Couple of feature requests

When looking using strategy option could the weights taken with a given action be shown in % rather than fractions of 1. Using hotness feature the strategy option shows % for bet/check for example 30% bet 70% check. When using strategy in the main viewer it is shown as 0.3 bet and 0.7 bet

Secondly is it possible to add an option so that when you click on bet is shows all the hands bet in a range and then following this when you click on Ev or equity it shows all the equities/ev of hands in the bet range. At present it resets to showing the whole range to get to that given branch.

This would be very helpful to quickly see what the worst ev/equity hands are bet on a given street etc.

thanks

This is available here:
https://gyazo.com/9a1bcf7adb32c1f18d4f0e60f517eec4 (Tools->Configuration in the top menu)

You can get that by just navigating to the bet branch and showing EV/EQ/range etc. there.

Again, click on bet in the top tree view, ask for EV, see where it's the lowest.

For a Startingstack Game 180 and Blinds 5 and 10. What does a OOP Total EV Number of 4,003 means? Does the oop player wins 4chips per hand or what is this 4,003 number?
regards

Yes, all EVs reported by the solver at in chips per hand (in Range Explorer we normalize it to % of the pot to make it easier to compare to EQ).

HI!

Could you explain me the most important difference between pro and edge version of Pio?

Is it possible to work with packs in pro version?

I ran a script last night, the trees seem to work fine, then when I do aggregation report I get this:

https://gyazo.com/a2869a7a469cf0f197003d08d54a7491

Then the tree doesn't work again until I reload it:

https://gyazo.com/7e54fbfe3446b6b14324dac4b658e402

https://gyazo.com/8fa996dfe9fda3fac338de7efbc132b2

When you do this though asking for EV shows the EV of the entire range to get to that branch not just that are betting. Unless I am doing something wrong or am misunderstanding what you mean.

For example here

Whole range to get to river

https://gyazo.com/4efd3538415ed7b232d5a43e913fe3a9

Clicking on bet 172 option

https://gyazo.com/728b3c5adeba515439c13d9f671ea288

Clicking now on EV which bring up whole range not betting 172 only

https://gyazo.com/c8d4a7c7385c9141416df590a4767c42

The main reason to get edge version is to use the preflop solver. You need very serious hardware for it though (it's recommended to have 64+GB of RAM and fast modern Intel CPU).
Edge version is also useful when you have dualbox setup with a lot of cores as it's significantly faster on those machines.

It's possible to open preflop saves (so packs as well) with both basic and pro version.

I ran a script last night, the trees seem to work fine, then when I do aggregation report I get this:

The first thing is to make sure you only have trees from the same script in the folder. It can only works then as an aggregation report assumes the line you are aggregating in is available in all the trees.

The second thing is to make sure you are using the newest version of the solver and the viewer (both 1.9.2). If that's correct then try downloading a bugfix for the viewer (1.9.2.4) here:

https://www.dropbox.com/s/gesnfhgrbo...iewer.exe?dl=0 (download and substitute in your Pio folder).

If that still doesn't fix the problem please drop us an email but it really should fix it so please double check

Yes, by "navigating to bet" I mena clicking small "bet" rectangle at the top of the viewer window (in the tree you use to go to lines), not the big one on the right which shows range for an action. Specifically here:

You clicked on the button which shows the range for bet 172 option but you haven't navigated there. To do that you need to click in the small "bet 172" rectangle in the top middle of that screenshot (on the right to yellow check rectangle). It will move you in the tree and then you can show EVs, EQs and all other data working already with a betting range.

Ah thank you I see now what you mean
It's a little unintuitive I guess and would be nice to do it how I wanted to. Perhaps it's a bit of a luxury for programming time.

Sent from my SM-G920F using Tapatalk

It's not only about programming time.
What you want is that clicking on a button which shows range for an action "moves" the focus to that action so then when you click EV/EQ it shows those in the place of focus. Imagine now that you want to go back to original action, there would need to be a button for that as well.

The way it is all of this "focus moving" functionality is integrated with tree navigation. We are in one place in the tree at given time and if we want to move to another place we navigate there and ask for results there. I realize the interface could potentially save one click for jumping between children of a node instead of going to children, back, to another children etc. but we see it as a relatively minor benefit as of now which would complicate and crowd the interface.

I hope that makes sense. Specific ideas for interface design improvements are always welcome though

how much faster can i expect pro to be?

currently my laptop has an i7-4712hq@2.3ghz, 4 cores, 8 threads. geekbench score is 3600 for single-core and 11400 for multi-core.

for a tree that is roughly 900mb, and with 4 threads in the settings, it takes me 18 minutes to get to .1% of the pot. my cpu percentage in task manager is 60%.

interestingly if i bump up to 8 threads in the config settings in piosolver, it takes me 16 minutes to get to .1% of the pot. cpu percentage being 85%. (thought 4 threads was max?)

i have a desktop that has a xeon e5-2670@2.6ghz, 8 cores, 16 threads. the setup i have scores 2500 and 14000 for single and mulit-core in geekbench.

for a 900mb tree, how much faster would it be if i were to run it on the desktop with xeon, using the pro version of piosolver with 12 threads? and for comparison's sake, the basic version on the desktop?

(bench for both laptop and desktop is 12 seconds, slighlty longer with the desktop, which makes me thing there's something wrong with it)

I've seen a lot of questions on node locking, but not really this specific one. Is it possible to set a node lock for all the flops done? For example, if they cbet the turn 100% of the time, can I lock that when I have the script run so they always cbet the flop and turn 100% for my sample? I'd like to see the ev difference. Thanks.

Basic version uses 6 threads as of now (I didn't update the description but it's compiled to use 6 thread since I think 1.8). 0.1% is a very good accuracy you likely don't need. Pio is quite fast to 0.25% but after that the progress is not so fast anymore, especially once you get close to 0.1%.

It should be 6 threads

If the CPUs are from similar generation the speed is proportional to number of cores * frequency. The bonus for hyperthreading is around 15%. Newer generations are faster than older ones but not significantly so.

In the next version pro will work with 16 threads so that won't be a problem on this hardware. It's hard to tell how much faster it would be, you can try to estimate from the information above.

Check energy settings, them being on is the main reason performance is not up to par.
Another reason might be some overeager resource hungry antivirus doing "online scans", "online file protection" or similar things which suck a lot of performance out of the machine.

Not right now but...

For 0% or 100% type of actions you can just build a tree with only this action allowed.

Sure, just remove the option to check. We added an interface to remove/add arbitrary lines in 1.9 version. Please see here:
https://www.youtube.com/watch?v=zLGpcZavxeQ (point 1)

The problems start when you want to lock general type of condition like "always bet top pair on the turn". This is not available as of now.

i'm definitely not the most knowledgeable about computers.

if the you compare two cpus of the same generation but one has twice as many cores, the pio bench of a cpu with double the cores should have a 2x quicker time?

or let's say you have one computer as a constant, with pio basic and pro. the computer has 6 cores with 12 threads. would pro be 2x faster, plus 15% on top of that?

though on my laptop, 6 threads was faster than 4 threads by 2 minutes out of 16, so an improvement of 12.5% with 50% more cores.

would be great to test out pro for a day if that is possible.

If they are clocked at the same frequency then the answer is yes - almost 2x quicker.

If it's the same computer it will only be around 15% faster with pro version. Using 6threads instead of 12 on a hexa core i7 (for example) is about 15% slower (in case of Pio) than using 6 threads even though task manager shows you 50% CPU usage (this number doesn't really mean anything in the era of hyperthreading).

Hyperthreading is confusing (and on purpose, it's a good marketing to make you think you have 2x cores than you really have).
You have a quad core CPU (4 cores) on your laptop. Using 6 threads instead of 4 isn't 50% more cores. It's the same number of cores but two of them now run 2 solver's threads. Performance gain of 12.5% makes sense in the context.

We don't have trials in general (free version is for that) but if you drop us an email with your current key I will have something for you.
I would advise against buying pro version for performance reasons only unless you have 8+core machine to run it on (or you want to run it on two machines).

Did you get any information from your friend regarding powerful plugin?

Ok, so right now there are really minimal but we are willing to listen to what you would like to have. Please drop us an email with some usecases/desired functionality and we will see what can be done.

I am not sure how to do this, I've only known how to set it to force flop bets, but I have no idea how to lock it to always continuation bet the turn as well IP or OOP as 3bettor, thanks.

I'm getting to know all the features of pio and have a question about node locking. I tried to experiment with it like you did in the tutorial video. I took a hand and node locked (intermediary question, what's the difference between selecting "lock all hands" and "lock selected combos" in the node locking window?) a tighter raising range for oop player (in a hand that would be a check/raise). As I understand, all following actions don't adjust to the new strategy automatically, right? So I have to rerun the calculation to get the optimal solution again. I did that, but then while my strategy was MOSTLY in place, pio took some combos of the few hands that I locked as raising hands and moved them into flop donkbetting range. Why is that?

Another question. Lately I feel pio has been calculating hands slower than before so I thought that maybe somehow the cofiguration has changed. As I understand the speed increases the bigger the number of threads used. What number should there be and how can I find the OS default? I have i7-6700 CPU @ 3.40GHz processor and the performance of other programs hasn't changed, i.e. everything works well

I linked to the video explaining that here:
https://www.youtube.com/watch?v=zLGpcZavxeQ (point 1)

Just in case here is an example configuration where OOP player always bets the flop and if called always bets the turn as well:

http://pastebin.com/re7iSBTP

To load it in your Pio please copy the text and then go to Tools->Paste tree building config from clipboard

Notice this part, which is responsible for checking option being removed:
https://gyazo.com/0094b8d97c75785d4567ad10a9ffb941

When you lock all hands the strategy for all hands is set to what you chose in node-locking view. When you use "lock selected combos" checkbox then you need to also mark combos which you want to be locked (by choosing them with weight 1 in the range selector that appears when you click a small square next to the radio button). The remaining combos are left for the solver to modify during solving. More here:

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

Yes, you need to run the solver (click "go") again.

You probably didn't click "set strategy and close" when leaving node-locking window byt for example cancel or maybe you just close the window. The way to verify that things are locked to what you wanted is to look at the node in the browser and verify that:

1)strategy is set to what you want
2)there is "LOCKED" flag in node description, here:
https://gyazo.com/f1fc010bba214e6e2036d2ea263b57e4

The most common reason for this complaint is that you are just building bigger trees. Pio started using way less memory in 1.8 version so trees which are 2GB now are way bigger than trees which were 2GB before. They also take longer to solve. Pio 1.9 is minimally faster than previous version. Things to verify in your system:

1)make sure energy saving settings are off
2)make sure no "online scanners", "online file protections" or similar anti virus features are running as this is huge resource drain
3)open task manager, sort by CPU usage and see if there is no application taking a lot of CPU when the Pio is running

Putting 0 there means using OS default.

It's very rare for a program to be CPU bounded (that is very rarely CPU is the limiting factor). For example for things like databases it's usually disk speed and even if it's not the program is only CPU bounded if it's able to use all the cores which is rare.

I am a new user so I apologize if this is a dumb question. When I start to solve the tree in Pio, I don't get any updates on how exploitable the solution is until I stop the solver. In the Getting Started video, the narrator seems to get updates every 5-10 seconds as the exploitability decreases. I see no updates. Any ideas?

hey can anyone help me? I can't seem to open the pio updater no matter what, I downloaded it but whenever I try to open it the sand timer next to the cursor flickers then nothing happens, anyone know how I can fix this?

Maybe your tree is very big and/or CPU not fast enough. Try solving some very simple tree first (preferably just river one) to see if the updates appear and work from there.

Random behavior like that is usually caused by buggy software having admin privileges like Avast (and some other antiviruses but Avast is notorious for hanging other applicatoins). Make sure to disable that and report if that helped.

So it actually was giving updates, just very slowly. I was just solving the QhJh2 s flop for limit holdem. I let it run for about 15 minutes and it got down to 0.35% of the pot per hand exploitability. I am running a 2012 Macbook pro via bootcamp.