that list would be insanely long, depending on stack depth.
donkbet sizes field being empty should have higher priority than "use only one size" checkbox

i m forced to use 1.8 til this is fixed, i probably have 1.8 backup somewhere

For each hand combo, could EV pls be displayed with a measure of variance in EV across all runouts?

Maybe I missed to select range .
Now when I go run script, I get message that I don't have enough memory (Missing like ~500mb). I got 8gb ram..

Hi,
In the browser tab, if I uncheck "square size proportional to weight", the solver gives the strategy for all hands (full range). How should I interpret this? I mean am i correct to say if "I somehow arive with this hand (although its not in the solution range) in this spot this should be its strategy"?

If by "insanely long" you mean like 4-6 lines then yeah. The good thing is that you need to only prepare it once and then just copy-paste.
I understand that you don't like the change and maybe you are right about expected behavior. It is noted.

That's one possible feature although unlikely to appear in the near future (there are many many more useful features and that one isn't a 30 minutes job to code).

Kill Skype and Chrome tabs. Try making the tree smaller (remove some sizings etc.)

Yes.
There is no practical difference between a hand not being in range anymore and it being in range with 0.00000001 weight. That's why it doesn't make the slightest sense to remove hands not range when showing EV or strategy (those don't depends on current weight of the hand). It does make sense to scale it though, that's why we make "square size proportional to weight" checkbox available.

Hi guys, I'm considering upgrading to the edge version. Anyone knows how much memory required for a typical preflop tree like SBvsBB (one bet size postflop, 74 flopsets, 50bb eff)?

It depends on some factors but if you just want raise or fold then call or 3bet then call or 4bet then call or all-in which is 3 preflop exits then you might be able to fit it under 32GB (on 40-50 flops) and definitely under 64GB even with additional options. Once you start including a limp 64GB is a bare minimum and 128GB would be recommended for bigger trees with additional options or bigger postflop play.

If you just want 50bb effective then 64GB should be fine for all practical purposes.

Hi, I have PioSolver Pro. When i first got the program about 4-5months ago I noticed that there was no updater file in the root folder. Its always been like this, not sure how to upgrade to the latest build as a result. Im currently on "build 058e759". Any advice would be a great help.
Thank you.

Hi,

I was wondering how to run an aggregated report in the new version. What I want to see is flop turn and river frequencies over multiple flops, but I can't input a range of flops anymore like I used to.

Thanks!

Does piosolver run faster with xeon than i7? The pro verson only supports 12 core, so I wonder if there are difference between using one or two E5 2683 v3 which is a 14 core 28 thread cpu and how much ram would you recommend?

What are the important specs for pio?
Does RAM speed/timing have a significant impact?
Is 4core (~i7-6700k) good enough, or is going for 6 or 8 cores basically 50% and 100% faster?
Looking for a combo below ~800$ (i7-6700k+64gb ddr4+mobo), but if an extra 200-300$ is significantly faster I'd consider it.

It was an oversight from me corrected in the 1.9 version. Please watch this video:
https://www.youtube.com/watch?v=1PEmsGaGLvk

I talk about the updater and how to use it for first 3 minutes (see the video description for a direct link to the relevant part)

I am not sure what you mean by "range of flops" the functionality for multiple file reports should be the same as before (+ some improvements/corrections). Can you tell us in a bit more detail what's impossible to do anymore?

Depends on the Xeon and an i7...

That will be faster than quad i7 but probably only marginally faster than something like 5820k (assuming pro version).

It's hard to say what RAM amount to recommend. If you already have a commercial version of the solver you can estimate the size of various trees and see for yourself. There aren't many practical postflop trees which don't fit under 16GB of RAM in my opinion.

It's probably a good idea to browse last few pages (assuming 100 posts per page) of this thread ctrl+f'ing hardware/CPU.

No, DDR3 @ 1600Mhz is fast enough.

The speed is proportional (more or less) to number of cores * frequency as long as we are talking modern Intel CPUs.

Check prices of 5820k. That will be both faster and very likely cheaper.

Want to purchase Pio but not sure if I should start with basic or go ahead and get pro.

If I purchase basic can I purchase pro for the difference in price at a later time or would I have to pay full price for both?

What are the differences? The site doesn't go into much detail. Just that it is faster??, scripting, and node locker. I've seen a lot of guys using the node locker on videos and seems like a good tool for exploits. Not really sure what scripting is. And is it really faster? Anything else?

Cheers

The upgrade from basic to pro is 249$ so the difference + 23$.

pro over basic:
-you can use it on 2 computers
-you can use scripts (to automate tasks, solve trees when you are not in front of your computer)
-there is a very slight speed difference for quad core i7s and significant one for hexa core CPUs; there is no speed difference for quad core i5s or dual core CPUs

This and all other features of the pro version (with the exception of scripts) are available in the basic version.

Can I safely move saved trees to an external hard drive without affecting functionality (simply by cutting and pasting the 'saves' folder)?

Yes but if you want the configs to load with the saves you need to copy .txt files as well as .cfr files (preferably just copy the whole folders).
I would recommend copying first, checking if it works and deleting the original files later instead of cutting and pasting.

How many iterations are needed to get reasonable approximate of EV for a hand (5bb/100) when solving preflop?

I have no idea. Answering this question requires a lot of testing on various trees and we don't have resources to conduct such tests.

How do you get access to the Skype group?

There is a link in your registration email (if your license is older then please email us and I will send it to you).
Sadly we have hit 600 people limit and it's not possible to join right now. We will reset the group as a short term solution after next bugfix release but longterm Skype group won't be a vialable option anymore as it's impossible to scale beyond that limit.

https://discord.gg/VNqehjD

^^^ A channel on discord which is a lot like skype (for pio)

I would love stop using a stinking buggy pile of **** that Skype is soI will check that out after the 1.9.2 release.
To be honest though the group with several hundred people is just too big and it's overflowed with support requests. Let me remind everyone that the proper support channels are:

0)Reading FAQs here and especially here as well as watching FAQ videos: part1 and part2
1)This thread
2)Email for questions not suitable for this thread (mainly licensing/transaction issues and business inquiries)

Asking on Skype has an advantage that 600 people see it and those online may take time to help you. It had a disadvantage that we are unlikely to see it and very unlikely to respond (we respond occasionally but we don't monitor all the chat). Asking here is nice because it's easier to link and more people see it that's why you are the most likely to get well thought response if you ask here.

I've generated 3 scripts and when I go to Tools -> Solver -> load script, I get message "ERROR: Command 1093644 not recognized." It is same number for all 3 when I load.

It really sounds like you are loading some .txt file which is not a script.
You can run the script directly after generating it: Script->run this script in the script generation window. If you are saving them - make sure you are loading exactly those files and not something else.

Just in case you haven't seen it yet, here are short instructions explaining how to run a simple script:
https://www.youtube.com/watch?v=VhG_YsEvfwU (question number 6 - see the video description)

If you are sure you are doing things correctly please send us the script to: support@piosolver.com (preferably paste it to pastebin first).

EDIT:
Guessing from the number it seems like you are trying to load a full save as a script. If you would like to load a solved tree just go to File->Load tree in the top meanu or drag&drop that file to the viewer.

Hi, I don't run PIO (Mac user, are there any Mac options beside Parallels?) but I am using the PIO weighted flop sets for calculation in CREV, so I thought I could post my question here.

I am calculating the EV of flatting 3-bets for a range X against a 3B range Y. I have set up a post flop game tree (100bb starting stacks and three bets/raises to get all-in) and I play the ranges against each other post flop with the CREV equilibrium solver for each flop in the PIO 95 flop set.

The post flop EV for a hand in my flatting range is then the weighted sum of the hand's EVs on the 95 flops, divided by the sum of the weights, right?

Now what puzzles me: How card removal effects affect the results.

When the ranges X and Y are given (especially when they are narrow) the probability of being dealt a particular flop changes. For an extreme example, if the 3-bettor has AA only, two aces are always out of the deck and the chance of getting an A x x flop is reduced dramatically.

To account for card removal effects in my calculations, should I recalculate the PIO 95 weights by multiplying each weight with the probability for getting the flop the weight is associated with?