Trying to decide optimal line vs villain for max value in this spot is indeed interesting to me and I'm still not 100% whats "best". In-game I think I'd have opted for the ~$80 turn, call all shoves from villain and plan to shove river our self on most rivers. I like the advice/logic on just shove too. I hate giving too cheap a price to the FD but also hate losing my value-targets to a big bet. The solver exercise that demonstrates that the EV difference probably isn't significant anyway seems sensible, and how ChaosInEquilibirum broke it down by showing how solver changes as you add more value target combos vs flush draw combos makes perfect sense.
Knowing the clientele at the Orleans and the likely hands this villain can have, they are probably folding all the PP's under the K to any size, probably coming along with any of their top pairs as long as the bet isn't TOO big. Maybe they'll even call one more street with JJ-99? But I feel to any real pressure you lose near-100% of your value targets with this demographic. Plus with this demographic, in a 3-bet pot they may already have hapily folded everything but top pair or a flush draw to our flop bet... so they show up with even less possible value-targets on the turn.
What they do with the flush draw on the turn is likely call what they feel is "reasonable" and fold to a big bet, but you do get just occasional turn-shoves here over your bet when they don't wanna fold their FD or their improved FD to also have a pair or a SD also. Shoving on a 50+ year old Orleans shorter-stack player just seems like I'm losing all the value I should be getting with them when they show up with KQ or KJ here and they fold to the pressure. Maybe size to give just barely incorrect odds to call for the FD (does this make sense even though we are essentially committed to this hand so they will realize all their implied odds when a club comes in and they have it?)
Thinking about it now the advice I'd give myself is in-game maybe just use whatever sliver of info you can gain from the opponent via physical tells, how quickly they called, how they called, etc., to just guess as to if this is a value-hand they are holding on to, or a flush draw. With absolutely no info whatsoever, I'd probably think that any hand that would have called flop bet and will call the turn bet too weighs more towards a flush draw with this demographic. Thus either shoving or at least sizing up enough to make sure you're giving incorrect odds to the villain for his draw (including the implied odds since we are committed). So maybe bet a little higher like 120 (awkward but maybe still gets calls) or just push as a couple others suggest.
Quote:
Originally Posted by ChaosInEquilibrium
I checked PIO with input of our range = {AK,AQs,AJs,TT+,KQs, QJs,JTs} and a few different settings for opponent range, e.g., in the first sim opponent range = {77-JJ,AQo,KQo,AJo, suited broadways (except AK)}. ...
Locked the 33% sizing for the flop bet. Gave two bet options for turn: jam and 1/3 PSB..
PIO jams turn with AK pure. OTOH PIO bets turn for 1/3 PSB with AA almost pure (~80% of the time).
The EV difference for jamming turn with AK versus 1/3 PSB is only $3, so super minor difference.
So the moral of the story is that probably either bet size is fine from theory perspective.
I'm practicing my solver-savy-ness so plugged into GTO+ as well. Used same hero range and similar villain range. Usually how I setup the solver for unknown/random villains in a spot like this where they flat call a raise then flat call the 3bet while now OOP is heavily weighted to hands like TT-77, AQs, AJs, AQ, AJ, some JJ and a sliver of QQ and AK... then some KQs and ATs... and then a sliver of a bunch of other hands this presumably not-so-good-player can show up with. (So like 10%-50% weighting on 44,55,66, 10% weighting to suited connectors, weaker suited broadways, and even the other unsuited broadways. Not perfect but I want the solver to consider the small-chance of all the "weird" hands a bad player can possibly show up with. Honestly I should probably be adding a sliver of other stuff as well.
Exact range matrix used:
I gave more flop-bet-size options for hero, but solver was betting 1/3psb for 100% of range so didn't need to node lock that. I had to node-lock the villain's responses to this bet, because when left to the solver it either folded or raised with villains range - it never called this flop bet. I just made assumptions about the villain and node-locked it to fold all whiffs, call all top-pairs except AK where I presumed villain would check raise some percentage. Then just call all non-nut FD's but mix in check raises half the time when villain had nut flush draw, and do a mix of fold/call on PP's under the K.
On the turn for hero the solver had options of check, 1/3psb, 2/3psb, and shove. For AK The solver bet small near-100%; it did mix in a very small percentage of checks when hero held the Ac in their hand. It never shoved or picked the 2/3psb. (Albeit as ChaosInEquilibrium pointed out when he ran this in PIO with slightly different ranges and bet options, the EV was also very close. It was about a $1.20 difference between betting small and shoving)
Quote:
Originally Posted by ChaosInEquilibrium
Here’s another tidbit:
If you want PIO to start mixing bet sizing options with AK then you need to give opponent more Kx relative to flush draws. In particular, if you add KJo to the above baseline range then PIO is indifferent to the two bet sizes with AK.
So:
Baseline: If opponent has 6 FD combos and 10 worse Kx combos then PIO pure jams AK (EV difference relative to 1/3 PSB is $3/$225).
Baseline+KJo: If opponent has 6 FD combos and 16 worse Kx combos then PIO is indifferent to bet sizes with AK (and mixes with slight preference for jam).
The hands that really prefer the smaller size in the baseline scenario are just TT/KK, which makes sense.
The range I input would have had more top-pairs for villain even if I only have them a sliver of a weighting, but I am surprised this range and options I put in my solve mostly imply betting small vs a jam.
I'm always up for practicing/learning more with the solver so if you want to tweak what we are running & compare further just PM me.