Yes, only 0s and 1s make sense there. I think fractions are rounded to 1s although that's not a defined behavior.

It is easy to implement as long as you give up the ability to lock hands across the whole tree which we don't want to do because of some plans for the future.
The reason it wouldn't be possible across the whole tree are memory requirements (right now you need only one bit to say if combo is locked but with frequencies you would need at least 7-8 more).

We may add that in the future for limited use cases (like only being possible on first street) but it's not the priority right now.

It compares two actions. It shows EV of both and then the difference in favor of the better action.
On you screen the hands in red are the ones for which betting is better, tha hand in white are vry close and the hands in green are the ones for which checking is better.

Thanks for all the answers! Much appreciated

hello , quick question about node locking... , when you lock an opponent strategy to say fold more hands than he should on the turn , pio adjust your strategy into an explotative approach ? i mean pio will start to add more hands to hero's betting range right ?

Is your "solver install" video (https://www.youtube.com/watch?v=ahEuEa9oY2E) for the edge version only or for all of basic, pro, and edge?

In general yes but there are exceptions. For example if you lock 2nd decision of Villain (for example you lock too much folds after bet-raise) then Villain might start betting less in the first place so folding more after raise won't necessary influence them (the hands you locked as folds just won't be bet anymore so they won't make it to folding too much decision).
In short: solving after node-locking produces optimal strategy for both players assuming the play in locked decision can't be changed. They both do everything they could to exploit/adjust to this fact.

It's the same for all versions although only the edge version has more than one solver. Basic/pro versions have one.

In tree building, the "use only one bet size if there was a raise before". Not sure what this does. What does the % refer to?

As to the function of the button labeled ">": This reminds me that I've thought that the Greek letter delta (∂ or Δ) would be a better label for this button.

% refers to bet sizings in lines where there is already a raise made, for example:

check-bet-raise-call/...all subsequent bets.
This is useful if you want to have multiple bet sizes in some spots but 1 is enough in others. It rarely makes sense to have more than one bet size once there was a raise already (because the stacks are very shallow).

In the next release there will be a mechanism to construct any tree without much complications.

This is a good point

Would it possible to solve preflop in a way IP or/and OOP player enters into the pot with hand its EV > X amount defined?

It's not that simple because that would need to happen during solving (you don't know what EV a hand is going to have until you solve the whole tree). It's possible to program and hope it converges but it sounds like not a very useful feature but still not the easiest to implement which means it's not going to happen.

We are pleased to announce PioSOLVER 1.9 release. Highlights:

1)More powerful tree building form
2)Frequencies across runouts are not shown for preflop trees as well
3)Many PioViewer improvements/fixes

More here:
http://piosolver.myshopify.com/blogs...er-1-9-release

and here:
https://www.youtube.com/watch?v=zLGpcZavxeQ

Edge version is already shipped, pro and basic one will be available in coming days.
Have fun!

https://www.youtube.com/watch?v=zLGpcZavxeQ

20:25 Can you elaborate about why you do not recommend using multiple post-flop sizings for pre-flop trees? Would you still recommend not to do that even with a large amount of available memory?

-from the tests I've run it seems they have minuscule impact on preflop strategies; it's usually better to add more flops instead
-preflop solver is less precise and has problems if you made postflop trees very complicated; it mainly has problems with big overbets

It's better to use the preflop solver to get preflop ranges and then use the postflop one to get postflop play precisely.

I mean, you can do it and the 1.9 solver is a bit more stable and should handle those cases better than 1.8 one but still I just don't think much benefit. It's possible I am wrong about some cases though (it's impossible to test all so my view about impact of those additional sizes is formed by things I've seen which is not too many as solving preflop trees takes a long time).

Hello,
Is possible to remove the Add Allin option in "Remove Line"?
Which is the magical word?

"Allin" (without quotes) should work, try it

Here is the way I use the terms:
optimal - equilibrium, Nash equilibrium, a state where neither player could improve by varying their strategy

most exploitive strategy (MES) - a strategy which wins the most against fixed strategy of an opponent (their strategy at given time)

The term "GTO" was introduced (I think by Bryce) because many people in poker community use "optimal" to describe what we mean by MES and other people use "optimal" the way we do, the idea was to avoid confusion.
"GTO" is what we mean by optimal.

We all know tournament life is utility. Could preflop solver display for each hand frequency across runouts a player is loosing all his chips, ideally more than X defined?

There isn't option like that at this time although it might happen in the future.

Hi,

1)Regarding equity realization,does IP's EQR + OOP's EQR always sum up to 200% on a rakeless game?

2)And if we add rake to the solution,the reason IP's EQR+OOP's EQR won't add up to 200% is because of rake,right?

Thanks!

TY for 1.9 release!

Q.
1. Does Basic version viewing of preflop solutions allow generation of preflop hand charts?

2. Alternatively, can edge user include preflop hand charts in solution provided to Basic user who can then browse preflop hand charts?

Nice new features, TYVM.
I would still like to see a range picker which tells me what percentage of the whole range it is.
For example:
I would like to know what % of a betting range contains a nut flush draw.

There is no reason for it to sum up to 200. For example if IP's equity is 75% and OOP's equity is 25% but IP's EV is 90% of the pot nd OOP's EV is 10% of the pot then EQRs sum up to 90/75 + 10/25 = 1.6 (or 160%). On the other hand if equities for both players are 50%-50% and EVs are also 50%-50% then sum of EQRs is 50/50 + 50/50 = 200%

Yes.

Yes.

This is available for a long long time now. Just navigate to a bet and open Range Explorer. There are checkboxes you can play around with. Now admittedly there is not a checkbox for "nut flush draw" (only for a flush draw and combo draw) but you can pick them manually, like I am doing here: (2nd barrel on the river, BTN vs BB):

https://giphy.com/gifs/26BRBp8lTP9pr9c3e

I am showing that IPs turn betting range has 13.3% flush draws and combo draws but only 2% nut flush draws.

Hello punter,

1. if I buy a basic version of piosolver and later want to upgrade to pro or edge would I be able to pay a discounted price (just the difference would be nice ;-) ?

2. I presently only have a dell xps 15 9550 with intel i7-6700HQ @2.6Ghz + 16gb ram (maybe expandible to 32)

Would this be enough to solve pre-flop ranges up to 200bb EffectiveS, lines: 2.5bb-f/c/3bet 9.25 - f/c/4bet 24 - f/c only?

3. I am still confused about terminology, sorry: can your software find MES and Equilibrium (ie optimal, nash etc as per your previous post)? How do I know when is MES and when is NASH?

Best, great software

The prices for upgrades are the difference +25$.

That's a very nice laptop.

Laptop and preflop don't go well together.
You could probably fit about 40 flops for those trees with 3 exits under 32GB of RAM assuming 100bb stacks but I still wouldn't recommend it. If you attempt it make sure you have a cooling pad (laptops aren't really made to run many subsequent hours at nearly max CPU).

The solver shows equilibrium approximation for both sides (the longer you let it run the better the approximation is). It just shows EV of MES against that approximation so you can see how far away form perfect the approximation is. Notice that in the perfect equilibrium the strategies are MESes against each other.

thanks for your always prompt answers, my plan was to solve just to have the gto ranges for different sizes for OR-c/f/3b-c/f/4b-c/f for up to 200bb.

Then, having the ranges, run postflop solution with 100 flops for specific ranges matchup.

Would that be doable?

If not, can you give me a quote for running such calcs on your servers?

Best