So, I recently ran some sims and I have to admit I'm not totally clear on the drawbacks/viability of the settings/etc I used. I'm hoping folks in this thread can help me learn more about what I'm dealing with.

I wanted to run some 40bb 4max PLO preflop sims (as kind of an abstraction of 6max, which I'm experimenting with a 40bb buy-in for), first to see what preflop ranges were deemed optimal in a high-rake environment and then to see what ranges were optimal in response to different highly non-optimal "strategies" that are commonly played in these games.

Before running anything, I was playing with the settings and saw that doing perfect/30 flop abstraction with none/15 turn & river supposedly allowed me to run these situations on my home computer (which is relatively powerful but not amazing).

First thing's first: am I just sacrificing too much with my abstraction? Or will my results be somehow worthwhile, still?

Second... in order to create the exploitative sims, I'll give an example of what I did and see if folks think I'm on the right track:

In order to model a whale in the CO (or earlier; they aren't really positionally aware... reasonable abstraction imo), I went into my database for RFI and open limp frequencies for that player profile (as determined by DriveHUD). I then started a new run of the game tree I describe above, paused it ASAP, and then did the following:

Right click CO's CALL column... Force CALL frequency... Force frequency... type in their open limp CO%. Ditto the POT column and their CO RFI%.

Then I ran everything until it got to 10+ I/N. Rinse, repeat for different player profiles in different positions.

I mean, the short version of what I'm asking is: did I do everything right here? Haha.

Many thanks in advance.

I am having trouble loading a saved game, and then running it. When you load the saved game (all streets HU FLHE) it starts running right away, no? It doesn't seem to be doing anything and is frozen. And, if I hit stop then it cancels the whole sim and goes back to the state it was before any iterations. BUT, the status line reads

'Status -> Running, 00:04' and the run-time changes, but the solution does not really change. I find it hard to believe it finds the HU solution almost immediately, and I'd expect it to change.

The iterations line stays at '1,283,559 0.00/s 0 thread(s)'

and the rest of the lines are stagnant. Is it running or not?

I want to check some of my runs with you guys because I'm really confused about some results.

I've done several runs of defending vs locked 3b OOP whose size is 9bb vs our 2.5bb open and my monker is mostly folding and 4betting, almost none calling.. even vs range which is like ~14% wide (linear). Rake is set up to 750 mchip and postflop sizings are fine + there are several nodes, so it's not like oop is lacking some options of raise or whatever. If I lower IP's 3b size to ~8bb, then monker starts to call more.
Similair thing is defend vs locked cold 4b ~7% linear.. OOP will defend a lot more for 11->25bb then vs 26bb.

Seems like this 1bb is making huge difference.

Did anyone notice this?

^ fixed it... hit the '+' button to add threads. the process is running on 0 threads when you load it!

One open spot for getting hands solved. Pm for more info.

I had the same problem supert seemed to have, sometimes the monkey will briefly pop up but often it does not. I have 64GB of RAM and have changed the xmx variable to 58, still nothing happens, any ideas?

is there a way to take a solution and then put it on anther machine amd change the settings to continue to solve it?
if you purchase a tree can you continue to solve it with different hardware?

Hello guys: I try to make the payment of 499€ to buy the solver., but appear one message saying:
Payment could not be processed
What happens? Someone to whom the same thing happened?
Thanks!

I tryied to pay with 3 different cards and with everyone the same message.

How do I only let the solver raise to full pot? Like if I add "33%,50%,75%" it can bet and lead for that size but also raise to those sizes. I want to be able to bet such sizes but raise to full pot only.

Another question regarding the syntax: how do I filter for KKss only? So pocket Kings with two spades for example that are King high.

Hey, I was wondering how does the force frequency feature work when trying to solve PLO turn strategies? I tried locking it at 45% and after 30-40 iters it was at 70% then 40 iters later it went back down to 15%, keeps going back and forth

Hey guys, having a weird issue with a 40bb 8h tree. The moment I change three stack sizes to short stacks the RAM requirement for the tree immediately jumps from 120gb to 3000gb. I can only handle 256. When I click around to determine what got auto added I don't recognize anything out of place. Any ideas? Thanks

I have changed the _Java_options part to -Xmx250 for my 256GB drive, I've uninstalled all other javas prior and installed the 64bit offline and I still get the java error w/ no workarounds, anyone still have this issue with these alterations?

I recently downloaded the trial version and then updated my Java to windows offline 64 bit but now whenever I open the solver nothing happens and nothing opens.

i have 32gb and a 2core cpu 3.8hz. what's the max i should set the thread count to in a solve?

2 newb questions

1. does anyone have a guide on how to use macros? like what the syntax is?

2. how do you change out the hand rankings monker uses when you enter ranges by percentile?

im assuming support no longer responds here. They don't respond to my emails either.

Does anyone else experience monker to be freezing whenever you copy paste a range and then try to ctrl+a mark it to remove it?

How do I force close Monker in win10? It seems like it doesn't appear in the task manager at all

Maybe not allowed to post this here but so many folks wanting help and getting none.

If anyone wants monker tech support I'd do it for $150 an hour. pm me

In the task manager, shut down Java. It is easy to find since it'll be eating your resources.

I would recommend loading fewer ranges upon launching Monkerviewer. Monkerviewer is known to be slow starting up when you have a lot of files in your ranges folder, even on relatively fast hardware. I wouldn't recommend upgrading your hardware solely to try to resolve the issue. It likely won't hurt, but I wouldn't expect it to completely resolve your issues.

I wouldn't recommend setting texture settings to none for the turn and river for any solution that involves betting past the flop.

If you change any unlocked settings, you invalidate your run.

In order to continue solving a run, it needs to be saved with the option to continue calculation later. It doesn't matter that the hardware is different.

Use a semicolon after the first sequence of bets, so "33%,50%,75%;100%".

"ksk:ss:wwxz".

Are you solving starting from the turn?

I think this could be considered a "bug" in the current official release. What you are seeing is real though, the RAM requirements are higher when you have asymmetrical stack sizes.

A future update should address this issue. I would recommend trying to email the developer at support@monkerware.com with your issue. He can possibly help you out.

Are you using the free version of Monker? It should work.

Depends on what the exact CPU you are using is. You should be able to see this in 'System Information' under Windows. Either way, a dual core CPU is not suitable for solving with Monker. You may want to look at an entry level build like this one. It will be much more suitable to the task.

For the most part you can filter hands using PPT generic syntax.

Look for 'Java(TM) Platform SE binary'. Monkersolver will be running under that.

Does anyone have any recommendations for servers with >256 gb of ram? I've been using contabo but would like to run some heavier preflop sims and 256 gb wont cut it. What are my options? Thanks!

Is there a way how to drastically reduce the size of .mkr files when saving? I bought a course recently that featured a couple solutions with .mkr files that were around 5MB each but still functioned properly and could be opened with the solver, as well as edited. My own .mkr files are usually at least several GB.

How much would you need and for what kind of tree? I have access to a computer with 1024GB RAM, you can PM.

Unless you require triple perfect quality with hundreds of buckets most trees can be simplified into smaller trees with lesser RAM requirements.

hey guys. how i can increase memory in mac os x? cause i don't have MonkerSolver.l4j file there
i tried to redact start2.sh (change -Xmx6g to -Xmx12g) but it doesn't help me
can you help with it? thanks