So what are you going to do when you've solved everything, good luck with that by the way. Memorise it all !

added you on discord

My memorizing skills are way above average, but unfortunately I'm just a human being (lol).

It's all about finding patterns man, it's all about finding patterns.
Patterns are way easier to memorize.

Merit in trying pal i'll give you that.
Better to have the knowledge than not, just dont over do it and confuse yourself.

I've already spent several hours studying some solved databases. Finding patters wasn't easy but I can say it's doable.

You mentioned confusion and I have to say that it is a thing. Sometimes, even after finding solid patterns, you still find some solutions that behave unexpectedly, so it's easy to get confused and/or find the wrongs answers.

But as I said, it's still doable. Thanks for the kind words!

If the goal is to model the real life game, is a solver the right too for the job? Adding more sizes to the solve in attempt more accurately model a real life game is strange thing to attempt isn’t it? Sure, a human isn’t restricted to sizings the same way a solver would be, but a human will be unable to balance those different sized ranges anywhere near as well as a solver can. So, if you wanted to study how to respond to a 66% bet vs 75% made by a human for example, studying the solver’s response to these two different sizings doesn’t seem like much bang for the buck. What I mean by this is that I don’t think there’s more to gain by studying against a 10 size betting strategy than a 2 size betting strategy. Maybe it’s better if you’re building a bot, but if not, it seems like overkill.

Again, If I were building a bot the silliest thing I could do would be to open a public thread in the biggest poker forum ever.

Anyway, you get way more accuracy and more information using more bet sizes as I stated before. This databases is going to require time and money. If I’m doing an investment I want to make sure that I’m doing it in the best possible way. I can’t see a reason to make an investment knowing that I won’t be making in the optimal way.

I’m suggesting that solving for this number of bet sizes will likely have diminishing returns—which is something to consider given your concerns about the cost. Also, I suspect that the returns will be very small, to the point where it won’t be valuable to a human player. This isn’t to discourage you from your project, only to give you something to consider so that you maximize your efforts and money.

I agree. For the purposes of improving your own EV, this is overkill and will offer diminishing returns.

But there's scientific value. Like if you want to research optimal sizing strategies, a big dataset like this would be useful.

There is none reason to solve all 1755 flops to find "patterns" if you really want to study 1755 flops in every spot and every formation, you will spend your whole life and will not learn any different things than you would from the biggest pio or gto flop subset, if you used them, they already have all kind of boards, and if you say you want solve all 1755 to find average frequencies this is the reason pio e gto+ subset exists, they will not differ from the 1755, this topic is just hipocrisy, there is no point in solve all flops for all spots to study, everybody know you want this for use as live assist, everyone who wants things like this or is taking a complete off track study option or in 99% of cases just want to have solutions to of all spot to cheat, topics like this should be deleted and the creators and people who showed interest should be banned

Enviado de meu motorola one hyper usando o Tapatalk

I am interested. Just sent you a PM

i do assume you are using a basic storage for your massive srp solve of 1755 flops?

Jesus Christ, did anybody tell you of the existence of charts?
You know those fascinating draws made by a horizontal and a vertical axis?

What's more, you can export results to excel sheet and if you are able to do so you might even use some machine learning script to find patterns for you.

Again, any form of live assist is illegal, unethic and unfair. I would be an idiot posting a thread here asking for help to do so.

we are going to study on aggregate reports and charts.
if we used basic storage we could not do so.

Unfortunately, we have got to store the whole tree.

of course you have aggregate reports with basic storage ?!

are you gonna be able to access a fully saved srp spot with all 1755 on a "normal" computer or are you only going to access it on the rented server

We are going to be able to tue solved files on our normal computers.
It would be practical to rent a sever everytime we want to see the solution xD.

the storage needed for such a solve has to be > 1TB ??

it's going to be heavy as hell but less than 1tb for sure xD.
It's probably going to be 200 gigs or so

literally everything youve said so far from calling yosurself a "GTO freak" to talking about how EXPENSIVE it is to rent a $250/mo server to insisting u need all flops with like 20 sizings on each street to insisting that you can somehow build a database of full saves so parametrized and make it <100s of TB is cringey and stupid

Hundreds of TB???
I assume you have never run some sims like that before LMFAO.

Also, you don’t know what server it’s going to be rented so I can’t see how you made up that price.

I kindly ask you not to talk about stuff you have no clue.

This is a bump

I'm curious what your game tree parameters will look like.

What are your sizes on all streets? Are you planning to include donk bets? Multiple raise sizes?

What rake structure are you solving for?

What preflop ranges are you using? Are they derived from preflop solves specific to your rake structure? Or are you using generic presolved ranges?

What do you hope to achieve with this massive solution?

There are going to be donk bets.

We are simulating rake as is we are playing NL100 on pokerstars.

sizes 30,50,80,150 - sometimes allin

1-2 xr sizes. Possibly 2 for every street but we are still working out on this.

We are using preflop ranges by acevedo.

We hope to get a good edge on all of our opponents on the most common spot of the game.

Cheers!

I haven't read this book, but my understanding is that it mainly focuses on short stack play--50bbs and less. Is this true? If so, are the ranges based on those short stack sizes?

the book offers several ranges.
we are using 100bb ones