People are putting in a lot of work with solvers but aren’t grasping the rationale behind the solver recommendations. It seems everyone makes a play because it’s “solver approved” or “what the solver says” instead of considering WHY the solver says that.

This severely limits people as it is impossible to memorize every solver approved play you might encounter, but understanding the rationale for why the solver recommends it affords you the flexibility and adaptability necessary for a winning session.

This is such a cold take.

not to be too dismissive but I genuinely don't understand who is out there not trying to understand the underlying game theory concepts, that has any level of credibility whatsoever.

Obviously you need to understand the factors driving the solver behavior for it to be useful. This is like saying people shouldn't learn math by hitting buttons on a calculator, like no ****. But at the same time, saying ppl are too dependent on calculators would also be bizarre, its a great tool that can be used ineffectively.

Meh take

Think you are a couple of years too late with this take. At some point it was very true. But I think there are tons of good free content out there where you get all kinds of explanations/theories on what solvers do. Even like a lot of the super old school live regs sound like they studied heaps.

The way the majority of players seem to be using "solvers" is to arbitrarily input a few bet sizes and lock/prevent certain actions, that actually do occur in practice, so that the majority of hands in a given range end up being mixed frequency plays utilizing all sizes at some frequency. When you do things this way no matter what betting action is being checked you will generally end up perceiving yourself to be correct "according to the solver" while causing a nonnegligible EV loss compared to the full game tree that you didn't even come close to approximating.

Imagine if you were using a chess engine and you determined that if you chose any of the top 7 best candidate moves at a given node that you made the right decision... This is pretty much what the majority of players "using solvers" are doing. I'm not saying that these mixed frequency combos aren't mixes, I'm saying that if your input sizes only included 33% and 50%, but that the actual optimal sizes vs an opponent would have include 41% that both 33% and 50% could only ever be "close enough," never actually giving you the highest EV strategy.

I wonder just how different things would be if the first mover was named PioCALCULATOR rather than PioSOLVER.

Watching the triton series I've noticed less and less bluffs. Not sure what that means

That sounds more like some sort of exploit because from what I've seen solvers bluff their asses off. Humans even the best in the world bluff way less than the little guy in your computer so bluffing less would be moving away from the solver

depends on nodes, plenty of nodes are overbluffed

Some people misuse solvers.

In the oughts, some people misused Poker Stove.

For that matter, when most of you were in diapers, people were misusing the Sklansky-Malmuth Hand Groups.

[i]Plus ca change, plus ca meme chose.[i]

How bout this: without utilization of tech a lot of players wouldn't be very good.

And ANY type of real time assistance is cheating.

Not really accurate. If you wanted to say, regs bluff less than solvers when it is hard to find natural bluffs. I can definitely get on board.

Regs and Fish bluff way more than solvers in a ton of spots.

solvers were basically designed for people who already understand poker to take their game to the next level. They weren’t designed to teach you how to play.

So, the exact type of person who might be a good poker player in the 21st century( a kid in diamond 3 in lol) will google/ search the dark web for resources and realize the importance of solvers to modern poker immediately. A certain kind of person like this might essentially learn poker from solvers which has almost nothing in common with the way someone learned poker in 2005, let alone the 1990’s or earlier.
They are learning “correct” poker whilst having no understanding about what it takes to play your best for 10 hours after getting coolered 5 times or even what it feels like to get check raised in a big pot.Some will never figure out how to put anyone on an applicable range or even understand what ranges are beyond regurgitating what chatgptwizardpro says. In regards to live poker, its not far off from reading a book about rock climbing having never climbed a mountain. Especially in regards to exploitation of weak casino players. so these players will absorb a lot of theory but have no idea why an omc in tulsa might call 200 bbs with pocket 10’s but fold pocket 8’s to a 4bb bet against an opponent with 60% vpip. . etc etc. This kind of thinking is a cousin of the thinking of people who dont actually play poker that think there is a certain “correct” way to play the hands .

My sense is that those players who cluelessly learn poker from solvers who are most naturally inclined towards poker will end up reverse engineering bad player tendencies in a way further from but not disimilar to people like me who played early 2000’s online without casino experience. They wont truly understand the knowledge they have absorbed without practice but will pick it up over time.
Then there are those who are slightly less inclined towards poker who despite learning from the best courses and solvers online will never truly understand the fundamentals of poker.

none of this is fundamentally different from the past. I misapplied all sorts of concepts and ideas throughout my life.

The truth is you should learn how to play poker well before using solvers, but those who are empathetic and adaptable that started from a solver based approach should be able to learn all the other stuff over time. And eventually, given enough thoughtful study, have a truly deeper understanding of poker than their solverless opponents.

its “plus ça change, plus c’est pareil”

this **** just baffles me like, who is out there, that has any idea what they're doing, checking solver outputs and patting themselves on the back every time they see a mix?

Like maybe it is true that "the majority of players" have no idea what they are doing, but that's true whether they're using a solver or not. That's why the game is beatable. Materials for beginners teach u better than that.

Solver or no, most players are bad at self evaluation and will justify their actions in a way thats less than objective. It's not like players who don't use or rely on solvers are better on this front.

Solvers are extremely overrated.

Poker is full of fishs.

The more you understand about solver outputs the more you understand that there will always be spots that are way too complex for humans to replicate, that this leads in practice to either simplifications or outright errors that are exploitable, which leads to possible exploits that in turn are re-exploitable, and that great players will always be those who can navigate the best in these kind of spots. Thats what you should use a solver for but in my experience most people just use solvers to memorize charts and ranges in various spots.

So yea, fish with access to ranges will have a better understanding than before of pre-flop ranges because of charts, but thats where the knowledge stops for most people. Even a lot of decent regs I know dont go much deeper in their studies than knowing jam ranges in various spots up to 25bb eff.. and to be fair to them this is enough to be massively profitable in games up to 1k buy ins live and mid stakes online - poker is far from dead.

Take a random realistic deviation from GTO that you have to deal with when playing against humans and ask them how the solver reacts and most regs wont be able to answer. Like "CO v BU 25bb eff, BU takes the bottom 10% of the GTO jamming range and flats instead, how does this impact your strategy on a broadway rainbow flop?" 95% of players who tell you "Im in the lab every week" wont be able to answer this question correctly because they spend their time on rote memory of easy spots instead of deep understanding of GTO.

Memorizing GTO is very suboptimal way of studying GTO and poker in general. But nodelocking realistic ranges, learning how and most importantly, trying to figure out why the solver deviates as it does to overfolding, underbluffing etc, is where a solver is extremely helpful and can turn you into a great player.

Disagree

I am working on an app (with an early pre-flop MVP) to memorize solver ranges with spaced-repetition (evidence-based learning technique).

So firstly, nobody thinks you can memorize the game tree. So why memorize anything? It's similar to memorizing a bunch of vocab if you want to learn Spanish. If you memorize the 5000 most common Spanish words, you will not be fluent, in fact you won't even be conversational. But that is still a fantastic foundation to build on top of, and you even have more survival skills than someone who knows 0 words .

Similar in chess, memorize opening and endgames won't make you a grandmaster but it's usually a good start. There are cases where "a little knowledge is a dangerous thing" but overall you're making moves that will put you in better places.

Ultimately solvers are a study guide and not an answer key though. I don't think even the most hardcore solver nerds dispute this. It's not realistic to memorize the whole game tree and it's not realistic to implement that memorization, without the help of a computer, so impossible live.

There's going to be an art to implementing what we learn from solvers, and as some have mentioned, some counter-exploits of people who have studied solvers poorly. For example, a ton of trainers on the market, pretty much all the popular ones, do things like say you did "Perfect!" when you pick a dominant strategy in a mixed solution. Say, the solver says bet 70% and check 30% and you say "Bet". Well, if your opponent's strategy is fixed, those two options are the same EV. But if you always bet 100% of the time, your opponent could adjust their strategy to make your bet exploitable. So it's not "perfect" in any sense of the world. And in a world of people using trainers poorly, you might want to specfically study adjustments against people doing pure strategies where they're supposed to mix.

But I'm also not surprised that it happens because as a software developer and PM working on GTO trainer, you have to give people what they want. And people ask for a trainer where they "play a hand of poker" and then they either here it's "Perfect" according the solver, or it's not. They don't want to learn mixed strategies correctly because that's confusing. I think there's ways to make it less confusing, but they mostly center around not studying a specific hand at all, and instead strictly studying high level strategies on various board types. Talking to people, there's a small number of other hardcore poker nerds interested in this, but not most people.

So it goes back to, do you want to make a product that's successful in terms of adoption or successful in terms of helping people win - if those are not the same thing? And I think most people selling GTO content or software are going to do a mix of both, but if someone wants to pay you to study solvers poorly, you can point them in the right direction but you're probably not going to fight hard to stop them.

The biggest value add in solvers by far is just being able to answer questions with a lot more confidence. If you go back to 2p2 threads from before 2015, you can find dozens and dozens where people would argue in circles based on their opinion. And one person's envelope math based on equity said one thing, another person's said another thing. And many people would say "What are your reads? You need reads!" which sounds great except if you just sat down with some unknown player and find yourself in a sticky spot, you simply do not always have reads to go by. Solvers let you answer these questions - given certain constraints of course. So they are massively valuable tools.

Solvers will not kill poker (live at least) because it's impossible to implement the correct strategies and people will always be exploitable. Players like Patrick Antonius can naturally sniff out these exploits without the need for a computer to help. And I think people would be surprised how often these "natural crushers who don't need computers" have strategies that align with solver strategies. But there's a big class of people, myself included, who don't naturally find these exploits but the solver gives us a chance to really step back and visualize the game tree at a higher level and see how strategies interact at a higher level and manually find the exploits based on that. And it's obviously an effective strategy because for every 1 pro who does no solver work, there's at least 10 who do, moreso as you move up in stakes.

So the only people who are "too dependent" on solvers are people who don't know how to use them properly.

But, I agree with others, this discussion has been hashed to death and nothing I'm saying is really a unique or new take at all, it's more of the consensus among most good players, and the only confusion is stemming from people newer to the game still figuring out what all this means.

I’m pretty sure sklanky and mason have a book where they just make up the EV of things and then say it’s better than diff solver solutions.

You need both imo. Understanding the mechanics is great, but you can't recalculate every spot in game from first principles and then it's easier to rely on intuition of how your hand would play built from seeing thousands of sims

I think solvers are too dependent on people frankly

these are the insights that make someone a 500nlz coach

what good is mathematical probabilities if the results are skewed by interpretation.

People at low stakes are so far off from what solvers suggest it's not even funny. The mythical poker player stuck on low stakes while doing a ton of solver work doesn't exist. More like some guy who is in general a donkey and once in a while looks at solvers and finds some justification for his play