For this, simply create a tree where you only bet (you can use the editor to remove the "Check").
After that, build a database from that tree.

Excellent, thank you!

Oh, i just catch the "screen resolution can be below...." option, thats what i want
thanks anyways

Hello, I currently have an i3 it is 2 cores at 3.2ghz

I am definitely looking to upgrade bc it took ****ing forever to run a blind vs blind srp sim

Looking at i7 says it has 8 cores at 4.5ghz

If it took let's say 45 minutes to complete a 1% dEV with the i3 how much faster you think an i7 would make this? I mean it should be quite extreme difference I would imagine?

Your website says the formula of (core*speed) is how fast GTO+ runs the sim so my set up that is 2*3 = 6

with an i7 it would be 8cores at 4.5ghz so 32?

My pc has 4 threads and i7 has 16 threads. does this matter?

Also what about memory? I have 8gb memory and sometimes it says I don't have sufficient memory to run the simulation, I definitely dont' want that happening. Would an i7 fix this issue or would I need 16gb memory upgrade as well?

Thank you for your answer

Is there a way to node lock decision by frequency? For example let's say that the optimal frequency for a certain combo is 100% bet divided to 70% bet size A 30% bet size B, and let's say the tree has sizing C which I want to node lock this sizing for 70% betting frequency of size A, is it possible or theres only an option to node lock for 100% frequency?
Thanks!

I think you answered you own question on the RAM issue. You need more or you need to simplify your game tree variables.

I ran a really intensive solve off the balanced subsets included. I want to grow that database now. Should I run another with the same tree and then merge? I do not seem to be able to choose an option like Add X random flops to the database. Thanks for your instruction.

Okay so I've resolved this same exact 80 flop subset twice now, making sure the database and all the ranges are updated and I'm not getting an aggregate report. I have an earlier iteration of this database and several other databases with no issues. Any thoughts?

Maybe the file was corrupted somehow. I re-built the second solve within the same file. I am now trying from a blank slate. I will update if this doesn't work.

Two cores of 3.2gHz is basically a 6.4gHz system.
And 8 cores of 4.5gHz is 36gHz.
So it would roughly be 36/6.4=5.6x faster.
A 45 minute solve would then become 8 minutes.

You would need to upgrade memory as well.
16GB should be sufficient for any regular use.

No, because just the value 70% doesn't tell us which hands to lock.

Yes, I think you should indeed be able to merge.

Are the preflop ranges perhaps not symmetric? Aggregate reports can only be calculated if suits can be switched. For example, if not all AKs hands are present (AhKh AcKc AdKd AsKs) then it's not possible to build an aggregate report. So, for all suited hands all 4 combos need to be present, for all pocket pairs all 6 combos need to be present and for all offsuit hands all 12 combos need to be present.

So I didn't know about the suit selection would prevent an aggregate report from being generated, however I know that the OOP and IP ranges needs to be symmetrical, not to each other, but throughout the database. However I've always done weighted ranges with no issue up until this point. Starting from scratch and running it a third time yielded me the same issue.

Ok, if you need me to take a look, then please send a sample database to support (just 15 small trees, basic storage, etc).

That's very kind of you but actually I manually entered the ranges instead of just preloading them from the newdefs and I found a bunch of hands that mysteriously got suit selected in Flopzilla somehow. Issue resolved. Thank you.

Okay so I figured it out. I believe the issue came from me exporting a GTO+ range from a database file into Flopzilla, and importing into Flopzilla reveals a suit selection that presumably came from the GTO+ solution, which then carried back over into GTO+ and denying me an aggregate report.

trying to see diferent strategys as BB vs BTN on flop, and i was do next

run the solver, then blok IP strategy on flop, then edit oop strategy, then run solver.
On board AKQ, i was do it, just blok ip cbet strategy then edit oop just with 0% xr and with call/fold that i want, then run solver.
How it is possible that the EV of OOP in that edit solve is a bit bigger that the EV of OOP in original solve?
I can´t understand it
thanks

Yes, if either a flop or dead cards are entered in Flopzilla, it will remove them from the range.

Can you send a savefile to support?
It should make it easier to answer your question.

sure, but what you mean with savefile?
just copy paste last message and send to support, hope it is, if note says me please

Create a savefile of the tree with "File->Save as ...".
Then mail that file to support.

oh okey, so i do it with both, original and nodelocked one?
but how i can upload the file on the support box question?
EDIT: ok i see you on email thanks

Hello, need help please



CO opens, BB defends, srp

It looks to me like the solution is saying the EV of x'ing and betting TT on the T76 texture IP is exactly the same EV showing as 10.17. This seems very strange to me.

Obviously vs weaker regs and recs I'm never x'ing but let's take one of the top regs in a 500NL pool for instance. How can x'ing back TT here net us the same EV as cbetting? I'm kind of confused by this.

Thanks for any clarity

I have another question please

Imagine an EP vs BTN cc situation where BTN is fishy and has a fairly wide cc range

Flop is AJ9r and Hero has AA. The solution suggests that BTN should be raising our flop cbets with a huge range of 25% - things like 9Ts, some underpairs, a lot of Jx and KQ/KT combo's - kind of confusing why it would do that - but obviously nobody would be raising 25% like this

What I was able to do was edit the BTN's range to mostly call all those combo's (assuming passive fish) but the EV's didn't seem to change for the Hero EP strategy.

So the question I guess would be - after we edit the flop decision closer to what we think the villain would actually do - how can we re run the simulation to take into account for this change?

I tried "play against the solution" but I can't access that it's greyed out - any other ideas? Thanks.