The two 5-bet sizes are just how the solver splits up its 5-bet range to extract max EV with certain hands (mostly KK, AKs/AQs, some QQ & A5s). Notably, AA rarely jams or small 5-bets, as the solver uses it to protect the rest of its flatting range, which has to continue pretty wide versus the small 4-bet. Good rule of thumb is, the smaller the 4-bet you're facing, the less frequently you should 5-bet AA, and if you do decide to 5-bet, you should go for the small size. With the exception of the exceedingly rare 4-bet shove (which I excluded from the final sim), 5-bets are where the solver really likes having two quite different sizes.
After facing a small 5-bet, the SB just 6-bet shoves the hands it would have called a 5-bet shove with anyway, but it's forced to fold the majority of its 4-bet bluff combos. The only 4-bet combos that can continue versus the small 5-bet are the suited Aces, suited Kings, and a little bit of AQo.
Just because you don't see it from your opponents doesn't mean you shouldn't implement it yourself. Chances are your opponents won't know how to play versus it and will make mistakes in terms of which hands they 6-bet, call, or fold.
This is taken from a rakeless sim, which was then re-run with only weighted average raise sizes (mainly for simplification purposes in my own studies), and the outputs rounded to individual combos for further simplification (so to the the nearest 25% for suited hands, nearest 16.7% for pairs, nearest 8.3% for offsuit hands).
82.4% 2.15BB RFI:
25.8% 10.25BB 3-bet:
9.6% 22.9BB 4-bet:
14.9% jam, 8.5% 39.5BB 5-bet (green=jam, blue=small 5-bet):
and the SB response to the small 5-bet (response to jam should be obvious):
Last edited by donkshovetheturn; 04-16-2021 at 04:56 PM.