Yes, that should work perfectly fine.


No, that's not necessary. You can use GTO+ for this. Simply enter your tree, and any play for villain as you see fit, lock the play for villain in the relevant decisions and then run the solver. GTO+ will figure out how to best exploit villain's leaks.

Hey Scylla. Enjoying using GTO+ for some basic solving so far but I'm slightly struggling to get quick solutions when I use multiple bet sizes (mostly the fault of my less than stellar PC) so I have one (possibly stupid) question:
When I solve something to 0.5% dev, say, and then lock a decision (say I stop it from donking 1.2% of the time OOP) and run the solver again, after the first iteration, it's back at 50% dev or something. Surely it's possible to retain the majority of the existing solution without such a big deviation, or am I missing something? Does it start from scratch? Is this necessary?

If at all possible try using the option "With only 2 bets left, get the money in smoothly" for turn/river play. Your trees will be smaller, solve faster, while the solution itself is barely affected.


In this case you may instead want to consider using the editor to remove the branch entirely. With the tree being smaller, it will require less memory and solve faster.


Retaining part of the solution will not significantly affect solving speed. The solving algorithm is a learning mechanism that needs to figure out the complexities of the problem. Providing it with a solution that worked in a different spot (even if slightly different) may even result in as much as it needing more time.

Cheers for answering my question

I have been using this option - it also reflects how most of my opponents and I play, so it's a good feature.

I have done this a few times. Doing it for the hand I was looking at yesterday, I realised that it was giving the OOP player a load of turn/river donking sizes when I only wanted them to have those if they checkraised the flop - is there an easy way to distinguish between these OOP bet out spots in the tree builder without digging around in the tree? If not, it would be very useful.

It's probably easiest to just build the tree so that it only has single bet sizes for the turn/river (and flop, if you so prefer). After that, go to the editor, load the tree with "IMPORT TREE", add whichever bets you like to the check-raise line. Then, click on "ACCEPT CHANGES" to accept the newly created tree. For a demonstration video on the editor, please watch the third video here: www.gtoplus.com/videos

Im a little confused by starting pot and effective stack, like is it in bbs or raw chips. For instance to analyze a 1/2 hand where I minraise button and bb calls is it pot 9 and stack 196 or pot 4.5 and stack 98. Thanks.

Either approach will work, as long as you consistently apply it everywhere (so for example you can't use bbs for the stacks, but cash values for the bet sizes; it needs to be consistent).

GTO+ is pretty amazing for the price. I can’t think of a good reason to spend 4-7x on PIO when this product exists.

One thing I was curious about though is if we could see the betting frequencies of specific hands on the turn / river report. We can see equity / EV of specific hands within your range but that’s not quite the same. It would be really interesting to quickly look at how certain hands play various turns and rivers.

Is there a way to get it to solve multiple hands while Im away from the computer. I have an old laptop and solving takes forever. Would be nice to set up some hands and let it solve them while I sleep.

This is already possible. When running the solver, select the option "Medium storage" or higher. In the report, to see data for individual hands, select the "Specific hand" tab. See the screenshots below.

Yes, first of all, you can create a database of files, and then solve the database. This is described in video 2 here: http://www.gtoplus.com/videos/. Other than that, it's also possible to solve all files in a given directory. For this, click on the "Directory" icon in the upper right of the interface (see screenshot).

When you enter a specific hand it only tells you the EV / Equity on various cards but not the bet / check frequencies. I'm looking for the latter so I can analyze barreling opportunities with specific parts of my range on various run outs.

Also, I think a filter statistic on the river for missed OESD / Flush Draws would be useful for determining which kinds of hands are give ups and which follow through.

Does difference between dEV of 0.5 and 0.25 is significant in practice,particularly in micro limits?I assume 0.5 dEV is solved faster.Should we worry about losing 0.25% of pot or quality of play?

U can't copy solver play anyway so u better find logic when balancing combos.On river I guess it is all based on choosing a bet size and balancing in same ratio.Missed oesd's r betting in correlation with what hands r blocking.Missed flushes probably not bet at all in 3th street because they also block chasing flushes.It depends on spots...

I'm well aware of the limitations of solvers, thanks.

I was talking about limitations of humans

A dEV of 0.5% should be fine; the play has already converged at this point.

Ok, I will take it under consideration.


We can add it, however, when it comes to poker, it doesn't matter how a certain part of the hand was reached; all that matters is the current situation (stacks, pot, ranges). When it comes to river play, what is mostly important is whether your hand has showdown value. Typically OOP will bet his nut hands an non-showdown hand, while checking the rest. See below for a pic for a betting range on the river. We have selected the "Bet 29" tab for OOP on the river. River play will always almost exclusively look like this. OOP will always bet his hands with equity of almost 100%, but will balance this by bluffing with hands that have roughly 0% equity. For river play, showdown value is the leading quality in a hand to look for when deciding whether to bluff or not.

Glad to hear it!

I think this is a bit of an oversimplification. In many spots betting all the hands that don't have showdown value will mean drastically overbluffing because our value range is too narrow. In that case we have to be very careful with which hands we continue betting with and the solver is definitely giving up with air a pretty significant portion of the time from what I've seen. It would just be interesting to quickly look at which of your big draws that missed happen to be suitable combos to bluff with and which are not and try to understand why. That's one of the spots where solvers shine, imo.

I can't open GTO+. When I click on icon to open I get error message that says "Failed to create empty document" ?

Please help

Is there a way I can use CREV to work out push/fold charts for short stacked 6max cash games (taking rake into account)? If not I think there is at least a way to work it out for SB/BB? Can you point me towards a tutorial?

SB vs BB is possible. For this just set up a heads-up push-or-fold tree (see pic below). After that, set the stacks with Ctrl+F2 (or by clicking on the 'Stacks' icon) and set the blinds/rake/etc with Ctrl+F1 (or by clicking on the 'Blinds+rake' icon). Then, run the solver (F1 or click on the yin-yang icon in the toolbar) to get the solution. For multiway play you may want to consider using the historic data from your tracker. In that manner the ranges you get will actually reflect the games that you play in.

That was a bug in one of the earlier releases.
Can you please upgrade to the latest version?

I did not mean that you should always bet your air. On the river, it doesn't matter what draw value a hand had previously. A 6 high flush+OESD draw on the turn is simply 6 high on the river when it misses. It's more of a human emotion to still think of it as a draw. What you need to look at on the river is showdown value of your hand. Given that 6 high has almost no chance of winning, it's a candidate for bluffing (that doesn't mean to bluff it all the time; just that it's a candidate). An A high missed flushdraw on the other hand is probably better placed in the checking range, given that it has a chance of winning in a showdown. So, when composing your ranges, the leading factor in deciding where to place your bluffs is a hand's showdown value. If you need to bluff 20% of the time, then the bluffs should first and foremost be chosen from the non-showdown part of your range that has almost 0% equity anyhow. Betting will be the only way for those hands to still have a chance to win.

The selection will be due to showdown value and to a lesser extent card removal. The solver is a mathematical algorithm. It doesn't know that the hand was a draw in the past. Even moreso, it doesn't even know what a draw is. That being said, we can consider adding past draws to the list if that's what you're looking for.

To be clear, I'm not disputing any of that. I just think it's pretty well established knowledge that the worst hands in your range are some of the better candidates to bluff with. The reason I'm curious about how missed draws play on the river is because we are barrelling with them so frequently with our equity that they make up a pretty significant portion of our non-showdown range and it would be interesting to quickly assess which we are deciding to bluff with and try to figure out why, as well as detecting various patterns after using the solver more. For example, if straight draws were consistently betting more often than missed flush draws, that could be useful information. Of course we can already see that to an extent, but certainly adding that would streamline the process. Wouldn't put it at the top of your to-do list but if it's not very difficult to add in I would certainly appreciate it.

To clarify, I would be looking to see the range of all missed draws which would include flush draws, OESD's, and gutshots, even if they rivered somed sort of showdown value.