Is there any way of adding the Eq. solver command in a script?

If not please can that be added in a future a update

Hi,

It should not be too long now until our next release. A new, currently not yet existing, type of approach to "scripting" will be added in a release not too long after that. We will provide more information on this when the release is available. Our new concept has taken about 6 months to program, however, I think most people will be very pleased with it.

Cheers,

Scylla

Hey scylla howd you get to this area/function on crev, is it in postflop editor?

---

Second question, how do i see how many value combos and bluff combos villain has in a certain spot whenever i hover over i always have a hard time because it just lists 5% 3oak 15% tp's etc... is there an easier way to find value + bluff combos of villain in a certain spot or is there a certain function im not clicking on?

Thanks.

via Imgflip Meme Generator

It should all be done in about a week, after which we'll probably need another week for preparations (release video and such). We will provide more info next week when the development is done (I'd hate to post screenshots and such, only to find out that more development time is needed).

It's the previous analysis tool. The new one has quite a few more features, however, not the particular feature that was needed for the question. To bring up the tool, mouse over the EV field for the entire decision or an action and press Alt+A (after an EV run).


The problem there is that it's basically impossible to distinguish between value bets, semi-bluffs and bluffs. It's typically a spectrum with very strong and weak hands on both ends, but in between those areas there's hands that are very difficult to assign to any category.

For example, see the equity graphs below. Here, SB is the blue line. The graphs refer to SB's betting range. The right pic is on the river. On the river, betting ranges are typically polarized between value bets and blufs. So SB will bet his strongest and weakest hands, but check the intermediate hands (he wants to see a showdown with hands that have showdown value, and tries to avoid a showdown with his trash hands by bluffing them). And, given that it's on the river, draws are no longer relevant; a hand has either gotten there, or it hasn't. However, on the flop (left pic) the story is different. Pretty much any hand has equity, even the weaker ones (in this pic, the minimum seems to be about 25%). After all, in the end, even if you bluff with 22, then you still have a chance of hitting a set on the turn. So, on the flop, pretty much any "bluff" is still partially for value. It is for that reason that for flop and turn play, it's not really possible to categorize anything as a value bet or bluff.

Okay mate thanks for reply , so if we are OTR and all the run outs are known is there a way still to click on a button that displays hero's value bets & bluffs and villain's value bets & bluffs instead of just counting them manually on the computer screen by hand.

Or like even on a certain street i don't understand why we can't just see how many value combos + bluff combos villain has, like i'm not trying to do a nash equilibrium tree or anything just like viewing them and stuff sort of like matt janda style where he constructs ranges manually/creates defending ranges etc.

My apologies. For some reason I had not realized that your question referred to an earlier post in the previous page in this thread. My confusion was due to what exactly the definition would be of "value bet" and "bluff". Given the fact that many hands are semi-bluffs, people might have different opinions on when exactly a bet would be a bluff.

That being said, if I understand you correctly, then you would like to use the definition that a hand is a bluff if it's -EV versus villain's calling range. If so, then I think you should be able to use the approach with the "Delete" action that I suggested earlier. After that, sort the table by EV and check at which hand the range switches from +EV to -EV (hand 71 in the pic below). Also check how many hands there are in total (322 in the pic below). As long as no weights are applied in your tree, this will tell you the exact "value" vs "bluff" ratio, given the above definition of these two terms.

Hey guys, where can I set the different betsizes when making the tree? EG. flop bet can be 38% or 75% or 100%? Cheers!

Click and the nodes, go into advanced option and add your alternative bet sizes. It takes a while to set up the tree but it gets easier with practice.

We are excited about the coming updates, scylla, keep up the good work!

Our efforts up to v333 have been focussed on the speed of our algorithm and memory compression. Other than that, we have also taken time to develop our analysis system (for more on this system, please see the second video here: http://www.cardrunnersev.com/download.html).

In order to facilitate our development we have decided to keep our tree building wizard basic for the moment. Right now, if users want to consider multiple bet sizes, then they can enter them manually. Just to be clear, our solver is perfectly capable of handling trees of any size and with any number of bets, so there are no restrictions there, however, we will offer more elaborate options for the tree building wizard at a later time.

In our upcoming release we will not only be expanding on what we have built so far, but also introduce new features. I expect that the new release will be well received, so please be sure to check back in about two weeks.

Cheers,

Scylla

I think I should add that, right now, if you want to use multiple bet sizes, then just add the relevant actions. After that, right-click the decisions following those actions to add a subtree with the tree building wizard. When selecting bet sizes with the wizard, just pick the option that creates bets closest to roughly 2/3rd pot (or whatever you would consider a reasonable bet size).

It would be very useful function to have , at least I spend 60+ minutes on creating desired trees with multiply betsizes. The worst thing is that the process of creating big tree manually has 0 value in it and feels like wasting a lot of time and very boring... In my opinion- Function to expand tree wizard to multiply sizes on each street would increase value of this software by tons

Hy Scylla! Big fan of your product..! Right now I am trying to automate (not CREV scripting) an equilibrium tool process using automated mouse remote control through python. For some reason the left mouse click inside the CREV window doesnt work. Do you know why ? Greetings

My guess would be that the focus is not on the window at that time. Windows can sometimes be a bit tricky when it comes to focus. That being said, although it will not be available yet in the upcoming release, we will offer a new type of scripting not too long after that. So a dedicated method of achieving what you're looking for should be available in the near future.

Cheers,

Scylla

Thanks for the reply!
I am sure its not the focus since i alt+tab into the CREV window during running the python script (also with remote).
Looking forward to the new script.

How much ram and how good does the processor need to be to run crev smoothly?

In that case I don't know. If the screen is actually clicked on, then there is no reason why it shouldn't register. Is there perhaps any chance that the window scrollers are used, causing the location of the click to be shifted?

RAM itself does not affect speed in any way; it only affects the largest size of tree that can be fit into memory. At the moment, as long as you have roughly 4GB of free RAM available you should be ok in basically any scenario (in pretty much any case you won't need nearly as much though).

As for the processor, the main factors here are the processing speed and the number of physical cores. So if you have a 3.2GHz processor with 4 cores, then basically you are running on a 4x3.2=12.8GHz system. The solver scales really well with both speed and the number of cores, so any increase in either should linearly affect the processing speed.

Other than that, should you be looking for a processor, then please keep in mind that often processors are advertised with their number of virtual cores, which is a concept that is, at the very least in this context, meaningless to the speed. The value to look at is the number of physical cores.

I have 3.46GB of RAM available and my processor is an AMD A4-4300M APU (2.50 GHz). Is this good enough to run crev?

It's definitely enough RAM, provided you are running on a 64bit system (most systems are 64bit nowadays, so I won't expand on 32bit systems unless you need that particular info). As for speed, on my 3.4GHz with 4 cores it takes 88 seconds to solve the file /equilibrium_sample_files/flop/flop_4bet.stx to 0.5% of the pot (for the number of threads, choose 2x the number of cores). So, if your system has 4 cores as well, then I would expect it to take roughly 120 seconds. As long as your solving time is roughly in that ballpark, then it should be ok. Should you only have 2 cores, then I would expect the solving time to be about 240 seconds (maybe a bit more, given that other processes will take up relatively more processing power).

To test for yourself:
1) Start CardRunnersEV
2) Go to "File->Open..." and select the file /equilibrium_sample_files/flop/flop_4bet.stx
3) Click on the solver icon in the toolbar (or press F1)
4) Enter "0.5%" in the dialog
5) Click on "Compute"

Should you want to alter the number of threads that the solver runs on, go to "Settings->Set number of threads for solver and math engine" in the menu. In my experience, the performance is best if you choose 2x the number of physical cores.



After solving, to see the results, close the solver dialog and click on F7 to compute (or click on the "Calculate" button in the lower left). For more on using the solver and the analysis system, please see the videos here: http://www.cardrunnersev.com/download.html

I have a bit of a problem. When I try to save the regkey into the main directory I get a message saying that I need to request permission from the administrator to do so. So I cant run crev with full functionality(only trial mode).
EDIT: problem solved

Ye, i understand equity doesn't matter OTR only nut hand combos really so i can see why the right graph would make sense OTR.

Not sure if ppl post strat in here, but this is sort of a query/question lol.

Not really that great with using/understanding equity distribution graphs (all this is new to me),
...so basically if the blue line is at the top left corner and all the way up for the majority of the graph that just means we have a lot of nut combos compared to when it dips majority of the time we have a lot of bluff combos.

And my question is why is this important again what are we trying to figure out between the 2 variables of equity % and range?,
...i'm guessing whoever has higher equity % means they will do most of the betting/raising so whoever has the higher line equity % wise within the graph comparison for betting

Yes, from what I can tell, on the river, the betting range is typically a combination of nut+bluff with not much in between. Intermediate hands are checked in an attempt to reach showdown. This is done in a mix with nut hands (as to discourage villain from betting by threatening a check-raise).

River play seems to be mostly about trying to get value from your air hands by bluffing them together in a polarized range with strong hands, while trying to get to showdown with your intermediate hands. So your intermediate hands are trying to see a showdown, while your bluff hands are trying to prevent this. A player with a stronger range will indeed have a far more comfortable position here to get value out of both of his intermediate hands and air.

Equity graphs can be hard to interpret, however, the most important thing to take away from it here is that hero's betting range is polarized. The checking range on the other hand is typically a spectrum, containing nuts, intermediate hands and the remaining non-showdown air that could not be bluffed.

when working with preflop only trees (only reraising/folding allowed) sometimes a strategy will sometimes lose ev if the other player deviates from nash. Does this happen beacause of the rake?