Consider that on the turn out of position with an SPR <1, I'm striving for the least worst option rather than the perfect option. The jam wasn't because I love punting, I merely like punting.

So pick your poison...if you take check-fold off the table which we should be able to agree is the most distasteful option given you forfeit a sickening amount of equity.

I don't like the small xr, BTN range has really few Bet/folds to this small sizing so you don't really accomplish much in terms of folding out equity. I would xc flop.

Turn looks ok, I mean we're obv stacking, just a question of shove or xc, and I think shove looks preferable.

I am just wondering how do you guys apply strategies that are closer to ones that solver offers...first of all, how do you know that your decisions are in align with solver's ones? secondly, why do you think that it's optimal? lastly, how do you implement frequencies like "raise a hand 60% of the time".

it depends

Meh, don't we have enough better hands to c/r here even at these stack depths? (EDIT: Actually c-bet here probably has some merit as well, so the sims that want a mix make sense to me.) And if we have to pick backdoor flush draws wouldn't it better to have ones where we don't have a pair? If we call flop and villain checks turn there's a reasonable chance we can win this hand at showdown if villain is prone to checking turn and river with Ax blanks just trying to get to showdown, and maybe we can consider a river hero call and/or some really thin value on certain runouts. (although going for thin value would be read dependent)

I'm not using a solver so maybe my logic's off base, but this feels instinctively spewy. As played, ugh, I guess we go with this hand on the turn but don't we have to have at least some give ups? And on that the solver agrees with me at least.

This whole thread is just a simple example of the problem of solvers. OP ran a play that he thought was the "solver" play. However, when erc ran the hand through Pio solver, the play that OP ran was in fact not the solver's top choice for that play. In fact it was a very rare exception line that was chosen.

Simply put, humans cannot duplicate the solutions given by the solvers. You may study one situation and memorize every variable the solver provides for the situation, but if you change any detail of the hand, that solution changes.

Pio solver is a great program, however, I don't know how humans can make those same decisions in game.

This begs the question then, is it worth it to try or are we better off keeping to solutions we can figure out in game.

Have you actually compared the pools frequency to pio in that spot? Using pool vs pool range. I guess you could also argue to at least check your range vs pool in that spot. Compare btn fold to pio agg report in the same spot? I know you have enough hands to do it. Also check for other frequency inconsistencies down that line

I agree with the x/c. Not disagreeing with you.
But I could be persuaded to X/R in that spot pretty easy.

We can't work out these solutions perfectly in real time but we can simplify them and take away the general points. It's still useful to see what a balanced x/r range looks like, the hands which are selected, and how we exploit when we tell it that villain isn't playing the correct strategy.

Sent from my SM-G935F using Tapatalk

If I am not mistaken , solver offers the perfectly balanced, unexploitable strategies and if u deviate from them even a little bit, it's not balanced anymore because those strats only exist in binary fashion - either it's perfect or not. So, what's the reason of those takeaways and simplification?

Pretty sure you are mistaken. If we are supposed to c/r 60% but c/r 61% we are still balanced. If we start check raising 90% in a 60%, then of course we would be unbalanced.

We intentionally deviate to exploit our opponent because he's deviating. It's nice to learn what the 'perfect' solution is first and see how that changes according to villain's range

Sent from my SM-G935F using Tapatalk

I am not mistaken. PioSolver refers to itself as "Nash equilibrium solver". Nash equilibrium can only exist in binary fashion.

So, no one from us and our opponents follows solver strategy that makes solver absolutely useless IRL. I rest my case
As it was mentioned above, OP likes to think that his game is close to GTO, but IRL:

1. No, it's not
2. It is spewing

Nah I completely disagree here. I will try to use these strats vs tough players and unknown players where I dont know enough about population tendencies. Also, its the solvers that can offer the maximally exploitative solutions for a given street, so technically we are still going by the 'solver strat'

Sent from my SM-G935F using Tapatalk

After checking, earlier posts have proven Pio likes to raise A2 somewhere between sometimes and alot in certain conditions.

I was under the assumption Pio would advocate a lot more flop checking in this scenario however I knew it wouldn't be 100% but I had a plan to play my whole range as a check. This means I'd already conceded that I was going to deviate from the Pio solution, sometimes called "GTO", for the sake of simplicity. Plus I think mid and small populations instinctively play further from "GTO" v a check than a bet on dynamic boards ip, giving me the opportunity to EXPLOITATIVELY hammer them.

Pio advocates checking A2 on the turn. I imagine this is predicated on villain jamming many of his draws. I'll remember that but do I plan to enact that against randoms? Of course not. Give a random an opportunity not to bluff while realising all of his equity after facing prior aggression and he will take it. So thats another snap deviation from GTO.

I consider myself an exploitative player. I haven't paid off a river raise with one pair in forever because they always have it, but Pio has made it easier for me to develop holistic strategies and to see where opponents intuitively play well ie. close to solver solutions and take them out of that portion of the game tree and drown them.

Solvers offer minimally exploitative solutions that are based on Nash Equilibrium. So, applying it for maximum exploitative strategies does not make much sense. The only "solver" that exists for maximum exploitative game is Libratus, but

• it's for HU only
• it's not proven to be optimal
• it's just NN deep learning algorithm that selects the path based on the known information. It's much like human intuition.

I saw this link in another thread : https://www.pokertube.com/article/ch...-no-fan-of-gto
I think that this guy makes a lot of sense.

The thing about Nash equilibrium is that your opponent must also be playing according to Nash or you are leaving money on the table. If your opponent is deviating from Nash, you are supposed to deviate as well.

And this goes to my above point. No one is figuring out Nash in game. Therefore, no one is playing Nash, therefore it's use in real poker is limited.

GTO style is maximally exploitative strategy if u know your opponent range in a particular spot.It's all about maximising EV.

Sent from my SM-J600FN using Tapatalk

No it is not. GTO makes you unexploitable. If your opponent makes an error in either direction, then you are supposed to adjust for exploitation.

Look at a simplified GTO example using numbered cards from 1 - 100. You each put $1 in the pot and your opponent can only bet $1 or check. You can only call or fold and if checked by your opponent you cannot bet.

With $2 in the pot and he is betting $1, he is supposed to bet 83 - 100 for value and bluff something like 1 - 6. And you are supposed to call off with something around 66 - 100

Assume now he doesn't bluff. If you call off the GTO amount you are making an error. In fact, calling off with 66 when he never bets lower is a huge error.

Now with a pot odds call, you are supposed to call off numbers 87 - 100.

If he instead bets everything, if you are only calling off with 66 - 100 you are leaving money on the table.

But you can factor this in with Pio, albeit with some flaws. I start with flop and first lock my opponents strategies based on what I think they actually do. Pio assumes then that everybody plays perfectly after that which ofc is a problem but then you can move onto the turn and river and do the same thing. In this case if all of our opponents strategies are predefined no matter what we do, the response for us is no longer nash is it? I would have thought here it is maximal exploitation, but perhaps I have the concepts wrong.

The initial nash strategies may not be directly applicable to most real life games but we can still learn from solvers by looking at the way it chooses to deviate.

Sent from my SM-G935F using Tapatalk

U r not giving a GTO example u r giving a toy game with predefined rules.

Sent from my SM-J600FN using Tapatalk

Interesting debate.

The more I use solvers the more I think its less and less applicable. What guys are doing here for sure is not making you a lot better player.

I think the only use of solver is to see how it ""thinks"" and deviates on certain texture vs suboptimal strategies. Thats it.


PIO is the best for lazy coaches who want to make theory video, but doesnt really want to give out too many info about how they play or talk the pop tendencies etc.
You fire up the solver, make it work while you are not even there then in your videos you explain what you see lol. Yepp very good value indeed. Furthermore we dont know what financial incentive big coaches have to promote the PIO.

Its more useful in heads up or spins higher stakes where you only play shallow stack, way less positions, and no ICM or other factors are present in game, and where also ppl are way better postflop than the average MTT small-midstakes guy. Except the crushers most avg mtt reg is in the dark ages postflop.

Which I find even more ridicolous is PIOsolver is still very expensive. Thats just a ****ing joke when you can get a decent solver for as low as 50 bux. Couple years went by, several software flood the market and you still selling your stuff as much as it was day 1? Also, the 500 $ version is with restriction on how many threads u can use. Also being one of the slowest solver on the market. Only thing its better than competition is the interface. Thats it. But edge is still over 1000 usd omg. While simplepostflop is you get 250 bucks and it solves preflop also.


As I checked the solutions here it doesnt make any sense. Its not HU postflop so the range of the BTN postflop is so much different. Copying HU solutions when it 3 ways its just amateurish. Or maybe i misunderstood something.

Better method:
whats poptendency in this spot?
what do I want to accomplish with this hand? Whats my range? What does a player type like him bets in this spot? Do i have FE vs that
Whats my plan on future street on certain runouts?
etcetc

I'm no mtt pro but just purely chipev standpoint I would cbet flop, planning on barreling couple good turns.
With this equity, vulnerability and backdoor EV for sure I dont want to xraise, nor xcall in this spot.

GTO+ is 75$ and it goes along with CardRunnersEV which could be the most useful software to become good player(think about OtBRedBaron)...

Sent from my SM-J600FN using Tapatalk

It is a myth that GTO is defending strategy(being not exploitable).It is max EV strategy.Think about limping with AA.U can stuck a gto player putting u on a wrong range but you r losing value not raising and risking to lose when he hits.

Sent from my SM-J600FN using Tapatalk

GTO stands for Game Theory Optimal and much more than poker. Toy games with predefined rules allow to study it effectively, if studying is your purpose. The Clairvoyance Game is the most simple example.
IRL GTO is used by poker players who like to think that it's optimal strategy.