Hey, I Use Dual core 2GHZ.
I use WIn 7, 64 bit version.

It looks like that crEV is 32 bit application, Does it run slower on 64bit machine?
Is there any way to boost performance, without buying new CPU or overlocking it?

For some conditions I need to run to too many trials, so it can get slow :|

No, that shouldn't be a factor.

The main factor is the CPU speed.
It's really the only thing that matters (with the exception of an external program like a virus scanner slowing the software down; in that case set an exception).

If you're still working on your project where you delete ~99% of the runs that are performed, this is to be expected. You should probably consider using 100k runs instead of 1M runs.

Hey scylla,

I just rebooted my computer and CREV somehow uninstalled by itself.
Thankfully all my saved conditions and filters are still there.

Any ideas what could have caused this?

Well, it can't have anything to do with CardRunnersEV itself.
It's just a program on your computer.
It can't uninstall itself.

Hey scylla, any chance in the future for a backdoor straightdraw using both hole cards condition?

Hey scylla, it would be cool, if you would be able to modify/add board setting function, maybe that would be Random Board with additional parameters, maybe Some new function.

Example:

SetBoard(street, paratmeters, SupressErrors)
street,
parameters: street cards,
a ) integer [1..52] entire deck
b) single : [1-13]. [1-4] [rank].[suit]

SupressErrors: for loops, that program wouldnt be stopped if it attemts to set (for example while graphing) wrong card. Outcome should be Nan, error or something, but loop wouldn't be interrupted.

That would allow all other runnouts.

New additional function:

Board, with some similar parameters, that would allow to output in graph or two dimentional matrix random generated board (or just part of it, like turn cards). There is no even need to render that in images, just [1..52] would work well.

Examples for first function:

SetBoard(0,48,49,50): set some Royal straight flush possible board, like TJQ of spades or something

SetBoard(1,52) turn A of spades.

Example of second function:
Add data point (Board(1), EQ(#1)), lists equity for action pojnt #1, and turn card


Thnx.

Yes, I'm looking into that.

Do you mean for scripts?

Scripts would have highest flexibility i guess, so yes.

Almost only thing that crev lack now is more flexible scenario analyzer when u plot exuities or freaquencies based on the board texture.

I would be very glad to loop over all turn cards and graph or make table of results and i would like to generate bunch of boards and try to find out critical board textures

Could you give me an idea of what sort of board textures you'd be looking for?

k, so first I will say that what you didn't ask for :

When I am talking about plotting different turns, then I want to have ability to do something like this:



That is equilab's Scenario analyzer tool. Unfortunately it shows equity for specific holecards only- not entire range. ANd of course it doesn't take into account any decision trees at all

So Having that in crEv would help a lot. Having that in action scripts would allow to plot you anything- interesting equities, plot difference between several lines (lets say check/call vs bet), plot any other data, accessible over custom formulas and checkpints. And best thing is that IMHO it is not that difficult to implement.

----------------------------

Now, what board textures I would be looking for.

Well main problems that I face are:

1) i want to work on some specific board textures (and that doesn't mean, that those should be similar, for example XYZsss and [2,9][2,9][A,Q]r, but I don't want to lose a lot of trials due conditions that will not satisfy me.

2) Empirical research. that is most interesting for me. I know what are breakpoints, numbers, however I don't know what conditions should be to justify those results.

Example: I want to find out all board textures that some freaquencies will be at [x,y] range.

What do I do now- I generate boards until condition is satisfied, then scripts stops. Unfortunately in that way I get just one board that satisfies that condition.

If I would be able to generate, lets say 1000 random boards then I would get much more data to perform my empirical research.

3) Situation is pretty similar to first one- when I have some board pattern, like monotone board texture. But I can't do that by just entering board, because variation of rank could have critical changes.

So 2nd ppoint is most important.in addition it would be combined with 3rd one in some cases.

Is that fancy stuff or really good feature? Well it is very good thing that I lack for entire period that I own crEV I even tried to automate those things with Auto hotkeys, unfortunately crEV provides everything in graphical way, so I can't read text and numbers in that way

Also maybe you will pick up some interesting ideas right here:

http://flopinspector.com/features/

and some introductional video:

http://flopinspector.com/tutorial/

how to use it on just one street ?

For example I want to find best 3bet range preflop but it takes me to the flop, turn and river too.

If you want to just play preflop and on the flop get your hand's equity, then just use a checkdown action on the flop. You can also assign a checkdown percentage so that the player doesn't get his entire equity, but just part of it.

So Scylla what do you think about that that all board output and also SetBoard, would it be possible? O ryou don't see how it could be valuable?

I'll just be honest here.
I think it's best.

Your post is full of spelling errors and incorrect sentences. Also, you constantly jump from one subject to the next and apparently expect me to fill in the blanks. And in the end you link to a pokersite "for inspiration".

Please understand that yóu are the one making a feature request. Personally I don't need this feature. Yóu are the one who has to convince mé. If you can't be bothered to clearly write down what you want without spelling/syntax errors then I'm not spending time on this.

You're asking me to spend weeks of my time to add a new feature, while you're not even spell-checking your posts.
So, no, I think I'll skip this feature.

Ok, thank you for your honest answer.
What likelihood is that you will implement this feature if I will spend more time to explain than in more clear way? Maybe if it is low you don't have time or will to work on that then I will just do other stuff and not bother you, however if chance is pretty high, then I would do my best to convince and explain that.

And sorry for bad english + not having spell checking software.
Sorry if my previous post looked offensive, but I got impression that you are not going to do that but you didn't tell that. so I value that honest answer much more than polite silence.

It doesn't work...
[IMG]http://s29.************/5goel0ph3/2014_02_12_17_44_04.jpg[/IMG]

As you can see, in the first green square BB calls the SB raise BUT you still gt to the flop and you are forced to continue on the nest streets.

Also, in the red square below, BB checked BUT the action continues on the flop and so one.

Why ? What is wrong ? What should I do ?

Thanks !

If you click on a decision button the "Advanced add action" menu will appear.
Here you can select the checkdown action:



Just use this menu to set up the following tree:



You are now checking down on the flop.

Do you think it would be possible to add weighting to the text range input? For example KcQc(50),KcJc(50),KcTc(33) would add a range to a condition with those weightings. Or possibly KT-KJ(25) would add that whole range @ 25%. I understand this might be not possible/too much work depending on how things are setup. Thanks.

I'd like to have the ability to put a postflop condition of "top x percent of made hands".

Like say you get to an unknown river oop, and you want to call with the top X percent of your range to be unexploitable to bluffs. It would seem like fairly simple programming to have this condition.

This is currently not possible and can certainly not be done in a single EV run (it requires two separate runs; the first for determining the range and the order of hands; only after that can the top X% be determined, after which a recalculation is needed). It might be possible to do this in the future though, but it is not as simple as you propose.

I cóuld, but could you please explain for what purpose you would need this?

I think I do a lot more splitting of combos than a lot of users (mainly OTR). Helps with blocker effects, can get more accurate solutions when trying to be unexploitable, allows you to compare taking different lines with the same hands etc. I also type in parts of ranges fairly often (mainly for the first two reasons, but sometimes it is just faster) so being able to weight them a bit faster would be nice, especially if some combos are weighted differently.

Could you verify this http://forumserver.twoplustwo.com/25...right-1415445/

Please ?

I want to see if I did it right or not.

Thanks !

Can't you just use the menu with the predefined conditions (left part of the preflop condition menu)?