ok thanks a lot, now is clear
some others information about that? what about if i put my full saves on a ssd, it will change something significantly?

i will try...but, so is it faster using small saves or full saves for the aggregation on multiple files?

thanks for the answers btw

Yes, it should be significantly faster.

This one is impossible to answer. Small saves are very fast to load but the solver needs to recalculate rivers needed for the repots. Big saves already have rivers but they are big and loading takes time because disk is slow and building the full tree from scratch is not free either. It will depend on details but I feel that working with small saves and reduced recalc accuracy (even if you reduce it a lot it should be on par with full solution anyway, as the default one is very good) makes more sense.

I've never used the console before. After I press ctrl+b, what do I type in, to round flop, turn, and river for both IP and OOP to 1/1? I searched the UPI documentation but am still confused what to type in. (And I think you are right about it changing the exploitability though I've found the exploitability only changing by a small margin in my trees. Maybe I will also try rounding to 1/2)

Much thanks punter.

I am sorry it was an answer written hastily.
Rounding command has the following syntax:

round_up_to IP | OOP number_of_chunks starting_street

"|" means "or".

For example:
round_up_to IP 2 flop

rounds strategies of IP to 0-50-100 (chunks of the size of 1/2) starting from the flop.
If you want to round strategies for both players, you can do:

round_up_to IP 2 flop
round_up_to OOP 2 flop

This can also be done in scripts if needed.
It's more convenient from the console because it remembers the history of commands which you can scroll through using mouse wheel so pressing ctrl+b, scroll, click is faster than than entering everything manually.

Is it possible to set it up so that all my saves get saved to my external hard drive?

Yes, just provide the path there when saving files or in script generation window.

I have trouble when loading saved tree.

After I load tree from [File>Load Tree], pioviewer does not automatically change tree building parameters as the one I loaded. Tree building parameters or settings remain empty (no IP/OOP range, no betsizing setting etc).

I load full-saved ones then pioviewer used to auto-change tree building parameters but now dont.

Plz suggest solutions to show tree settings when loading tree.

There is a config file along with every save (it is created when you make a save). If the config file is till in the same folder as the save it will load parameters from it. If it isn't - it won't be able to.

Hey, I just bought this software, I have two incredibly stupid questions that I'm almost embarrassed to ask, but since this is basically the support thread here goes.

I think I understood virtually everything else from the FAQ/youtube vids/Runitonce vids but these two baffle me:

1) Where exactly do I see what accuracy% the hand has been solved to? I know that I can set it to solve it to x chips accuracy per hand etc, but if I just click "go", where can I see the progress level? What number am I roughly looking for to have it "good enough"? My laptop isn't the fastest so I'm trying to optimise my waiting time basically. I can't see any sort of indicator anywhere that would tell me that we have reached 99% accuracy or whatever?


2) So this is the more stupid question... I feel like there has to be an obvious answer to this. Since we just input ranges, pot&stack sizes, bet sizes etc in the postflop tree building & calculations... how does the software know the pf action (as in, who's the aggressor)? Example: I want to solve a spot where I 3bet the BB and BTN flats. So I input ranges, betsizes etc. Now the solver suggests that on 7-2-2 board I check more or less 100% because it seems to think that I'm not the aggressor. Basically everything I run it seems to suggest that OOP checks more or less everything and that IP is the cbettor, but I haven't found a way to tell it who has the initiative.

I understand there's the Preflop tab but I don't get how I use this combined with the postflop tab to explain the "full" hand history to Pio.


Thanks!

My antivirus found Trojan in free version. Trojan(HEUR/QVM03.0.0000.Malvare.Gen)

1) at the setup settings at top right, % of pot or bb/100, then at the bottom of window comes the bb/100 or % of pot which it has currently solved it.

2) #lolmttregs : DDD, You sure u have inputted betsizes for oop player?

In the tree building tab on the right, here:
https://gyazo.com/dd64d9e509f89be3ced01fd160042835

I think I talk about it in the quick start video

0.5% of the pot for rough outline, 0.35% for quite a good solution and 0.25% is probably a level you don't need to go below unless you really are picky about things being perfect. You can of course stop it, browse it for a while, and as long as you have the tree in memory resume solving to reach better accuracy.

Being an aggressor and acting in some way because of it is a purely human concept which has no theoretical justification. The only thing that matters is your range. It's true that even in optimal solutions (as well as in human play) the aggressor ends up with stronger range usually so they bet the next street quite frequently on most runouts. For the solver though it's enough to know what the ranges at the starting point are. If you get a 100% check on 722 board in a 3bet pot you likely entered quite a loose 3beting range and maybe too strong a calling range. The best way to get better feedback is to use "copy to clipboard" button in the treebuilding tab, pasting it to pastebin.com or similar service and linking the config. This way we can build it locally and take a look (by going to Tools->paste treebuilding config).

I think that makes intuitive sense when you think about it (the only implication of the previous actions at the point of making a decision is what kind of range your opponent has) so I am leaving it without further commentary for now.

Antiviruses are just programs which use a lot of heuristics (simplification) and contain bugs (like any software out there). The description of this warning even says that it was caused by some kind of heuristic. Feel free to email us about it and give us the name of the software so I will drop them an email about the false positive but my experience with that is that those companies have very little incentive to fix things and a lot of incentives to show as many warnings as possible.

I went ahead and constructed a tree fitting your requirements using stealing/3betting ranges form one of the tries to estimate them theoretically (BTN vs BB, 6max) and solved it to very good accuracy. I've got 90% cbet:

https://gyazo.com/6e1a395100dd029011f88e439388ca25

You can find the tree here:
http://pastebin.com/TGH6MuKs

(to use it, copy the text and go to Tools->paste treebuilding config)

It should be a good starting point for those spots. It makes sense to remove the hands

Does script command "show_all_freqs local" use all flop ,turn and rive cards to calculate for exampale:

r:0:b39:c:c:c:c:c - 54.507%

Yes, this is aggregated frequency for the whole line.
The idea behind the command was to output data you can compare to your database to see where population tendencies vary from optimal ones.

what is for PioViewer 1.9.2.4 the distribution Plugins directory ?

We only support very simple plugins now (the project of powerful plugins wasn't very popular and it was a lot of work to support it). I am not answering the question fully because I need to ask my friend about details and I will be able to do that tomorrow.

Hey, thanks for the replies to my previous noob questions, I'm quite comfortably pioing now. I just have one more question and then I promise to take my noobness elsewhere.

Is it possible to somehow edit the ranges midway through checking results in the browser? Like say (random made-up terrible example here just to point out what I mean) we assume that the board is KQJ8 and Pio suggests that OOP should bet KK, QQ, AJ and 66 and check everything else. Now we decide that there's no way anyone is ever actually betting 66 here, and we'd like to modify villain's betting range to just KK, QQ, AJ so that we could then more effectively decide what to do with our range.

I assume this is probably not possible and I understand this is now slipping away from GTO to CREV type stuff. Anyway I thought I'd ask since it can't hurt!


PS. As a random feedback, I really think you should make a detailed written tutorial at some point. There are lots of oldschool people such as myself who prefer written manuals instead of watching videos. It's also really difficult to easily find advice about specific parts of using the program from videos, even if you use timestamps. It must be tedious for you to create one but maybe you could hire someone or something, I really think a proper manual should exist.

Questions are always welcome

There is node-locking thing which sadly doesn't have very good interface for now. There is a bit more about it here:
https://www.youtube.com/watch?v=a_OhaxBi7mY

and here:
https://www.youtube.com/watch?v=gJEPGSIpIBM (question 12)

and a shorter video from our friend here:
https://www.youtube.com/watch?v=Nbx6OLxFxek

It usually makes more sense to solve tree from scratch rather than locking and resuming solving.

I think this is a bad idea. I realize it would be very convenient for a lot of users. The problem is that writing it and more importantly keeping it up to date when the software changes is a huge time-consuming and not very fun task. Manual which is not up to date (and therefore often incorrect) is worse than not having it at all. If there was a written manual decent enough to be published there would be way less functionality and features. Manuals are great for software that isn't developing much or is more like one-time release (like video games which don't get any updates). It's also a good idea for software with much wider audience (then you can hire people to write it).

Notice that a lot of software and devices these days don't come with manuals (only quick start guides) and the reasons for that are similar - keeping up to date manual is a huge overhead.
I made a decision that this is a very bad idea for our project and I realize it annoyed quite a few users. I hope the reasons above make them more understanding. That being said we are looking for ways to make the experience better so if things are unclear we try to answer quesitons as clearly as possible and sometimes provide a bit of additional description or videos.

is there a plan soon to improve node locking with "range explorer" type selection?

Understood. thank you.

It's one of the things we want to do but we don't provide timelines anymore (as they never work out anyway).

Is it possible to run back to back scripts and if so what would be the best way.

I have a question along the theme of the previous one I asked, sorry if this is stupid. I'm simulating this hand:



I'm trying to see what hands I should bet the river with and to what sizing. However Pio keeps assuming that I still have hands in my range that I would have folded on the flop. For example J7s is a 100% fold on the flop (according to both Pio and myself). But simulating the riverspot it now pops up in my range again even though Pio already wanted to fold it on the flop:



What am I doing wrong? I feel like these hands shouldn't appear in my range anymore since it obviously affects the strategy quite a bit whether the bottom of my range here is a random hand I already folded on the flop or not?

Yes, the way to do that is to create separate scripts you want to run and then create a script to run them both. See Skype group FAQ here:

https://docs.google.com/document/d/1...=h.cnb4hjen5qo

You are looking at the strategy right now. Strategy has to exist even for hands outside the range at given point because the solver needs to know what the hand would have done if it followed a given line (to compare its EV with other options and be able to choose between them). The way to combine strategy view with range view (so you get squares not filled in 100% for hands which are not that often in range) is to check "square size proportional to weight" on the right.

When we are at it you likely want to check "refresh on new selection" one as well (then it auto-refreshes when you browse the tree) as well as "bars width proportional to weight" (this means the fillings on 13x13 squares will be created according to current weight of combos in a square). Please see here:

https://www.youtube.com/watch?v=VhG_YsEvfwU (question number 4)

and here for the explanation of checkboxes:

https://www.youtube.com/watch?v=5Ozjfnatwms (question number 9)