Hi,I'm looking to rent a server or instance(idk the correct name) paying per hour(instead of monthly) to solve postflop spots only,so high RAM isn't really necessary.I've seen Amazon selling some but I'm clueless about which one to pick.Any recommendations?

Thanks!

When I click OOP equity and sqsz prop to wght checked, I get the non-yellow grid with all the hands at that decisionode, yes they have equities but if I want to look at only the raisingrange I still have to click on *raise* and then get the yellow grid again without equities, so I want to see in only 1 yellow actionrange the associated equities of the lowest equity hands. Now I first have to click on equity OOP, look at the equities of all the hands in a grid which shows hands w/o specifying which action they belong to, search for the lowest equity hands and then click on *raise* in order to see of they show up there in yellow OR to find out that they are folded. Basically when u click on An action and IT is shown in yellow, there I would like to see the equities. Do u understand my point of relevance in terms of efficiëncy? tY

I don't have any suggestions when it comes to cloud providers although I know our customers are using Amazon, Azure and I think someone was using Google thing as well. It's a good question to ask in a Skype group. Sadly I have 0 experience with those so I can't help you with the choice.

I think you didn't follow my suggestion in the last pot.
The way to achieve what you want is to navigate in to raise in the tree (by clicking it here: https://gyazo.com/044edce4d354bdb119c5be7688d4fd8a). After that you will see equities of hands in raising range.

You don't have to do that if you follow the suggestion above.

I do understand your point. I also think you are missing a way to do that which already exist. If you follow my suggestion you will get exactly what you are asking for: only hands in raising range with equities and weights (as size of the squares)

Your criticisms of the proposed method are correct, but the current "all in threshold" method doesn't truly reflect how people reason about bet sizes.

For example, the current all in threshold method means that the inclusion of a bet size depends on how the street was reached. In other words, rebuilding a tree starting from the river can actually result with more bet sizes being available even though the same bet sizes were used because 0 chips have been invested if you start from the river.

The problem gets worse when you start to think that poker isn't always a 2 player game, so the current method implies that a player would think differently if the node was reached with 5 players investing lots in the previous streets (then folding down to 2 players) vs if it was reached with only 2, even with the same ranges for both players.

There is an alternate method which is better than both the current and proposed method. A method that stops tiny bets (like your example) and that also doesn't have the problem of a the inclusion of a bet size depending on how much the player invested in previous streets.

I call it the Minimum Stack to Pot Ratio. It works as follows. If an opponent's call of the given bet or raise would leave the effective stack sizes less than a given fraction, then the bet or raise becomes an all in.

For example, suppose we input fraction of 0.2. This means we can never have less than 20% of the pot as effective stacks.

So if the pot starts as 100 on the flop and effective stacks are 200, and a size of 80% pot is input into the tree as a cbet size, this would become an all in. This is because (100+80+80) = 260 would be the pot after a call and 40 would be the remaining stacks. And since 40 is less than 260*0.2, it would therefore become an all in instead of an 80% bet being available on the flop.

I think this better reflects the way players reason about bet sizes, I do believe most players think "how many chips will I have left after I make this bet" rather than "how many chips have I already invested into the pot, and what fraction of my stack would have been invested after I make this bet".

Whoops, that example should read

So if the pot starts as 100 on the flop and effective stacks are 120, and a size of 80% pot is input into the tree as a cbet size, this would become an all in. This is because (100+80+80) = 260 would be the pot after a call and 40 would be the remaining stacks. And since 40 is less than 260*0.2, it would therefore become an all in instead of an 80% bet being available on the flop.

Yesterday i moved the pio folder to my secondary internal hard drive.
I don't know if this is related but now im getting the message "Server Error: ERROR: Command set_eff_stack not recognized" when i try to build a tree

I see your point but I don't really want to re-visit this discussion right now. There are more interesting things to work on and the newest Pio version allows for removing/adding arbitrary lines so you should be able to work around any issue resulting in a bit different all-in threshold definition than you are used to. I mean it, I just don't want to reshuffle the arguments which were made tens of times in the past already

Make sure you are connecting to the newest solver. Maybe you have moved some old version and now are connecting to that? That command was introduced I think in 1.8 so it's likely you somehow ended up moving an older version of Pio.

Hi, a nice feature would be some sort of "automatic node locking" to reduce computation time. By that I mean an option to calculate the flop strategy with only a few turn and river sizings, then lock that flop strategy and introduce more turn and river sizes.

I suggested a completely new method of doing it that no one has spoken about previously; the only reason I revisited the downsides to all-in threshold was to explain that my idea didn't suffer from any of these downsides. If my suggestion was implemented it would put an end to these endless arguments about all-in threshold.

Sure, but the point of PioViewer is to make PioSolver easier to use. I think using a Minimum Stack to Pot ratio instead of an all in threshold would make it even easier to use.

can i load a preflop solution on a basic licence? I dont have to solve anything just loading a solved preflop situation

Do aggregated reports over multiple files work correctly if I use different settings to create the files, i.e., different bet sizes for different flops?

Ok, I think your idea is actually good. I am not 100% sure that's how people think but that makes a lot of sense to me anyway. The thing is it's really such a minor part of treebuilding interface and the change is always costly (people have their configs and they are already used to the way it is) that we just don't want to spend time on it. I know you have a lot of attention to smallest details but from pragmatic point of view, at least or us, it's better to just not have that discussion. I hope you understand

Yes but you need to update to the newest (1.9.2) version and you may need to connect to the preflop-browser if the viewer doesn't do that automatically for you (it should).

The feature expects the same tree structure as to run a report it asks for results in the same line. You can cheat it a bit if all the trees have the line you run a report in but to be honest we never tested that so I would suggest trying it on small samples first. Of course if you run the report in a spot which only exists in some trees then it will be an error.

Hey,

I just moved to a new computer and I am having trouble to reactivate PioSolver with my once deactivated registration code because the personal download link in the first e-mail I got from you isn't working anymore.

Hey, where can i see the computer minimum requirements for the basic piosolver version ?

We had some problems with personal Dropbox and personal links in the past. Older licenses are more likely to be affected. Just drop us an email to support@piosolver.com with some identifying information (key would be the best, registered email is 2nd best) and I will send up to date information to you. If you are moving to a new computer you need to deactivate the license on the old one first, like here:

https://www.youtube.com/watch?v=5Ozjfnatwms (question number 3)

I you don't have access to it anymore, again we need your key and information which activation to reset (in case of pro/edge licenses).

All you need is a computer which runs 64bit Windows.
That being said to have any kind of fun with the solver you need a modern Intel CPU and 8GB of RAM. It is possible to run the solver on dual core laptops but it won't be that fun unless you are only solving really simple trees. I recommend quad core Intel CPU and 8GB of RAM as something which makes the program fun to use.

Guys,
Is it possible to do the following:
1) I have a bunch of models solved
2) I want to modify them via node lock and save (without resolving each time)
3) Run some script to resolve all those modified models and save them (new ones in new folder)

The goal is not to modify a model + resolve it + save as a new modified ONY by ONE, one after another
but to modify them all manually + save one by one and only after to run resolving for all of them later (at night f.e.)

P.S. How can i reduces number of threads used by pio ? my cpu is always running at 100% for hours, guess that's not a great a idea)

This is not easy. You would need quite a lot of programming to do that right now (to include node-locking in scripts, re-run etc.). It's one of the things on the to-do though to add some node-locking functionality to the script generator.
I am not sure your case will be easily doable though as it seems you want to node-lock different things for every tree which will make any kind of automation more difficult.

It's a good idea if you are using desktop computer with good functional cooling. It's definitely not a good idea in a laptop.
Anyway, you can control it in Tools->Configuration->Behavior (assuming you are using 1.9.2 version). 0 means default number of threads for the OS (so max), experiment with other numbers there.

2) Thanks )

About 1)
Aren't there some script's commands like: open X file, solve and after reaching Y accuracy save it to Z folder ? All i should do then would be to compile those commands in one script and run it

If you are willing to do node-locking part manually then yes, it's not complicated (drop me an email and I will write a sample script for you).
In general it makes more sense to lock on fresh (or just barely solved) trees as locking and solving on already well solved ones is both slower and may result in way worse accuracy (because the solver assumes it's already far into solving process). There will be an option to reset the tree (while keeping locked nodes intact) in the next version.

Hey, Im new to PIO solver and watched a couple of tutorials so far.

This video explains "square size proportional to weight" and it makes a lot of sense why Q8o is still shown on the turn in OOPs range if you dont tick the box.

https://www.youtube.com/watch?v=VhG_...youtu.be&t=614

I just analysed a simple BTN vs BB spot, where PIO suggests betting JJ almost 100% on T76.

The turn is an A and PIO suggests bet 17% and check 83%. But the JJ square is now mostly grey and only a little red mixed with green, what is the reason for that?

Dont we still have 100% of JJ in our range?

Thanks a lot for your help!

It's hard to say anything without seeing the actual tree. The best way to share the config is to go treebuilding and calculation tab, click "copy to clipboard" button then paste the config to a service like pastebin.com and link here.
This way I can copy it to my Pio (by using Tools->paste treebuilding config), solve it and take a look at your spot.

Did you go to the betting branch? My wild guess is that you went to a calling branch or you're looking at a different spot. Again, show me the config and it will be clear

Hey Punter,

I'm sure this is possible but I can't seem to find how to do it-using the preflop solver, is it possible to specify which player acts first (the IP or OOP player)? In HU the IP (SB) player acts first preflop and 2nd postflop, but often in modeling 6max hands I'd need to be able to indicate that the IP player was acting first.

how come i can't find tools configuration and behavior? where is it located for windows 7?

Yes, the one which has less invested preflop (those fields are used for blinds). If they have the same amount invested then OOP acts first.

It will just work because IP pays a smaller blind HU and in 6max situations it's BB who pays a blind.
Just to make sure you need to input blinds here:
https://gyazo.com/bf0daec8924eeea16bc09b9c385ebc6f

Make sure you are using the newest (1.9.2) version. If not the updater is here:
https://cdn.shopify.com/s/files/1/07...90236514835682

I go through updating process on a video, here:
https://www.youtube.com/watch?v=1PEmsGaGLvk (point 1)