I need help understanding PIO results of a SRP LP vs Button flat on A82 rbw board. I ran 3 scenarios using the GTO sim and 2 nodelock adjustments for IP response to cbet. I was very surprised to see how oop cbet % decreased drastically between the 2 nodelock scenarios detailed below. The sim was run using 33%/66% oop cbet size options in a 200bb game. LP opens 23% and button flats 14% (represents the live games I play in). Any help would be much appreciated:
Original GTO sim I rancame up with a 70% oop checking range. I then ran 2 nodelock models for IP call/raise/fold range:
Model 1: I adjusted calling/raising ranges mainly by calling more pocket pairs and Ax (not raising Ax/pairs as original sim shows the GTO strat)
Model 2: I adjusted model 1 further by raising 2pr at a higher frequency and slowplaying less (i.e. IP raises sets/2pr 50% (original model has sets raising 50% and 2pr 25%)
My results were very surprising:
GTO model - OOP x 70%, IP Call 66%/Raise 13%/Fold 21%
Node #1 (slowplay sets/2pr) - OOP x50%, IP call 71%/Raise 7%/Fold 21% (i.e. reduce raising range)
Node #2 (same as node1 while playing 2pr+ w/ less slowplaying IP) - oop x 4%, IP call 70%/Raise 9%/Fold 21% (i.e. strengthen raise range/weaken call). OOP also folds 90% of range to a raise vs this range including AK, AKs (80%)
Main Question - Why does oop strategy change so drastically by changing it's betting frequency from 50% to approximately 95% between Node 1 & 2. Does it make sense that increasing the raising frequency of IP's strongest value hands should increase our cbet strategy to this degree? It seems PIO thinks so, so what is the rationale behind this?
I've attached 2 screenshots. The first one shows OOP strat under both scenarios and the second shows IP strat under both scenarios.
Based on comments these are a little difficult to see but main point is both IP strats are exactly the same vs a bet outside of 2pr raise frequency (approx 25% in slowplayed model vs 50% in non-slowplay version). Sets are raised at approximately the same frequency.