Dumb question I know, but can someone break down the basics for me? I am familiar with HEM and PPT but no other poker software so I'm a little baffled at what exactly this magical software does and how. What inputs does the software use and what conclusions do they provide and how does the software arise at those conclusions? How possible is it to be a winning player these days not using this software? Explain to me like I'm five.

As I understand:

The solvers really only work perfectly in HU pots. HU it calculates what you can possibly do and what he can possibly do and it figures the spot where you break even no matter what either of you do. As of now, I think it start calculating from a given point, usually flop or later with inputted ranges for each player.

At significant levels I'd guess it would be hard to be a good sized winner without studying this off the table.

Basically, solvers generate a strategy such that no player can change their strategy and win more money from their opponent. So they generate strategies that are impossible to exploit.

Usually what you do is put in some preflop ranges (though these can be calculated in solvers as well), give a few betsizes, and click go, then the solver will spit out an unbeatable/unexploitable strategy for every hand in your range.

It gets to these unbeatable ranges by having each player try to exploit the other one for as much as possible. So player A plays strategy A, player B exploits that strategy with strategy B, then player A adjusts to strategy C to exploit B, player B adjusts to exploit strategy C with strategy D, and so on. After enough iterations, neither player can adjust their strategy anymore to exploit their opponent, and that's the "final result" that the solver gives you.

You can be a winning player at micro and lowstakes games (and maybe up to 5/10 live?) without doing much solver study, but strong regs at say 100nl online or playing hs sngs are certainly studying this stuff. Possible to make 5 figures, would be difficult to make 6 figures without delving into solvers.

As a beginner, it's really not worth thinking about too much.

This is not true -- solver strategy pairs will basically never be breakeven. The key point is that there are no further profitable adjustments to be made in any strategy generated. Things like range advantage and position still exist and cause massive amounts of asymmetry in solver strategy pairs.

Put another way, if you look at some btn vs bb raise/call spots in a solver, btn will be +ev in general as they'll have hands like overpairs and sets that are tougher for the bb to have, and the bb will have a bunch of random weak trash in their range they have to defend with pre. The solver will spit out a strategy for bb that the btn can't exploit any further, but btn will make money in that spot just due to the structure of the game and the way ranges work.

Yes, I should have said figures the spot where neither player can exploit the other rather than figures the spot where you breakeven. Thanks.

Solvers cannot do it now, but if they solved the entire game (rather than isolated situations with input from the user) and 2 players played such strategy against each other, then the expectation would be for them to breakeven (assuming they played the same number of buttons/big blinds).

Right, this is correct

Thanks for the answers. Still a little too much for me to wrap my brain around but I guess I'd have to experiment with using one before I can fully understand.

If that's the case then I'd recommend not experimenting with one

robots don't know how to play poker. just find games with ******s and nit it up against them. poker isn't about being the best, it's about playing the worst.

So, solvers try to find the correct balanced action that when taken in specific situations, is neither exploitable nor exploitative, and is agnostic of opponents biases or tendencies. The actions can be difficult to understand without runnig simulations or understanding some fairly advanced math.

In other words, it is really not a great starting point for someone who is trying to develop a feel for the game at lower stakes. Most low stakes players and beginners have exploitable tendencies, and using the strategy suggested by a solver is really not going to be best.

agreed. you gotta play with worse opponents or turn those good players into playing badly.

I guess more my uncertainty is exactly what "actions "does it tell you to do? Whether to call or fold? Whether to raise? What percentage of the time you should take certain actions (ie bet 70% of the time, check 30%)? Bet sizes? All of the above?

Do you put in ranges like with PPT to get the software to do its calculations?

Basically all of the above.

I don't think any of the current solvers just tell you bet sizes. I think that is one of the inputs you have to make. But you can input multiple bet sizes for the solver to consider (although the more sizes you input the longer it takes to solve). But you can see what bet sizes it uses most often and then you can try removing the ones it uses least and see how much that change affects the EV and so by using it like that perhaps you can find what may be close to optimal bet sizes in certain situations.

The above is how I recall that this was explained to me once. I bought a license for a solver quite a while back, but haven't really used it much, so I'm not too adept at its use.

This is also correct. With that said, it's worth pointing out that you can run turns with ~inf sizes in well under a minute and rivers with inf sizes in a second or two, so with the exception of flops figuring out pseudo-optimal sizings in short amounts of time is very feasible.

That's whats wrong with poker currently.

Why put effort into something, if you're not going to aim to be the best mofo there is?

Its like settling down with an average salary and slave yourself for the next 40 years trying to pay out your house + taxes
fk that.

Nothing publicly available anyway.

You might want to take a second to think through the logic of what you posted there dude.

You think the poker playing environment would be better for us if everybody playing studied to be the best?

Stick around BQ long enough you will see plenty of these barely disguised brags...its all about being the best yada yada yada yada

Wrong. It is exactly what PPG said it is. Finding fish and stacking them. Ask any real pro (preferably one who is not trying "promote" poker in some way)

On top of that, the concept of 'best' is very difficult to define in poker. If I learned GTO perfectly and could actually solve a 10 handed game and always made the best, Nash equilibrium decision in every case, would I be best? Or if I played a perfect TAG that maximized my winnings while limiting my losses against unbalanced players, would that be best? The answer is, of course, it depends.

There is no such thing as a best style of play. It all depends on the games you are in and how your opponents are playing. Too many newish players fall in love with this concept that GTO solutions are the perfect way to play, when PPG is absolutely correct, playing a simple explotative style against exploitable opponents is going to be far more profitable than finding a GTO style that is really only applicable in nosebleed games that 99.99% of us will never play in.

Poker is not a game of linear improvement. There is no such thing as an objective standard for 'best'. Play the style approrpiate for the games you are in.

.

No. There is one, and only one, objective definition of "best" for cash games, the highest EVbb/100 over sample sizes that approach infinity.

Really? The 'Best' player in one game is going to be the best player in other games? There isn't such a thing as being best for specific situations, while not being best in others? You can boil the complex and dynamic nature of a very reactive, situational, and evolving game to 'highest bb/100' and think that this is a meaningful statement?

Yes. I noticed you took out the two parts that makes it meaningful, highest *EV*bb/100 and *as sample size tends toward infinity*. Those imply best overall strategy against all strategy's over huge samples. FWIW, I don't think this is worth an argument as it's pretty much self-evident that it must be true.

It’s 100% the opposite. Why would you want to spend all day studying and playing poker in tough games when you could make twice as much in easy games without any extra effort?

I did ignore the EVbb/100, as I am not exactly sure what that means, so I just assumed it was a typo. I did ignore your comment on sample size approaching infinity, as any applied math problem that uses the statement 'sample size tends towards infinity' proves my point that 'best' is not a realistic concept in poker (or any other real world use case). i don't expect you to respond, since you have already asserted that you are correct, so I am sure you have moved on to making other confused proclamations.

So, you are correct, in a sense. If a player played an infinite sample size, which would mean that he played in every hand played, and had the best EVbb/100 (still not sure what EVbb means), then yes, he could be considered best. In the absense of the poker player who has played every hand, I stand by my statement that poker is too dynamic and evolcing to have a concept of a static 'best' state. This isn't worth arguing, as it is pretty much self evident.

EV adjusted bb/100 is the measure without actual variance included. I.e., what you should have averaged if you ran true over the sample size.

I'm not implying an actual infinite sample size, an infinity of anything is impossible. "Infinity" is a concept of method in mathematics and doesn't exist in metaphysical reality. I'm saying very large sample sizes.

Large sample sizes, say ten million hands, means (assuming six max) that you've likely played many different limits against all player types. It implies that on average you've generally made the correct adjustments and all that in the majority of cases.

So therefore the best player overall would have the highest EV adjusted bb/100 compared to any other player with a similar sample size.

A funny thought, somewhere in the multiverse someone has that and has actually lost money (due to variance). They would literally be the most unlucky person in poker. People should keep that in mind when they think they run bad.