This is essentially a toy game based on a hand from the Thinking Poker Podcast.  If you don’t already subscribe to this, stop reading my babble and subscribe.  You’ve got a lot of great listening to catch up on.  I’ll be here when you get back.

For the rest of you, here is the set up.  Hero has cold 4 bet from the SB, the original 3 bettor, who was in early position flat calls our raise.   SPR on the flop is 4.The flop is 963r (one club).  Hero has AKcc.  Hero ranges Villain as TT-QQ.  What should hero do?

I set up a super simple toy game on PIO:  Hero’s range is KK+,AK.  Villain has TT-QQ.  How should the action go?  I gave the hero multiple bet sizes at first, but to simplify things, I just gave him pot on flop and turn (in reality he can definitely go smaller to get all in by river).  PIO pots the flop with its entire range, Villain folds 100%.  Of course this suggests betting AKcc on the actual hand.  As Andrew pointed out, if our range is AK+KK+, and we want to have bluffs, having the BDFD is the hand to do it with.

I then changed the range to include 99.  Most players don’t have this in their range for a 3 bet vs EP, but I just wanted to see how that changed things.  Hero still bets pot with its entire range.  Now, of course Villain continues with 99, but does not raise.  So, they have to balance that with some calls; they call their other pairs a little less than 1/2 the time.  Since all 3 pairs have equal value, they only call with the hands that block a back door flush.

On the turn, which was a 6c giving hero nut flush draw hero jams.  Of course, Villain snaps 99 and calls about 1/2 the time with his pairs that block the flush.

One interesting note.  When I first ran this, I stopped the sim at about 1% accuracy.  This wasn’t nearly enough distinction to get PIO to strictly prefer calling the pairs with the flush blocker only.  It was still mixing some calls without a club, and some folds with a club.  This made no sense to me; I assumed it would ‘figure that out’ much sooner.  This definitely makes me question just how accurate most sims are; it just isn’t practical to get down to the very lowest levels of exploitation; I got it down to 0.03 or so, at Andrew’s suggestion to see that PIO was doing this correctly.

I do think there are some real world conclusions we can draw here.  Against a straight forward player whose range is medium overpairs, we can barrel them off their hands on dry flops, and be balanced with our AA/KK, at least at this SPR.