So i've been studying BBvBTN 3bet pot at 100bb.
Ive come across some interesting turn sizing from BTN here in the XC XC line on AKXX.
with 53 chips in the pot, and 73 chips behind, BTN prefers to bet 36 chips OTT??

This kinda blows my mind because Im used to seeing a polarized range take a geometric size, and this size is 50p to set up a 30p river jam... clearly this is larger than geo.
Anyone have any idea why larger turn sizing here is better than geo?
or have any metric for when to apply this?

funny enough it actually seems like a good exploit vs players who wont put the 36 chips in unless theyre committing to the hand. But from a theory perspective i dont see the logic.

Im guessing it has something to do with BB having so many traps in range: AA/KK/AK. The traps are the only thing differentiating this from a perfectly polarized range vs bluff catcher toy game where I know geometric size would be preferred.

What does this oversized turn bet accomplish? Does it somehow decrease the EV of BB's traps?

Solver uses 75% 38.5 of the time on this turn


Across all turns never use the 75% turn sizing, using 33% a lot more often than the overall range


I was going to remove AA/KK from OOP in Pio, but I ran a normal sim before that which resulted in only B50 used


Ran it to 0.002% so maybe the Wizard solution is just solver noise?

Geo is only optimal in a game without changing equities.
The actual polarised sizing ingame would tend to be bigger than geo due to draws in both ranges

Yup - mathematically speaking, the aim of the geometric size is to force villain to put in the max amount of chips across the remaining streets. Which makes sense when you have a hand that can't be outdrawn, but if your value region is somewhat vulnerable then you actually want to bet a bit bigger on the turn, while your hand is still good.

it doesnt really make sense to me, because in most turn spots where the bettor is getting the chips in across the remaining 2 streets, the bettor has the polarized range and the caller has the catchers. And of course, the calling range usually has >0% chance to suck out with some of its catchers. So then wouldnt you expect the bettor to always bet a size >geo according to what you are saying?

there are other reasons to bet other sizigns, like you not having a very polar range in the first place, but if your range is very very polar then yeah >geo is usually what you see

i ran some pio solver toy games and I see that you are right