Quote:
Hi. I don't know if this question has already been asked here. Are there any plans to make a version for Mac os? And if not why not?
MacOS version won't happen.
The reason is that it's a lot of expensive work to port software to another OS and the market for Mac version is rather small. We just don't have enough resources to do it.
Quote:
I then copied the ranges with which the IP player would call and the OOP player would check and I tried to solve on one of the 30 flops (ACTD2S) with the same postflop betting options.
The results I got were wildly different. Why?
It's hard to say without seeing a full tree config (feel free to send it to us, I mean the preflop one).
In general though it's likely that your postflop solve went to much better accuracy. The preflop solver uses limited precision (16bits instead of 32bits) to save memory and as the trees are big you don't get as many iterations on them. if the spot is such that betting and checking is very close in EV it may take a long time for it to converge in the preflop tree while in the postflop one it will be able to.
This doesn't for the preflop ranges though as EVs for bet/check are likely just very close so overall EV won't be influenced.
Quote:
On a further but related note, what ranges should I use when trying to determine my optimal plays on a flop for a blind level? Is it right to use the ranges that you get from solving the situation starting at the preflop (I read a couple pages back that once you solve a preflop tree for 60-70 flops you get a very close to optimal preflop strategy)
Yes, using those ranges is the way to go, especially when you solve on a sizeable subset. Of course there will still be irregularities when having 60-70 flops so you may want to smoothen the range by hand (keeping the same %) to remove random noise but in general yes - use those ranges.
Quote:
when i browse the solutions every time i go to turn i have a short pause of 1-2 secs (when pio solves the turn)
the question is how i know the accuracy of provided turn solution?
It's called "recalc accuracy" and it's defined in Tools->Configuration->Behavior
You can change it and see how it influencs the speed of re-solving. Rivers are in general very fast (usually below 100-200ms) but turns my take a while.
Quote:
How is this possible? Shouldnt mix strategy have the highest EV?
There are two most common reason for our users gettig it wrong:
1)Comparing EVs anywhere else but the root of the tree. This is useless because that means comparing EVs of different ranges.
2)Using rake to run simulations (this is useless because raked games are not zero-sum so multiple solutions with varying EVs may exist, comparing those doesn't result in anything useful)
Quote:
Also in future is it possible we can have a feature like card removal and how it impacts the solution. ( like excluding our 2 hole cards )
Solving when excluding cards is not very useful (we think) but it will be added to browsing the tree so you can see how frequencies change depending on what you hold.
Quote:
is it possible to convert small saves (flops+turns) into micro saves (flops only)?
Yes but you would need to write the script yourself.
Like this:
Code:
load_tree mysave1.cfr
dump_tree minisaves\mysave1-mini.cfr no_turns
load_tree mysave2.cfr
dump_tree minisaves\mysave2-mini.cfr no_turns
... (repeat for all trees)