Is it possible to use node locking 100% cbet strategy when using scripting to analyze multiple boards?

It's better to just make a tree without an option to check. You can remove a check by using add/remove line feature in tree-building form. I talk about it here:

https://www.youtube.com/watch?v=zLGpcZavxeQ (point 1)

A simple tree with a check OOP removed (3bet pot) is here:
https://pastebin.com/We4QAnn3

To load it you need to copy the text and then go to Tools->paste treebuilding config in PioViewer. Notice that OOP player can only bet on the flop (can't check) thus they are betting 100% of the range.

This may be a dumb question but it's an area I don't know much about. I've always used windows and never had any experience with other operating systems. But when it comes to cloud computing, Linux is always cheaper. Can PioSOLVER be run on Linux?

And an unrelated question. Is there some sort of node locking for pre flop? Let's say I wanted to lock my open range and my 4-bet range to see how an optimal opponent will play vs me. Is this sort of thing possible?

A slightly related question. If I was solving a RFI vs BB scenario. If I instead gave BB 80% of hands instead of 100% (if I knew BB will fold way more than 20%), would this speed up solving and still get the same final ranges?

Not right now. It's possible we have a Linux version for the solver itself in the (distant) future but the graphical interface is very unlikely to happen.

Yes, it works the same way as node-locking postflop.

In this case instead of node-locking it's better to make a tree with just one choice (to raise but not to check/call). It only makes sense to use node-locking when you have 2 or more decisions and you want the solution in all the branches.

If those 20% of hands are not in stealing range either it will be a bit faster and the ranges are going to be the same. It won't be huge gain though and the EVs will be skewed (as now those 20% folds are not counted towards EV of the raising player). Other than that it will work.

How do you use the preflop solver for limit holdem?

You can create the preflop tree (by the way, is every action needed? Or can you assume there is no folding to reraises and it still works?) but then how do you let Pio know that it's limit holdem and you don't need to add bet size percentages on the bottom right box but use instead fixed betting amounts? If I leave the no limit box empty and try to run it from the limit holdem window it says pot/stacks are not up to date.

Thank you.

The same way you would for NLHE.

You can add/remove actions at will and it will work. It's a good idea to cut lines aggressively especially with limit Holdem as those trees are going to be humongous no matter what.

You can use "c" postfix to tell Pio bet sizes are fixed amount and not percentages. For example 100c or 200c.

I've made a full preflop limit holdem tree config, here:
https://pastebin.com/451GhyMV

To load it in the viewer:

1)copy the text
2)go to Tools->pastetreebuilding config

Notice that to be able to run limit trees even on relatively small flop subset you will need to cut lines both preflop and postflop extensively. Full trees are just to big to run on even very big hardware.

I compared how different the results of 0.8% was to 0.25% in a single raised pot with wide ranges and there seemed to be less than 0.5% differences between frequencies everywhere, river as well. You said earlier that 0.5% already is not that accurate on river but there is practically zero difference with what I just looked at between these two.

I had one sizing for ip on flop and two sizings for both on turn and river, one raise size for every street, max 4 bets/streets. 100bb deep bb vs btn 6m ranges.

Here's gyazos

This is 0.8% https://i.gyazo.com/6c273a524a99e4d0...57e670d7c6.png

And this is 0.25% https://i.gyazo.com/c7a7e83a2c81aea0...e4086218ed.png

Pretty much the same, what spots did you mean earlier when you said 0.5% is not giving too accurate results on river?

Yeah, the solutions are going to be similar but i nsome spots EV differences on mixed actions are still big which means the balance isn't exact yet.

This is always a judgement call what is and isn't accurate enough. I recommend running more trees and comparing solutions, especially those on the river. 0.5% is surely enough to have very good overview of what is going on.

Hi punter, is there a way to some how transfer the data from a solved postflop tree (.cfr) to a excel (.xml) file? If not are there any plans for this?

You can import part of the solution (ranges or strategies at given spot, aggregation reports for given line etc.). There is also text API which allow programmers to import bigger parts of the tree to either other formats or to their tools displaying the results (our PioViewer uses that very API to show the results when you browse the tree).

There isn't and won't be a tool to import the whole tree to other formats. The reasons for that are:

1)The solutions already contain very minimal information, only what is necessary to recalculate necessary data on the fly. The saves don't even contain all the strategies and they don't contain EVs nor EQs (all that is recalculated on the fly when you browse). This means tha format like Excel sheets would need to contain way more information

2)Any human readable format is inefficient in comparison to binary we are using even right now (and we may implement compression of saves if needed although small/micro ones are already small enough for most purposes even if you are storing hundreds of trees). This means the saves would be at least several times bigger than ones in our format.

3)If you intend to build tools to work with massive amount of data in the solutions it's more efficient to use the solver and its API to do that than to import everything to largely inefficient format first.

You can crawl the whole tree and dump all the data using the API but it will take a very long time and take a lot of space (because of considerations in 1) and 2)).

Where could I find the details about the text API?

Here:
https://cdn.shopify.com/s/files/1/07...68544248083387

Please read the introduction at the top of the pdf as well.

Is it a big difference in performance between using 2166 mhz ram and 2666 mhz ram?

There is none. The solver is CPU bound, not memory bound.

Hello,
Ive just bought a new PC.
What are the steps to install PIO again (I bought the basic license).

Regards,
Orojanjivu

You need to reset a license on your old computer first and then activate it on a new one.
You can do it yourself, like here:
https://www.youtube.com/watch?v=5Ozjfnatwms (question number 3)

If you don't have access to it anymore please email us attaching your registration key and in case of pro/edge licenses approximate date of the activation (so we know if it's the earlier or the latter one).

Is it big the difference in performance between ryzen 1700 and 1800x? Is it worth it the 200$ price gap?

To be honest I have no idea, maybe try asking on our Skype group or on some hardware forums. In general Pio performance is going to be dictated by number of cores multipled by base frequency number.

Any update regarding the new product?

It's ready and then another one is almost ready but we need to change the way our licensing and updates work that's why there are delays. Monday or Tuesday next week are very likely for the first release.

What are the new products to be releasing?

Using Jesolver, whenever I load a solved tree and browse to the river, I get this error message :


Any idea on how to fix this?
Second question (maybe related?)
I've been looking for this option but cannot find it, where is it?

Thanks

I am not even sure what the error means (it looks like it's a generic Windows error so it's displayed in your local language) but you should definitely ask Oskari.

If you are using the newest version of the viewer then:
Tools->Configuration->Data presentation - other->Don't show average line frequency when browsing the tree

I hope that helps

Thanks, enabling this solved everything

Total newbie to piosolver so excuse my ignorant question. Trying out the free version before purchasing.
Can I have one range be my two actual hole cards and the other some defined range I am playing against and then specify the flop and turn and then explore different betting options after the turn for their EV?
I am just trying to re-create scenarios from actual play and see if I made the best pseudo gto move.