The type of disc that you're using for storage will be of influence here, with an SSD being faster than a regular hard drive.

Currently it shows 3.9 combos out of the available 6 combos. I can consider adding another output toggle here, although adding all sorts of different output modes may turn out to be a bit confusing.

I broke GTO+...

Here's a "prisoners dilemma" spot that it probably wasn't designed to solve. It's a chop board AKQJTr with very high rake and a high SPR. The optimal strategy is to rangejam and force villain to rangefold due to the rake. Instead, OOP range-check-folds, incorrectly calculating the EV of a jam.

https://forumserver.twoplustwo.com/1...money-1799307/

https://drive.google.com/file/d/1t2l...ew?usp=sharing

Pictures: https://imgur.com/a/uuzHEQg

Ok, I see the problem.
The line is discounted very early on.
In zero-sum games this can never be a problem, but apparently we need to set an exception for non-zero sum games.
I'll take care of it for future releases.
Thank you for pointing this out.

Thanks Scylla!

How does card removal switch work? Can you explain the math or provide a brief conceptualization of how it adjusts the frequency?

We were trying to work this out for a toy game in the theory forum and it doesn't seem obvious. For example, A8s, 88 vs AA/73o on AQ962r with an SPR of 1.

Are you using matchups? Or some other method?

If there's 100 combos in total, of which 35 combos bet, and 65 combos check, then, without card removal, this will be calculated as 35% vs 65%.

If card removal is ON, then GTO+ will also take into account the card removal effect by villain.
So if villain holds a lot of aces, then it will be slightly less likely for hero to hold an A as well.
As a result, the frequencies may be slightly affected, such as turning them into 34.6% vs 65.4%.

If ranges are wide, then card removal will usually only have a very small effect.
If ranges are tight, and there's an overlap between hero's and villain's holdings, then the effect may be significant.
This will often apply to rarely taken aggressive lines, such as bet, raise, 3bet.

Sure, but how exactly does it calculate this?

In the example I posted earlier, I thought it would be something like (probability of calling GIVEN IP has already bet), but that doesn't appear to give me correct frequencies.

It's not an entirely straightforward calculation.
Unfortunately the discussion of algorithms is beyond the scope of this thread.

Thanks for reply!
I found a tricky spot w/ "fantom" bet size. I solved AQ7FD flop for BvB spot w\ 3 sizes OTT (on the turn) and found that 33% used rarely only for FC (flush closer) cards w/ low freq.

So i removed 33% cb and was trying to simplify some turn decisions (leave just one size).
I picked 6h turn and exported it in new window


In new window i added 33% bet and now GTO+ uses it quite a lot comparing to initial strategy for 3 sizes OTT.


What is it? I lean towards to believe more to turn reports than recalculated turn tree. I can't formulate an exact question. It's a bit confusing, pls help

I would need to see savefiles, but a reasonable explanation would be that the subtree behind the newly added "Bet 4.95" line is not identical to the original one. If you want me to take a look, then please send the relevant savefiles to support.

Can we import HRC ranges to GTO+?

So i want to drill the same line always which is "X,b,r and call" Wich bring me to the turn OOP.
When using "drill current decision" in turn we can only have that exact turn always. Would be nice to be able to change turns keeping the same action/line (X,b,r and call in this case).

We don't support that particular format.

We can consider it for future releases, although in some cases it's not as straightforward to accomplish as in others.
For example, if "Basic" storage is used, then a recalc would be needed before starting every hand.

Is there a way to choose a bet size in BB instead of %

An input field can apply to multiple stack-to-pot ratios.
Because of this, it's not possible to set a fixed bet.

Just a suggestion in the future releases, can you add some feature where if we node lock we can automatically set the whole range into one action? For example like double clicking the Check Button while "Lock + edit decision" is enabled, the whole range will be set to check instead of doing it per hand strength.

In v139 it's possible to merge actions together.
For this, click on "Lock+Edit decision".
Then, click on the "Merge" symbol.
A menu will now come up that allows you to merge actions together.



If you want to merge all actions into a single action, then you can also double-click on the bar below the matrix (it contains all hands in the range).

Recalc wouldn't be a big deal. Please consider it strongly. The new round up option + keeping the same line, your trainer would be the absolute nuts!

Ahh never saw that bar. Thanks for the screenshot

Im using 6+ version, when will 139 be available?

Usually we first release the regular version, and GTO6+ a bit later.
I will release the GTO6+ version on Monday.

Is there a way in GTO+ or Flopzilla to see the % of each card rank in a given range?

Can you please provide some more details as to what you're referring to?

Basically I want to see the relative representation of each rank in a range. For example in a range of AKo, 22 we have 18 combos. We set 18 as the 100% benchmark

A & K appear in 12 combos so they each get 66%, 2 appears in 6 combos it gets 33%. If this data is presented in a visual form kinda like this:



It would make it very easy to see at a glance what types of flops interact well with it. For example in this case I could quickly see that the range is weak on K high boards. To take it even a step forward it would be nice to compare different ranges and see in what card ranks there exists asymmetries. If you can see the rank asymmetries at a glance you already get a feel for how any particular board should play even before running a solve.