znzznz

I play midstakes and have a fairly streamlined way to model PLO spots with GTORB. I am looking for study buddies who have good knowledge of GTO and must be actively using either Pio, GTORB, SP. PM if interested.

It is known that when OOP, you should be betting when your range is stronger, and checking when your range is weaker.

Before, I thought that generally when your range is stronger, to always bet with no check/raises. When your range is weaker, you checkraise hands that can likely get 3 streets of value vs opponent, as long as you can bluffcatch enough by the river (and still keeping a few strong hands in your range). I feel like when range is weaker, that is definitely correct, though how much you check/raise and cap your raise is debatable and based on how much weaker your range is. If it's a 3bet pot 100bb in PLO and the flop is A44, you have no check/raises really.

This is all based on the completely polarized toy game, where IP has no incentive to bet. However in practice, the more similar the ranges are, and the more protection IP needs, the more incentive he has to bet. Most boards, IP will have some hands in his range that can bet 3 streets for value.

I thought this it would look more like this, when OOP for different range strengths. Betting frequencies are approximate, and assume proper bluff/value ratios. This is also factoring in position, so an even range advantage below would be a slight edge if positions were removed.

edit - check/call and check/raise %s are from your checking range.

Super Strong - Bet 75%, check/call 15% (below min def but not exploitable I think since betting so much). No check/raise.
Strong - Bet 60%, check/call 30%. 0-5% is ok if you can balance it.
Even - Bet 50%, check/call 40%. Check/raise 10%
Weaker - Bet 0%, check/call 50%. Check/raise 7%
Super Weak - Bet 0%, check/call 50%. Check/raise 0%.
Super super super weak - I remember seeing an example where you'd have to defend less than MDF if you are super capped. This I'm really not sure about. On one hand, most people don't turn enough hands into bluffs in this spot and you can safely fold. On the other hand, what happens vs someone who does turn enough hands into bluffs?

As for which hands to check/raise when your range is even to stronger, probably your very strongest hands.

This also assumes fairly uniform ranges though. There are a lot of spots where one player will have more nutted hands, but the other player will have a lot more stronger marginal hands. I suppose in this situation, the player with more nutted hands would just be betting more polar with a wider check/call range.

Be interested to know any of your thoughts. I've come to these conclusions just through play, logic, and an NL range frequency thing I made with excel macros. Curious of how it compares to sims.

It is pretty easy to know which type of hands to raise on the flop.  To pick the best hands at the proper frequency however is a little more tricky.  It doesn't seem to be all about raising your best folding hands that have equity.  I also see good players raising draws they could call the flop with as well.

--------------------

There is a list of things to consider besides equity:

- If I raise certain hands as a bluff, some turn cards may leave my calling range vulnerable.

- At the same time, but a smaller effect, raising range should be balanced to where your value + bluff raising range isn't vulnerable on most turn cards.

- I'm raising as a bluff more than 50% of the time, so it would be good to have some bluff raising hands that can defend vs a reraise.

- Raising bluff with equity is equal to less of a bluff.  For example if you raise 10 hands with 20% equity, it would count more like 8 bluffs.

-------------------

For example, on a flop of 974 rainbow, you defend the bb vs a co open.  The following work as bluff hands:

KJs, KTs, QJs, QTs, JTs, JTo, J8s, 86s, 65s.  How would you guys rank these?  Some of these hands definitely work as a call as well.

.

Hand is a made up hand.  Villain opens 30% from the CO to 2.5x, we defend with a range of roughly 25% (shown in picture below).  With an R of 60, we should be defending at least 20%.  The range shown takes away hands we 3bet (QQ+, AK, half of AQs, most JJ, half of TT).  Villain is a solid reg with no other reads.  The flop comes up 9d5c2c.  He cbets 70% pot.

I colored in Photoshop how we proceed with our range, blue is check/call, red is check/raise, and black is check/fold:

Check/call:
All AQs-A9s, A5s-A2s; A8s-A6s of diamonds and clubs only
A9o+, A8o with a club
KQs-KTs of diamonds and clubs, K8 of clubs
KQo-KTo with a club, K9o
QJs, QTs, Q8s of clubs, Q9s
JT of diamonds and clubs, J8, J7 of clubs
T8s, T7s of clubs, T9s
98s, 97s
87s of clubs
76 of clubs

Check/raise:
J8, J7, T7, T8, 87, 76 of diamonds
55, 33

Some key stats that I came up with:

Check/Call Flop - 57%
Check/Call Strength on the turn - 36% (Made up of at least a 9, strong draws, and floats that hit)
Check/Raise Flop - 4.5% (13/295)
Check/Raise Flop Value - 46%
Check/Fold Flop - 38%

Note:- Calling with the KQ-KT combos is when we have a backdoor flushdraw.  Same with A8o.  Actually most check/calls that make sense only with the BDFD is only counting the BDFD (like A6s/A6dd as well).

Questions:
1) We are check/calling way more than raising.  How does this ratio look?
2) Our check/call strength seems to be low.  The 36% number is hands we would have a very high chance of calling the next street, so unless we float a lot more, we will be folding over 60% to the next barrel.
3) Our optimal bluff frequency is over 40%, and in this situation we are over 50%.  Is this too high?  We also have some equity with our bluffs.
4) How would you change this range?

Say we need 33% optimal bluff frequency in a spot.  We bet 10 combos of value that have 80% vs an opponent's calling range.

Do these 10 combos turn into 8 value combos and 2 "bluffs"?

If it does, to get to 33%, we need to add 2 extra combos of bluffs.  Totaling 8 value combos, 2 "bluffs" (losing value hands) and 2 extra combos of bluffs.  8 value, 4 bluffs, 33% optimal bluff frequency.

However, vs our opponent's pure bluffcatchers that all our value hands beat, we still have 10 value vs 2 bluffs, so only a 17% bluffing frequency.  This seems wrong.  So I am wondering what exactly do we need to consider on river bets.