You are focusing on all the wrong things, the exact hand selection of those ranges doesn't matter when we just want to gauge the rough magnitude of the effect. I just locked the EP ranges after a few iterations, they are not close to fully converged.
Let's break down a really trivial example, step by step:
Just a simple 10bb (no antes) push-or-fold scenario, so we don't have any unnecessary complications. Let's look at SB-vs-BB.
The no-bunching solution looks like this:
SB push 58.2%:
22+, A2s+, A2o+, K2s+, K2o+, Q2s+, Q7o+, J3s+, J8o+, T4s+, T8o+, 95s+, 97o+, 84s+, 87o, 74s+, 76o, 64s+, 53s+, 43s:0.51
BB calls 37.4%:
22+, A2s+, A2o+, K2s+, K5o+, Q7s+, Q6s:0.45, Q9o+, J8s+, JTo, T9s
The bunching solution looks like this:
SB push 52.5%:
22+, A2s+, A2o+, K2s+, K3o+, K2o:0.99, Q2s+, Q8o+, J4s+, J3s:0.30, J8o+, T6s+, T5s:0.80, T4s:0.04, T8o+, 96s+, 95s:0.30, 98o, 85s+, 75s+, 65s, 64s:0.12, 54s
BB calls 32.9%:
22+, A2s+, A2o+, K5s+, K4s:0.99, K3s:0.01, K7o+, Q9s+, Q8s:0.63, QTo+, J9s+, JTo
Original SB pushing EVs within the no-bunching game vs BB calling 37.4%:
Not lets plug the ranges from the non-bunching solution into the bunching sim and see how the EVs for these ranges look like with the skewed distribution.
SB pushing EVs against the same 37.4% BB calling range, this time with bunching from the early position ranges:
Those are some pretty sizable mistakes if you play the nobunching ranges on a table that opens anything like the bunching solution,...
Last edited by plexiq; 07-08-2021 at 02:33 PM.