When i am trying to copy a range from pio and load from clipboard into monkersolver it doesnt work. Can you fix it?

When defining the amount of buckets per street for NLHE HU trees, is there some kind of maximum number, beyond which there is no improvement? I mean if the number is absurdly big, then each bucket will contain not more than one hand and there won't be any point in increasing the bucket amount.

I was just thinking in terms of usability, that it would be helpful if there were a time counter, perhaps next to 'Status' 'Running' to see how long a solution has been running for.

Also under the tab 'Tree' where it says 'Save', I think it would be helpful to see previously saved trees, so that they could be overwritten or just to see naming schemes for previous trees etc.

Is it safe to run two instances of monkersolver from the very same folder? It seems that solver creates some flop/turn/river bucket files (named mmnested[streetname].[amount of buckets]), won't there be any problems with different trees using the same amount of buckets?

Was wondering if you could make an official MonkerSolver Skype group (like there is for Pio) for users to be able to discuss things?

Been needing a new computer for a while so gonna take this release as a reason to upgrade, how does this sound?

CPU: Intel 6800K
Motherboard: Asrock X99 Taichi
Ram: Samsung DDR4 16GB x 4
GPU: MSI GTX 1050 2GB
SSD: Crucial Mx300 275GB SSD (probably gonna change this to 500gb actually)
HDD: WD 1TB HDD
Case: Akbo Nvore G200 Thanos
Cooler: Coolermaster Hyper 212 LED


I haven't kept up with computer hardware in years but I have figured out I shouldn't get Ryzen due to the 64gb memory cap, and I want to be able to upgrade to more than 64gb if needed.

Is there a better choice of CPU or is this fine?

If you can, I would wait for x299 platform and use a virtual machine for the interim. If you want to build now, x99 would be the way to go.

Looking at your build, I would say it sort of depends on how much you value value over performance. Based on OP's comments, I would say the 6950x is currently optimal for Monker.

I'm not sure how SSD performance impacts performance in Monker, but there are much better options than the MX300. I would recommend the 960 Pro from Samsung (http://ssd.userbenchmark.com/Compare.../3642vsm182182). The performance would never be wasted since other programs, such as HM2/PT4, benefit from the improved read/write performance.

The remaining elements in your build are not relevant for poker, except for perhaps cooling, since you will want to run some sort of stable overclock on x99. Your cooler looks sufficient though.

that's basically the exact machine I built (same motherboard and all)

it works well

having trouble doing saves since the latest update; getting the not enough disk space error

Not at the moment. This would kind of require scripting functionality, so when/if that becomes a thing, command line interface might be added too.


Yes, with perfect information about the texture, otherwise it's not guaranteed. I don't think you will see much effect on the preflop ranges past 30 in hold'em though.

Yes, once created the files are read-only and can be used by multiple instances.

I don't have much benchmark information, but should be good. Obviously there are better CPU's, but also more expensive.

Fixed now. A workaround if you have a run in the old version you want to save is to manually create a file named for example "save.mkr", and then you will be able to save it as "save".

If there is some interest for this, then sure.

found a weird bug that occurs when you try to build trees with awkward stack sizes. i.e. if you build a 3h tree where the button has 20 chips, sb has 20 chips and the bb has 18 chips and then add a minraise option for the button some nodes will not generate a 'call' action vs the bb who has a smaller stack. This also occurs on postflop nodes. I attached a image below.

Hi, I have a few questions/problems after purchasing the full version:

Can you see estimate how large the tree will be before you run it? If so, how can you see it?

How to remove all-ins from all the raises? Is it through the settings in the Tree tab? What value for all-in threshold makes sense?

The scaling on a 4k screen is really bad. You can't see what bets are made (on the little table view), you can't read the chart with hand values (after clicking on the action in tree explorer), the range selection tab is extremely small.

And finally, copying from clipboard in Ranges does not work either. Is it a specific Pio format that needs to be copied? (eg. when you copy range from the solution matrix it is in a different format than if you copy it from range selection).

One more note, could not edit the previous post.

I think the software would really use some longer PDF tutorial than the few lines that are offered on the site. It's really taking a very long time just to figure out the basics of the software, how to build the tree so I don't have all-ins at every node etc.

Can we get aggregation reports?

Should be fixed, thanks for the report.



Yes, that would be the place. All ins which are more than a x% potsize bet will not be added. What makes sense is up to you, but it usually does not hurt much to have some unnecessary all ins as they don't bloat the tree size very much.
This is also described in the tooltip. Most (not all) non-obvious things have tooltips, so look out for those.

Should hopefully be improved. You can also set the scaling manually in the settings now in case your OS does not scale automatically.

The format which has been described to me as pio and which works is: AhAd: 1.0, KhKd: 0.5, etc

PPT format also works: AhAd@100, KhKd@50, etc

Agreed. Thanks for the feedback.

Could you describe this in more detail?

That's the solution matrix format. Thanks for clearing that up.

Scaling works great now.

Now, the tree estimation works for preflop but not if you start the tree on the flop. Is it by design?

I feel like after loading a tree I made earlier, the pot size was incorrect. I'm reasonably sure I had the correct potsize before. The potsize (game tree starting from the flop) after loading from the save file is 1sb.
Also, is there a way to edit the pot size from a previously saved tree? How about the ranges? If no, could that feature be added please.

No, it just hasn't been prioritized as there are not many people using the solver for postflop hold'em. It is added now.

There was a bug when saving trees with large potsizes (over 35 sb). This has been corrected now.

In the load dialog, where you pick the tree, you can check "edit potsize".

The ranges can be changed in the same way as always. If you mean change the ranges in the tree save, then save again and overwrite the old file.

Would you recommend Intel over AMD for monker? Or it doesn't matter for the most part?

is it possible to change the excel format and when export the hands the high card to be first? for example instead of 9dKh it should be Kh9d

Hi

I think i will talk for many people but Can you change the range viewer part and make it more like PioSolver, and other solver pls?

Like a big dial with the range of fold call bet etc like pioviewer ?

Thanks sure you understand

ok so are preflop solutions possible in plo?
So I can get a gto limping and raising(and folding) range in the SB vs the BB?
Is that possible?
How much ram and how many cores do I need approximately for this?

It should be possible, yes.

You can do it with 1 core+ theoretically. As far as RAM is concerned, I'm guessing somewhere in the 60-70 GB range. I don't think the author can really tell you, you would have to define the game tree and then use the built-in RAM estimation tool.

I'm not sure I properly understand this. I'm assuming these are percentages of a players stack? Does "Call committal threshold" mean when a player bets 60% of his stack, he is commited to calling a raise for example? Would "Fold commital threshold" then mean when another player bets 75% of an opponents stack, they are commited to either folding or going all-in? I'm unsure how this comes into play exactly.

When I try to click 'Add to all' -> 'Postflop nodes', nothing happens in this instance. Can you maybe tell me how this is supposed to work?