I'm sorry, but I was 0% serious.
Unless you're a bored trillionaire (do these guys still exist?) and are willing to make me a billionaire.
My guess is .... not.

well that makes me very happy

I still think you should charge for it, but thats just cause I think what youve doen is awesome work, and you deserve something for it imo

The maximal exploitive button is pretty cool, I spent about an hour with it last night. Im not 100% sure on exactly how to use it yet, but will give it some more practice and watch the video manual.

Good job yet again

lol. Sorry to cause a minor uproar. I'm not a billionaire so the program is safe!

I'm also amazed with the most recent upgrades. My mind is repeatedly being blown to say the least. This thing does what used to take me hours to try to do by hand and gets to a waaaay better answer in just a minute or two.

tyvm Scylla.

I have been fooling around with the maximal exploitative tool

One problem I have been running into is using it for figuring out how to play perfectly vs a given strategy preflop.

For example, When I 3bet preflop each hand in my range will have different amounts of equity it will claim postflop. For instance, when I 3bet AA and get called with 100x stacks, it is likely my checkdown % for AA should be well over 100% of its equity of the current pot. But when I 3bet A5o, and get called, this hand plays much worse postflop, and although its hot and cold equity is decent, it will not be able to claim a very high % of its equity, maybe 75% or so.

So How do I tell the program (if possible) to calculate the maximal exploitative strategy, while also telling it how different hands perform postflop

this problem can be applie to flatting preflop raises in the BB as well, or raising hands on the button, calling 3bets, etc

any help here would be great

How about creating multiple lines for the 3bet preflop, one for each approximate hand strength where you feel a certain checkdown % applies?

And then just check down with that % for each hand strength.

If u mean what I think you mean that doesnt work.

Correct me if Im wrong but what I think you mean is to do this

Set our opponents strategy to raise X% from the button. Then insert a 3bet range that has y checkdown %. The insert the next 3bet range that has Z check down % and so on, until we have defined the checkdown %'s in 3bet pots for all the different subsets of hands we can 3bet.

Then we would use the mazimal exploitive button and the program would go through it



That was one of the first things I tried, but what happens is that no matter what ranges you put in for each checkdown %, the program starts each one at 'all hands' when u press the maximal exploitative button. This effectively makes putting in a bunch of other checkdown %s not worth it.


Oh but I just had a thought! Maybe if I order the different checkdown %s from lowest to highest instead of highest to lowest, then that would effectively filter out the hands that I cant 3bet.

Im gonna try that, and Ill post results when Im satisfied one way or the other

mm i dont think that would work either actually.

One other comment/question is, Ive been noticing sometimes when I use the maximal exploitative button, the analysis sometimes tells me to do things that are very slightly -EV. This has only occured for preflop stuff, when its just going thru the monte carlo mode, not math mode, so that might be it, but it still seems odd

I get the feeling that this feature is really valuable/cool for analyzing single street postflop stuff, but maybe not all that effective for preflop stuff yet

No, works just fine.
However, all Monte Carlo results always have a small error in them.
So hands that are slightly +EV in the exploititive analysis may turn -EV in a later EV run.
Not much that can be done about that, whenever you work with MC mode you'll always have to deal with interpretting the somewhat scattered results.

Also, you may notice the same thing when working with the math engine postflop. If this occurs, turn on "math engine->forced enumeration". With wide ranges, when this is turned off, an approximation is used which may cause +EV hands to look -EV in a normal EV run.

Ok, what you can also do is create multiple checkdown actions with the desired percentages, where in the postflop menu under "holecards->range" you can enter a string with what hands should be contained in it.

Easiest way of getting such a string that I can think of would be to enter such a range in the preflop menu, press ok, open it again and copy paste it from its name field.

that was a great idea scylla, thank you!

Hi scylla great software
Im havin some really hard time figuring out several things

1. lets say I am SB with 20BB, BB have 100BBs.
According to the Unexploitable shove I need to shove with 40%(!!!)
This is not making any sense...could you explain it please , thank

2. after examining this, I went to check my statistics and I have found out that I lose 27bb\100 in this situation (all folded to me, I am SB)
Which means that I lose in average 0.27BB every time.

How can I see the bb\100 when using Unexploitable push in this situation?

Many thanks, hope I made myself clear enough.

sorry double post

Hi Hemelet,

Yes, about the same thing over here.
38% here, but maybe I have slightly different settings.
Can you show me any pushing range that you feel performs significantly better?
Because I can't find one.


Edit: Ok, just manually checked by making a 2D graph. I varied the sb's pushing range and the bb's calling range. In excel, for every pushing range filter out the optimal calling range. Then out of all the pushing ranges, pick the best one. Best pushing range found was top 30%, best calling range top 15%.

Performance: $0.26.
Unexploitable shoving result: $0.28, which is a better result.

I see no reason to doubt the performance of the unexploitable shoving tool.


Multiply the EV by 100 I suppose ...

Ok, new feature and need some help.

When a condition has been created for either
- Unexploitable shoving
- Maximum exploitable counter-strategy
a short description is now given when you hover over such a condition.

Let me know if I've missed some major category.

Thanks for the reply.

1. So you can confirm that all the stats are legit? that the 39.1 is accurate?
btw do you have and can show "proof of concept" ? ak screen shot of PT,HM etc showing a result of unxploitable shove, after X amount of hands? that would be great.

2. what I meant to ask in the last question is, by using this wonderful tool , I expect to lose -0.21

And as said in the previous post, I currently lose 0.27BB\100 which is incredibly close to the EV using the unexploitable shoving! (-0.21 right?) {with 20BB on SB )

Does it mean that my SB play is very well?

Many thanks in advance, really appreciate your assistance

when stox ev auto downloads the new update, my virus scanner detects a threat

it says "a variant of Win32/Packed.Themida potentially unwanted application"

All I can tell you is, there is no threat.
Problem is, with all the virus scanners out there, I simply can not prevent one of them giving a false positive.

It's possible that there is a slightly better solution, but somewhere around the hands that are given, yes.

Sorry, don't have the time.
Also don't have a large database with 20bb sb's pusing into a bb.

Unexploitable shoving focusses on push or fold situations. Sometimes a smaller raise will get better results. If you only play push or fold then you're close to the optimum I suppose.

A bit too complicated for a chick like me

That's about right. See the relevant chapters of MOP and the FT book.

Actually, on second thougt, bb probably doesn't call with the optimal range.
So you can probably improve significantly if you choose the correct pushing range.

I love the software scylla. I wanted to offer up a feature suggestion, how about allowing for copy pasting of entire branches?

StoxEV is really powerful but I've found myself wanting more than once wanting to make some simplified assumptions that an opponent would always take a specific action.

For example:

CO (50)
Button (50)
SB (50)
BB (50)

CO raises to 5 with top 30% of hands

Button calls top 50% of hands, folds the rest

SB, always: shoves top 5% of hands, calls top 5-20% of hands, and folds the rest.

BB, always: shoves top 10% of hands, raises to 15 with top 10%-15%, and folds the rest.

Right now, I need to copy and paste the actions of the SB and BB one by one many times. It would be great if I could just select all of an entire branch of actions and paste them to another part of the tree. It would make some problems much, much less tedious to input.

Somebody asked about this last year, scylla's comments from back then can be found here.

As stated in the link above, I don't wanna.

Hello scylla, I have been using StoexEV for a while and I am very impressed with it.

However, today it seams I discovered that I do not understand what it really calculates .

I did the simple routine below. I raise PF from the BT with 99, BB 3Bets me and I call. Them I suppose that the BB will CB on the flop 100% of the time and that I will shove 100% of the time. Then StoexEV tells me that shoving on the flop is a +EV play, however calling pre-flop is a –EV one!? How calling PF can be –EV, if I know that once I do it I will be shoving on the flop 100% of the time and that is a +EV play?

If anybody else knows the answer, please.

Thanks