You should pretty much just always use the GTO solver, given that it will figure out realistic counter-play for villain in the parts of the tree where you haven't defined play for him.

Well, the trees are different, so you will get different results. As for using the Max Exploit button, you should most likely just always have all three options (flop, turn and river) turned ON so that the max exploit play is calculated throughout all possible runouts of the board (unless you have some specific reason for only wanting to do this on the flop).

Other than that, I notice that you're using checkdowns to finish up your tree, as well as a custom written line in the bottom screenshot. It's probably better to use the tree building wizard here to finish up the tree for you with realistic play. Then, lock any play in your tree that you want to remain fixed and run the GTO solver to figure out the play in the unlocked parts of the tree. For a demonstration of the solver, the tree building interface and how to lock items in your tree, please watch the two videos here: http://www.cardrunnersev.com/download.html

Thx for your answer. Setting up a node-lock strategy for all possible outcomes take a huge amount of time because its so complex. :-/

I was playing today a little bit around with GTO+. Can I bypass maybe the problem with filling out the whole tree if I analyze like this:

At first I let GTO+ calculate the flop strategy vs. a node-lock and then
I checkout with what range I have reached the turn. Now I choose a turn card and then I set also a node-lock for turn. After this I click in GTO+ on the little circel in the upper right corner of the software
to let it calculate the turn and then I checkout the turn strategy.

This would save up a lot of time because I dont need to set a node-lock for all outcomes.

But is my method in GTO+ accurate enough or will I get bad results?

You don't need to enter play for every single decision. Instead, you can just leave most or even all of them unlocked and let the solver figure out the play. That's the great thing about the GTO solver. You don't need to define villain's entire strategy; you can just fill in whichever play you feel confident about and let the solver figure out the rest.


Yes, this is one of the reasons why we are moving the GTO functionality to GTO+. The CardRunnersEV interface was originally designed as an EV calculator, and not as a GTO solver (this concept did not exist yet at the time). Although the GTO solver is a great addition to CREV in terms of EV calculations, we felt that our users would appreciate a more GTO targetted interface as well. With GTO+ it's all now as simple as what you did here, namely you can easily analyse situations street by street, fill in any play you see fit and re-run the solver to get the updated results. So what you did should work perfectly. On top of that, more analysis functionality will be made available in GTO+ throughout the coming months.

Ah i see, so if we are OTT or OTR and cards have been set we use max exploit tool?

Pretty helpful posts above i didn't see until now...

So use max exploit tool only AFTER running nash solver to see EV's, and whenever node locking assumptions only EVER run the nash solver and max exploit is never used with node locking?

If running max exploit tool make sure to check flop-river?

Lmk if i'm understanding this correctly/anything i said was wrong plz

The easiest thing to do here is to just use the solver. You don't really need the max exploit tool for what you're trying to do. The great thing about the solver is that can do most of the work for you when it comes to building trees and defining play. All you need to do is fill in the play that you want to see in your tree and leave the rest unlocked.

So:
1) Create a tree
2) Fill in any play you see fit into this tree and lock it
3) Run the solver so that it figures out play for you in all parts of the tree that you have left unlocked
4) After running the solver, press F7 to calculate the tree and see the EVs

So there's really no need to go though the effort of defining play throughout the entire tree. The solver can basically do all of the hard work for you, especially when it comes do defining play on the unknown turn+river, which can otherwise be quite challenging.

Other than that, in our new software GTO+ all of this is made even easier. Navigating lines through trees is very intuitive, all data is recalculated live and on the spot and there's no need to ever even press a "Calculate" button. For a more detailed description of GTO+, please go here: http://www.cardrunnersev.com/gtoplus.html

Hi,

If I run a preflop sim in cardrunnersEV with 3+ player, does the software take blockers for hands not folded into account when using Monte Carlo to find equity?

Is there any difference if I input an opening range for BTN and make him fold all other hands, and then make a sim on SBvsBB after BTN folds and if I don't include BTN at all. I'm not really using it for that spot but it's easiest way I found to clarify.

Yes, that's the great thing about Monte Carlo. Everything is taken into account, including card removal. So for example if you have a six-handed situation with 4 players folding, and you choose to simulate those players as well then you should be able to compare how this card removal affects the EVs, equities and frequencies for the two remaining players. In some cases the effect of card removal is significant.

With just one player folding I would not expect the effect of card removal to be very influential, however, it will most definitely have some sort of an impact. To include BTN, just make him fold with his folding range and then enter the tree for SB and BB as you see fit. The effect of card removal by BTN will be fully taken into account by the simulation. See the pic below for an example.

Hi, do you have any Omaha products?

No, at the moment our efforts are focussed on GTO+.

Right the confusion for me seems to be which answer/EV's are correct using the nash or max exploit tool results.

Here are some more photos...

http://imgur.com/8Glo8mp
http://imgur.com/zoHPkAp
http://imgur.com/iNL8HUp
http://imgur.com/MvuLSF6
http://imgur.com/lZoThSi
http://imgur.com/B7CxUqc

---

So i ran the max exploit tool after the nash solver (that's how to use max exploit right?) and it had MP/hero raising @ 47.3% and calling @ 11%.
So my question is does the max exploit strategy yield more EV/is more profitable then nash solver strategy of raising 53.2%?

What's your stance on 3players+ models? I know that equilibrium exists but it isn't valuable as HU.

There's really no need to use the max exploit tool after running the solver. The solution you get by running the solver is what you need to be looking at. Now, admittedly, unless dEV is 0 then max exploit will give a slightly higher EV versus villain, however, the max exploit strategy is very susceptible to being exploited itself. If after running the max exploit tool you were to run it for villain (with his locks removed) then you'll find that your EV drops significantly compared to the GTO strategy. A max exploit strategy leaves you open to counterplay. The great thing about the GTO solution on the other hand is that villain can't do anything against it (he can of course play max exploit, but this strategy would again leave him vulnerable to your counterplay). So, if possible, you should just use the GTO solver.

Is it normal for the new math engine to take a lot longer than the old one? As an example, in a flop spot with unknown turns/rivers, an EV run is taking 25secs with the new math engine disabled and 2:38 with it enabled. I'm using a threadripper 16c/32t with 32 threads set in the 'set number of threads' box and it seems to be using all 32. Are there any disadvantages if I use the old engine?

My efforts are currently focussed on the development of GTO+, so unfortunately I have not had the opportunity to look into this subject in greater depth. Perhaps I'll have an opinion at a later point in time.

No, that sounds very unusual.
Is this a GTO solution?
Have you tried setting a lower number of threads?
And does this occur in every spot, or just a particular one?
Also, can you send a savefile to support?

Thanks for the continuous updates over the years!

I have a few observations / suggestions regarding GTO+.

Currently importing a hand range string from Equilab doesn't work correctly, some hands are not being recognized because the syntax is different from Cardrunnersev.

I would love to see a bit more distinction between various lines, especially when it comes to the blue/green/red colors. First of all, in different situations green is being used for both check and call which confuses me a little bit. Second of all, it would be great if there was already an indicator at the point of betting or checking if this is going to be a bet/call, bet/raise or bet/fold, but most importantly if it is going to be a check/raise. This would help navigating the hand a lot.

Lastly... how feasible would it be to have an option to turn the game tree into pure decisions only when the equity is nearly the same for betting and checking (maybe with a threshold that you can input yourself)? Currently there are just a ton of hands that for example are being bet 90% of the time and checked 10% of the time. This is not very practical and it makes the distinction between "Entire decision", "Bet" and "Check" nearly pointless because both lines still show up in the bar.

Thanks again

The preflop menu was the first menu that we designed, and, given that we did not have a full understanding at that point how the rest of the interface would end up looking/working (it's a creative process, so any plans at the start of the project are subject to change) we have decided to leave out some details (a predef submenu is another example), in order to circle back to it at the end of the design and fill in everything that remains. So, indeed, some details are missing here and there throughout the interface, but we'll get to them in the end.

We expect to offer customization in this area at some point, however, we will probably stick to using the same color for check and call, given that check/call is simply the passive continuation of a line.

We get this rounding request from time to time, however, in my view it achieves nothing in making the ranges easier to interpret. GTO solutions are admittedly hard to interpret by the human brain, however, this would only achieve switching from one system that's difficult to understand to a different system that's difficult to understand. An additional downside is that important information is lost. The reason why certain hands are a mix between two actions is because the EV for both lines is the same. So the player does not care which line he chooses. Rounding loses this very important piece of information. Even if the solution says 99% bet and 1% check then checking 100% of the time will still give you the exact same EV. The mix is not there as a preference for a certain action; it's only there to set up GTO frequencies for the subsequent lines. When trying to make sense of sets of large datapoints the correct approach is to organize the data into categories and graphs. For example, when displayed in terms of "top pair", "set", etc to me personally the data becomes a lot easier to understand. And a strategy like "check the stronger top pairs, but bet the weaker ones 50% of the time" is quite a bit more easy to remember as opposed to trying to memorize what to do with individual hands.

This should be quite easy to see after one of the upcoming updates where we will add certain tree navigation functionality.

Thanks for the response, looking forward to the updates

Pretty sure this was asked before but dont wanna go through potentially dozens of pages...how close are the GTO solutions from CREV in comparison to PioSolver?

The results are exactly the same; there's several GTO solvers out there and they all converge to the same results. You can easily compare for yourself by comparing turn/river spots, which both programs offer for free. In order to ensure that the trees are the same, please use the tree building tab "Version 1.0 style" in pio. Other than that, you may prefer using our upcoming tool GTO+ for this comparision, given that it is our latest iteration and has an interface that is dedicated to GTO research. For download links and further information on GTO+, please go here: http://www.cardrunnersev.com/gtoplus.html

Would GTO+ support solving preflop trees in future?

Hello ,
newbie question. I have to PM scylla in order to buy license for CREV , no reply from my 2 mails yet. ( sent in the contact list here http://www.cardrunnersev.com/contact.html)
I have try but i m not able to send PM. How can i do?
Thank you , and apologize if i have wrong forum section.

We are perfectly capable of building a preflop solver at this point, however, this is a feature that we are not particularly interested in. It only solves heads-up and requires a custom built computer with roughly 100GB of memory. We most certainly could, however we prefer to focus on offering our own unique functionality. There's really no loss on the end of our users here, given that all you would wind up with would be one more preflop solver; I've lost count, but I think there's about 4 of them by now. Building yet another preflop solver would mean having to take resources away from other areas where we hope to actually offer something new.