Using the practice hands vs GTO opponent, we are 6 handed. I did not mean to have this on for NL50 6max cash game but anyway, this is the question. We all have 100bb´s. Blinds are 0.5/1. HJ makes min raises to 2, CO calls, BTN and SB fold. So it is on me in the BB to call for 1 BB.

GTO Wizard says this is wrong. WTF! It says fold is correct.

I have recently started defending the BB a little lighter but how is a defend with A10 for 1BB when we are all sitting 100 bb´s deep a mistake? Maybe simply sue to the face that we can be locked up. The next hand of 56 is clearly a defend and shows as correct, however, it has 64 as a fold.

It's due to reverse implied odds. When you're this deep mediocre offsuit aces don't play well against two fairly tight ranges. If you were shallower or if the opener was in late position where they'll play a lot of off suit aces that you dominate then it would be different.

That being said, folding A10o still seems a little too tight to me, given that you're only facing a minraise. I would likely call in game. Most humans won't leverage their stacks to the max like the solver would against you in that spot.

It seems like folding vs. a 3X open would be more reasonable.

Here's the BB range vs a 3X HJ open and CO call:

hands that are likely to make 1 pair at best and and with a weak kicker at that just dont do well in multiway pots. You would rather have something like a suited gapper.

Cause AT is terrible multi way and it will just get you in trouble. If you flop an A and they don't have one gl getting any money out of them. If anything you can squeeze it cause you block a lot premiums, but being so deep its to dicey. I told people this before solvers ever came out from playing 10s of millions of hands spots like this that you are losing money here. And they always argue. You make a solid point out of having solver is not enough. They think they are geniuses and using a solver and should win at poker but they don't and say its bad luck. It is not actually getting the data that is important it is actually how you look at it. And if you look at the right way you will be successful in poker. If you don't you will not win at poker. The fact you don't know you have reverse implied odds in this particular spot just goes show you got a lot to learn. This leak seems small but it is spots like this where my students have lost so much money. I saw it over and over again as the main leak to a lot of players in cash games. The only time I would ever squeeze this spot is if the rfi is lag and the cold caller is a fish to push out the lag and get the fish who has much wider range hu. Other then that just avoid this spot. Trust me your losing a lot of money long term here.

Enjoy losing money.

You might be right, and this could very well be a leak. Are you basing this off of a significant sample of aggregated hands? If so at what stakes?

I play pretty much exclusively live these days, and people play pretty bad. My thinking is that many players are opening too many weak offsuit aces in the HJ and people are definitely calling too wide in the CO. That plus if I perceive that I have an edge postflop it seems that expanding my range a little would likely be reasonable and still profitable.

I wouldn't think 50 NL would be significantly tougher than the higher stakes games I'm playing live, and that was my reasoning for advocating defending A10o to a minraise. Obviously anytime you deviate from GTO you open yourself up to potentially losing EV, and you do need to tread carefully and know when to get away from a hand.

I've been thinking about getting back online and putting in 100,000+ hands at low stakes just to do some database analysis to look for obvious leaks (like this?) that would translate to higher stakes live. Playing live you never really know if your assumptions are correct.

The good online players I play with live definitely have their preflop ranges dialed in better than I do. My recent results improved dramatically just after I started doing maybe 5 min a day drilling GTO preflop spots through an app. So I'm definitely open to suggestions for how to approach improving my preflop game if you have any suggestions?

I think rake is the main consideration here. If the rake was lower, ATo would shift to a call. If one of the players is weak, then it will shift to a call. Solver outputs aren't set in stone, and the equilibria they produce are really unstable and very sensitive to the inputs they're given.

This is 50nl rake:




It's hard to see, but there is a sliver of calling, so calling and folding are the same EV. At worse, it is a very small mistake in theory. There are many reasons why it may perform better or worse in practice.

Here is the same thing, but at 1knl rake:



Now, it is a clear call, and we even start reaching down to A9o. Rake really punishes passive actions, like calling the BB.



The moral of the story is that the output you see from a solver isn't the end of the conversation, especially in real games. You should know what inputs were used to build the model and then deviate as necessary in your games.

It has to do with multi-round game theory. You'll usually be the defender in this hand, which means, simplifying things a little, that someone can bet his value hands plus lots of bluffs on the flop. On the turn, he'll bet his value hands but drop some of his bluffs. And on the river, he'll drop more bluffs. One of the results of this is that your hand on many flops might have a small range advantage but still be negative EV to play.

Mason

Thanks for the response. Yeah rake is an important consideration, as are the tendencies of the players in the hands. I definitely adjust based on the player types. If the initial opener is a nit A10o becomes a fold, and if both opponents are loose and will respond predictably to 3-bets then sometimes a squeeze with A10o is a good play as well.

To Mason, yeah that's what I was alluding to when I mentioned that most humans won't leverage their stack to the degree that a solver will. If you look at simulations for example for certain ace high boards the solver often uses large polarized sizings to threaten stacks, and the solver bluffs aggressively, including some non-intuitive bluffs. This can turn A10o into an indifferent bluffcatcher.

In live poker a lot of people use smaller sizings with the value portion of their range and don't bluff as aggressively as they theoretically should. When they do use large sizings to threaten stacks it's often with a value-heavy range. So in practice I think A10o can often be a fine hand to defend. This is provided we can read the spots where our opponent is unbalanced and make some exploitative folds when indications are that we're facing the strongest portion of their range.

Obviously this is player dependent, but even high stakes players can't balance a polarized range as well as a solver can.