My range explorer seems messed up. It always has 100% ranges for both players with a random flop that's different from my tree. I recently installed and used some free version of piosolver in addition to my pro version if that might have anything to do with it.

This one.
The optimal solution guarantees that your opponent will not win more by doing crazy moves than by playing optimally. They may not lose more (if the moves are not so crazy) or lose much less than a good exploitative strategy would win against them (if the moves are crazy) but they will never have higher EV than playing the solution.

The problem with your option 2 is that the solver would need some kind of mechanism to judge what crazy move is and what isn't. Option 1 is both more correct and actually easier to calculate as no judgement calls are involved.

Well, I am not going to comment on thinking process of contemporary poker professionals. Most of them are much stronger players than me anyway.
That being said if our opponent plays GTO strategy they have some kind of range as well (range dicated by the GTO solution). It makes sense to think about that because when deciding which hands you should play and how it's important to estimate what hands are the best at blocking his calling range (for example) and which have the highest equity.

Again, this is valid and it's possible to work with random range also called "full range" if we start preflop. Unfortunately starting preflop is very computationally expensive. It's also useful to solve postflop spots with assumptions about preflop play even if those are not exactly optimal because those solutions are going to be good in practice if your assumptions are correct.

It really shouldn't matter. Can you send us an email to support@piosolver.com? Preferably add a screenshot from the problem. We will try to solve it either by email or setting up a call with you.

Thank you for responses Punter!

The confusion in my mind is over. Now everything is clear.

I am wondering if there is a way to add a bet size but only on a certain turn card/cards. For instance, let's say the flop is A93r and on Qx/Kx turns I want to add in the option to bet 1.5P, but on all other turns I want to make it 3/4P, is there a way to make the solver do this or am I stuck having to enter 2 bet sizings for all turns?

What is the most efficient way to figure out optimal c-bet size? Is it merely trial and error? Should I add in multiple c-bet sizes and choose the sizing that is most frequently used? (Assuming I want to choose only one sizing)

When running a sim what solving accuracy should I use?

Also, when running a script is there anyway to run a script that will allow different bet sizes for each hand. For instance, let's say I want for board 1 to solve for a 1/3 c-bet, board 2, a 1/2 c-bet, and board 3 a 3/4P c-bet, is there any way to run a script to do this without adding multiple bets sizes making the game tree larger and therefore will take more time to solve?

thanks.

Hello i want to buy cloud server to solve preflop trees. Can you please tell me the optimal server config from contabo, or any service provider...
Is it mandatory to buy + Intel Xeon E5 2630v4 (10 x 2.20 GHz) for 39eur
What SSD to choose?

And why the f#ck are they charging 39eur for Windows option?

This is not possible. It was a design decision to simplify that. The gains are that features like line frequency always make sense and that some speed optimizations are possible. The cost is that you can't have different bet sizes on different cards.


Trial and error is the most correct way but solving a tree with many bet sizes and choosing the most frequent will usually give the same result so it might be a good shortcut. It's a good idea to verify from time to time if more than one bet size has significant frequency. By verify I mean running 2 trees with 1 bet size and looking at total EV to see which one is a better option.

It's a judgement call. 0.5% is quite a good overview, 0.35% has quite precise rivers already and there is probably no reason to go below 0.25%.

There is no convenient way to do that right now.

Unfortunately we don't have resources to keep with ever changing offers of cloud/server providers. The faster the CPU the faster it will solve trees. The more RAM the bigger trees you can run.

Microsoft's policy unfortunately. Maybe you can find another provider with older version of Windows which should be cheaper than that.

Hey guys,
I'm currently an amateur player however am looking to study intently and ramp up my game.

Would someone be able to explain the main differences theyve experienced with Piosolver Edge as opposed to the Pro? Have noticed Edge is quite more expensive, should I just opt for Pro?

https://gyazo.com/41568d133035f8b4b2f8b6824c83fc71

I am doing something wrong with Pio Preflop. Why it doesnt stop like normal tree? I choose 9 flops like 22gb ram.

Hi Punter,

I rented a server and I can't seem to solve some of my preflop trees to below ~8bb/100 exploitability. I have a 256 GB RAM server and the trees usually only require 65-120 GB RAM so I know it's not the ram. I believe I have 10 cores as well. Not great with computers but here are my specs:

Dedicated Server 10-Core (no setup) (38811)
Windows Server 2016 Standard Edition (64 bit), Dual 10-Core (377166)
+ Intel Xeon E5 2630v4 (10 x 2.20 GHz) (377164)
Windows Server 2016 Standard Edition (64 bit), Dual 10-Core (377166)

Do I even need all that?

Anyway, is 8bb/100 accurate enough? It gets to 8 BB/100 within 8-12 hours or so then just starts hovering around there or going back up to even 20 BB/100 over the next 12 hours so I give up and stop. I think I read you saying you need to get it down to 2-3 BB/100.

If you want to run preflop simulations or you own powerful hardware(more than 16 hardware threads) that you want to take full advantage of you will need the edge version. You can always upgrade to edge later if you need it in the future.

The basic version has all you need to solve and analyze postflop trees. The pro version is useful if you intend to use the solver a lot as it supports scripting (solving batches of many flops, for example 100 flops overnight), 2 computers and it's faster on higher end CPUs.
The edge version has the preflop solver this however requires very powerful hardware to run. At least 64GB of RAM and a fast modern CPUs are recommended. We make some pre-solved solutions available to all our customers as well:
https://www.piosolver.com/blogs/news/piosolver-1-10-13

You can load them with any of the basic/pro/edge versions.

Using % of the pot preflop is really not a good idea as it's not clear what the initial pot is. It's better to use bb/100 aiming at 2-3bb/100. On most trees the solver reaches 1bb/100 or below if you let it run long enough. Some trees are difficult for the preflop solver, things that make trees difficult are:
-very small flop subset (9 is really little)
-small initial ranges (3bet-4bet situations)
-too complicated postflop play (it's important to use simplified postflop play as that doesn't influence preflop ranges much and use saved memory to have more flops).

Feel free to send us the tree config so we can take a look and see if there is anything wrong with the tree.

To be honest I don't know. I've never used GTO+ nor do I plan to. It's the best to ask around for more impartial opinions.

Feel free to send us the tree config (support@piosolver.com) and we will take a look at the tree and potential reasons it doesn't solve well.

Hi Punter,
I have a couple of questions before clicking the buy button, thanks;

1 ) My threadriper 1920x has # of CPU Cores 12 # of Threads 24. Does 16 "thread" max on pio-pro mean I will use 16 or 24 thread of my 1920x. I know the term can be a little confusing as to what is a thread sometimes.

2) Does two activations mean I can work off either macbook pro as well as my PC? Or do I have to de-activate one each time to use the other?

Thanks in advance

Hello.
I want to calculate the task preflop in icm. He made several simulations on the river. On the river, when we do call, he always writes that he always writes it as 'ICM_PAYOFFS'. How to do something that counts in icm.
http://prntscr.com/n3ga6q
http://prntscr.com/n3gade
http://prntscr.com/n3gam0
http://prntscr.com/n3gaqp
The second question is how to access these files? https: //piocloud.******.com/download-links.html
Thank you in advance.

Hey.
I want to calculate the task preflop in ICM. I made several simulations on the river, when we call, he always writes that “ICM_PAYOFFS”. Accordingly, ICM is not calculated at all. How to do what is considered in ICM.
http://prntscr.com/n3ga6q
http://prntscr.com/n3gade
http://prntscr.com/n3gam0
http://prntscr.com/n3gaqp
The second question: how to access these files? https: //piocloud.******.com/download-links.html
Thank you in advance.

It will use 16 threads that is 12 cores but only on 4 of them hyperthreading will be used. It should be this way at least, if it doesn't work straight away there are options to determine which cores a process is allowed to use or you can always switch hypethreading off to ensure 12 cores are in use.

It means you can use it on 2 computers at the same time and move the license (within reasonable limits) for purposes like new computers, long travelling or some experimentation. It's also always possible to recover the license should something happens to the computer.

Unfortunately PioSOLVER is Windows only. We have quite a few customers using it on Macs but they all have a way to run Windows there either parallel to MacOS (Parallels, other VMs) or booting into Windows (Bootcamp).

ICM_PAYOFFS mean that EV is displayed in ICM instead of chips so it works as expected.

Copy-paste the link to your browser and download?
The passwords are there right next to the links.

cool. I'll buy it now, thanks for being active on the forums.

Is there a way to solve for all possible Turns at once?
It looks like postflop scripting is for flops only.

My goal is to get rid of hands with ~0.0001 weights on the turn. Any help would be appreciated.

Copy-paste the link to your browser and download?
The passwords are there right next to the links.[/QUOTE]

Maybe there is another link, this one does not work for me? ((((

https://prnt.sc/n3pqpw

If we're using 2 separate bet sizes for examples sake 100% and 33%

The sim pot bets AA 90% but the other 10% it bets 33% of the time

Does this mean when we calibrate the line using a 33% sizing on flop and get to the turn pio is treating our range like we have AA only 10% of the time here making our range much weaker?

If so, should we remove the pot sizing if we think its not optimal due to population tendencies?

For now at least you would need to generate the script yourself.
Please email us (support@piosolver.com) as well I will see if we have a piece of code to generate all turns to make it easier for you.

Maybe there is another link, this one does not work for me? ((((

https://prnt.sc/n3pqpw[/QUOTE]

Are you using those links:
https://www.piosolver.com/blogs/news/piosolver-1-10-13 ?

If yes, maybe Dropbox is blocked in your country (or by your ISP). If that's the case, please let us know (preferably email) and we might be able to send you a link to files hosted at another server.

Yes, but you will not lose EV because of it because you make it up by having AA in the pot bet line.

There is an argument for removing some options to make strategy simpler to play or to do what the population doesn't like seeing. Surely that's reasonable with a small bet. I am sure people are used to it by now but it's still way more tilting than a big one as the whole point of betting small is to make them fold those 15%-35% equity hands which they have a lot in their range and they can't do much about so it will induce some tilts sooner or later in less confident about their understanding of the game opponents.

I feel that way about the pot bet that it recommends almost every time in 3b pots on FD boards oop. I think PIO is assuming our opponents are raising with correct frequencies so it sizes up for equity denial when in reality nobody is doing that and we make their life easier by being able to fold more of their range to a pot bet.

How to get from a file of this format,
https://prnt.sc/n4drav
like this?
https://prnt.sc/n4drly
Thank you.

Hi Punter. I recently purchased the basic Pio but found that I would like to use the scripting function. Is it possible to pay the different between basic and Pro and get upgraded?

Well, Pio isn't assuming anything. It just outputs strategy which guarantees the highest payoffs against all possible strategies of the opponent. You're correct though that if you know the opponent is doing something different it might be more profitable to play exploit vs that.

Use Libre Office or Google Docs. In Excel you need to find delimited settings and set it to a comma as well as decimal separator settings and set it to a dot.
This is a problem with Excel. The files are .csv which stands for Comma Separated Values, Excel failing to open those correctly by default is just a huge fail of the developers and I encourage you to file a bug report and call them on this.

The upgrade fr new licenses is listed here:
https://piosolver.myshopify.com/prod...ic-pro-upgrade

Please either use the same email as before or let us know which license should be upgraded after the transaction is done.

I ran a script to do 2 solves (2 different bet sizes) on 95 flops.

I then opened a solve, clicked on the browser and clicked on the cbet portion of the tree. I then selected report "aggregated analysis over multiple files".

When I open this aggregation report the numbers in the report do not coincide with the numbers I see in the solve. The IP EV in the solve is 71.3 but the IP EV in the aggregation report is 32.3.

Can someone please tell me what I am missing?

The goal in these reports is to analyze preferable bet sizing on different boards by comparing the EV.

I'm a newby to 2+2 as well as Pio. Please oversimplify. Thanks!