Quote:
Originally Posted by p0ker_n00b
Hello - I have a lot of questions. I hope you can help me with them. I will also say that I have a study group of 4 or 5 other players and we are debating on using PIO or CREV (I'm the only who uses CREV and another friend uses PIO and he says I'm using the wrong tools - but my coach said CREV is very powerful if you know how to use it) - Just adding that bit in there as myself and my study friends are unsure about which to use and hoping you could elaborate on some of my questions to help us decide.
I would recommend GTO+ for this. It's to a large extent CREV2, although it's more focussed towards GTO research. It also covers most of what pio can do, as well as offering several additional features, such as tiny savefiles, the ability to create databases of trees (which are relatively small, due to the ability to compress savefiles), internal analysis tools and the ability to toggle card removal ON/OFF. Furthermore, contrary to pio GTO+ will always converge to 0%, whereas pio will often hang around 0.05% (this is particularly noticable for turn calcs). When it comes to speed we are comparable to pio; sometimes a bit faster, sometimes a bit slower.
Quote:
Originally Posted by p0ker_n00b
I see there is a tree wizard. I just manually set up a tree by adding every possible option of bet/check/call/raise etc with multiple bet sizings but then remembered again about the tree wizard.
Is it possible to make tree wizard use multiple bet sizings? I want to have 1/3 pot, 1/2 pot and 3/4 pot.
If I do this for flop cbet, then of course turn and river pot sizes will be different too.
So what would you recommend is best way to set this up? Should I solve for just flop situation and then try to create a new tree that starts on the turn with new ranges?
You can use GTO+'s advanced tree builder for this.
It includes multiple bet sizes, and many additional options.
Quote:
Originally Posted by p0ker_n00b
Also I wish to ask - if you set up exactly the same ranges with exactly the same bet size options into both CREV equilibrium solver (not GTO+) and also PIO, and compare the outputs, is it possible to have relatively significant differences but both outputs are still optimal/GTO/accurate?
If the trees are exactly the same, then the results will typically be very comparable. The frequencies will match almost exactly, although there may be some differences in exactly which hands are used as bluffs. It's easiest to compare this by using GTO+'s "Basic" tree builder and pio's "Version 1.0" tree builder. In that case, both programs will build an identical tree.
Quote:
Originally Posted by p0ker_n00b
Because I set up a tree where I gave the worst flop possible (JT8) for a SB 3b range and UTG defend range where UTG has a lot of 88/99/TT/JJ and 98s+ and where SB has AQ+ and QQ+ with not much suited connectors and CREV still says SB should cbet 100% of range where as PIO has SB checking about 25% of their combo's.
CREV also suggested that UTG should raise at a higher frequency when SB cbets a larger size which seems counter intuitive and PIO suggested the opposite, where when SB bets bigger size then UTG should raise flop less often. I know it's a bit difficult to answer that last question bc I haven't provided much details but yeah, I'll be able to post a bit more details I think if you can't follow my lack of specificity.
Most likely the trees were different. Trees consist out of hundreds, or even thousands of nodes. When using the "Advanced" tree builders for both programs, the trees will always be a bit different, and the results will be different.
Quote:
Originally Posted by p0ker_n00b
Another Question has to do with your comments about bet sizings not being important in a GTO solution. This naturally confuses me a bit. I would assume it would be better to run the exploitative solver if we can guess to any degree of certainty how villain will react to different situations.
Because you always see "pro videos" where they set up like 4 different PIO bet sizes and test to see which one PIO "likes more" and then they "use that sizing for their entire range"
This entire approach doesn't work. All bet sizes perform nearly identical to one another. Typically the difference in performance between two bet sizes is within 1% of each other. Quality of play is far more important. Versus an optimal opponent there's not much to be gained from bet sizing, but far more from playing well within the bet sizes that you've chosen.
Quote:
Originally Posted by p0ker_n00b
So if bet sizing does not matter much in a GTO solution, why would PIO pick one size preferably over another?
You're basically looking at static. If you slightly change ranges, or the tree, or even just solve to a different Nash distance, then different frequencies may be reached. Even if you were to pick the line that the GTO solution least seems to prefer, then the overall EV would at worst be only slightly below the overall EV for the "preferred" line. There's very little to gain here.
Quote:
Originally Posted by p0ker_n00b
Also if I wanted to do pretty intense solver work, would you say basic CREV functions will suffice or would you recommend GTO+ - and what really does GTO+ offer that basic CREV does not?
I would recommend GTO+. It offers an interface that is specifically geared towards GTO, as well as the ability to create databases, improvements in the GTO algorithm, small savefiles, more tree building options, and several other features. For a sumup check the bullet points that the bottom of this page:
www.gtoplus.com