Hi
I am new to PIO and ran this sim at 34:11
https://www.runitonce.com/poker-training/videos/intro-to-piosolver-part-ii-liveplay-with-examples/
Basically, we 3bet 7d7s vs BU open, BU call and we face a flop Qs2d6s
I tried to simulate the instructor's simulation on my machine (using a rough estimate of the PF ranges by looking at the yellow area in the range section in the video). Initially I was getting results that were way off for example, in the instructor's sim, solver suggests OOP betting 87.24% OTF but in my sim, its suggesting less than 50%, I made a few changes in the preflop range (which I think are minor) and I manage to bring the betting range for OOP on the flop to 78%
What I have also seen in this case is by only removing the 6 combos of QQ from IP player's range, it makes a huge difference and that makes me wonder how only 6 combos out of 255 total combos makes a huge different to the strategy advised by the solver.

I do appreciate, with these solvers the solutions are only as good as the assumptions you input and in the above case, QQ is an important combo on Q22 flop but it makes things very hard when you are trying to analyse a hand and assuming ranges.
For example, if I had this spot in real life vs an opponent that I either know or a complete stranger, a lot of the times it very hard to say whether he has QQ in his 3bet calling range or not, and if that is going to make such a huge difference to the output solution, how am I ever able to trust the solution.
Either I am missing something in running the sim or I need some advice on practical aspect of the assuming preflop ranges of opponents we don't know, do we just assume they are playing GTO ranges pf (when they are clearly not especially incase of microstakes where people are making random plays everywhere) ?

Thanks