 27.2% equity to make this a profitable call.

Overall, the model predicts that our equity with KK on the turn when we call is roughly ~24.5%.

So overall, the EV of our call on the turn is -12.5bb. Which is not the worst call ever, I've made much worse calls. But it would definitely improve winrate if we can find the exploitative folds.


You can even fold pre, depending on the read. Best case scenario his range is A5s+AA and your hand is 0ev call, if he dose not have enough bluffs its just -ev.


by Benbutton k

I would suggest never folding the river here since you called the all-in on the turn so might as well table your cards. haha jk :P

This is an interesting hand, definitely feels like it's going to be AA a lot so I ran the hand through my predictive model to see.

What stakes is this at? I don't think it's mentioned anywhere, so I just assumed it was 500NL and ran it through the model as that, but let me know if it's different stakes.

I was actually really surprised at the results. The model predicts

200 NL. I didn't realize I got it in on the turn. It was all a blur.


This spot is a real nightmare.

What I can say is that if we’re considering a fold at any point in the hand, it seems most logical, if not only justifiable, preflop.

From the moment we call the 5-bet, we're assuming there’s enough 'weakness' in his range beyond KK+ to justify calling various types of hands (PPs, SCs, AK, AJs-QKs, A5s), at least partially, and that KK will take priority for bluff-catching postflop on a T-high board with just over an SPR of 1.

Following the same logic, we could also shove preflop.

If we think there isn’t enough, then we might as well fold preflop like the rest of our hands. Note that a balanced 5-bet range would consist of around 50% KK+ and 50% other hands like AK, AJs, KTs, A5s...


by Benbutton k

The model predicts the top hands in villains range are:

AA: 69%
QQ: 9.4%
KK: 5.6%
AK: 3.9%
JJ: 2.6%
TT: 1.9%
AQ: 1.3%
99: 0.7%
A5: 0.5%
AJ: 0.5%

So villain will have one of those ten hands roughly ~96% of the time! Funny to see A5 in there too lol, I guess some players at 500NL like their A5 5bet bluffs.

On the turn, we need to call 124bb to win a 455bb pot (ignoring rake), which means we need 27.2% equity to make this a profitable call.

Overall, the model predicts that our equity with KK on the turn when we

Can you explain how your predictive model works and what it is based on? It seems difficult to gather enough data in 5-bet pot.


Non all-in 5B = AA or someone not understanding hand strenght that well. I can't blame you but I can see fold pre/flop/turn


by JoSopinka k

200 NL. I didn't realize I got it in on the turn. It was all a blur.

Haha I feel you there, and I re-ran the hand through at 200 NL but didn't impact this specific spot very much.

Some other people are suggesting folding KK, and honestly it might be worth considering. My model predicts that villain is only 3betting ~9% of the time here, and then they are only 5betting ~4.2% of the time.

When villain does 5bet us, then KK equity is roughly ~43% against their range, and jamming would actually be -15bb in EV as they will be calling 92% of the time with a very strong range.

When villain 5bets here preflop, the top three hands in their range are:

AA: 53%
AK: 16%
QQ: 11%

I'm not sure if I would lean as far as folding preflop like other people are suggesting since we only have to call 37bb for a pot of 145bb.

That's almost giving us odds to set mine with KK, and there are also going to be times where villain will checks as well and lets us realise our equity better, etc.

For example on this specific flop, the model predicts that villain will check flop ~21% and will check turn ~25% after betting flop.

So overall, I think we can call with KK knowing that villain will have AA ~53% of the time and then get ready to fold to any big bets.


by GTOautistic k

This spot is a real nightmare.

What I can say is that if we’re considering a fold at any point in the hand, it seems most logical, if not only justifiable, preflop.

From the moment we call the 5-bet, we're assuming there’s enough 'weakness' in his range beyond KK+ to justify calling various types of hands (PPs, SCs, AK, AJs-QKs, A5s), at least partially, and that KK will take priority for bluff-catching postflop on a T-high board with just over an SPR of 1.

Following the same logic, we could also s

I think what you said is probably true from a GTO perspective. But I'm not entirely sure if it holds for population at 200NL.

My model predicts that hero's equity with KK is roughly ~43% when calling the 5bet, and that villain will have AA ~53% of the time when they 5bet us. So definitely not balanced according to your definition.

Given that we only need to call 37bb into a pot of 145bb (with 150bb left behind), we are almost getting odds to just call and set mine. Not to mention that villain will check sometimes and let us realise more equity or even get to showdown cheap, etc.

I think folding preflop might be -EV against players at 200NL in the long run, as long as we are able to fold to big bets like we see in OP. But if we aren't going to fold to jams, then I would agree it's probably best to just fold preflop.

by GTOautistic k

Can you explain how your predictive model works and what it is based on? It seems difficult to gather enough data in 5-bet pot.

Yeah for sure! The predictive model is essentially a set of multiple large neural network models trained to predict player ranges and actions in any spot for any stakes. It has seen billions of spots at a variety of stakes.

5-bet pots are fairly infrequent, but it is still likely that the model has seen millions of spots in 5BP, and in testing it has shown to be very accurate in them as well in unknown player pools that it's never seen before.


by Benbutton k

I think what you said is probably true from a GTO perspective. But I'm not entirely sure if it holds for population at 200NL.

My model predicts that hero's equity with KK is roughly ~43% when calling the 5bet, and that villain will have AA ~53% of the time when they 5bet us. So definitely not balanced according to your definition.

After checking on GTOwizard, 43% would be the equity of QQ against a GTO 5-bet range, and QQ has a +EV call worth 9bb.

By the way, according to your model, the pool’s 5-bet range is relatively balanced—it's just the equity of KK (as well as its shove EV) that decreases slightly, given that there would be significantly fewer KK in the opponent's 5-bet range.

by Benbutton k

Yeah for sure! The predictive model is essentially a set of multiple large neural network models trained to predict player ranges and actions in any spot for any stakes. It has seen billions of spots at a variety of stakes.

5-bet pots are fairly infrequent, but it is still likely that the model has seen millions of spots in 5BP, and in testing it has shown to be very accurate in them as well in unknown player pools that it's never seen before.

Ok, it seems like a pretty neat and useful tool.


by JoSopinka k

Can I just exploitatively fold the river (or turn even) here?

You should have extensive notes on other players. There are plenty of players in the 200NL pools I have played where you could fold KK on turn here. Vs vast majority we simply aren't folding KK though. KK with club is obvs less than ideal suits though.

KK is a bluff catcher here. If V is a supernit then KK is a fold. Any rec will have enough spew that we are never folding KK. Any decent reg will have a 5bet range that isn't just AA.


★ Recommended Post

its poker, could b deuces :P

Reply...