PLO Matrix recommends 3-betting 6622ds in the big blind vs a utg open raise. This seems crazy to me.

Is this a glitch in the Matrix, or is Matrix crazy like a fox? Interested in what you all think.

What rake structure?

6622ds in PLO trainer (PLO50 rake structure) is a call, but 6633ds and better is a 3Bet.

High stakes with 100 bb stacks.

I'm really interested in trying to understand the logic of 3-betting this hand. What flops are we c-betting? What are we check-raising? What are our bluffs?

With such small pairs I'd be super worried about stacking off to bigger sets. And we don't have any nut straights or flushes in our hand.

I need help unfrazzling my mind!

I'm still reading the Nandez book but I think I can contribute here.

UTG raises, everyone folds to the BB and because you have a ds double paired hand of 6622, you 3 bet. Decisions are easy because you'll hit a set about 25% of the time and when you do, you'll do incredibly well against an overpair that is c-bet and stacking off.

Additionally, because of the ds and the connectivity, low straights, flushes and backdoor flushes add to the value and have potential as well. As I recall, there are certain double paired hands that you would call rather than 3 bet, because they don't work very well together in terms of connectivity - for instance, 9922.

By the way, the above information that I posted is from the Nandez book. To me, the book is well worth the price and I'd recommend it to anyone.

I'm no expert but it seems like a pretty good spot to hit a set on lower boards when you're generally perceived to have high cards/high pairs.

Helps to balance your AA with some of these so you have board coverage, I imagine, just like in hold 'em. Makes you harder to range. Plus you can rep flops that are good for your perceived range (AA) a lot of the time.

If we get 4-bet, I assume we’re calling and can put villain on AAxx fairly safely. If we started hand with 100 bb and are heads up with the 4-betting utg, we end up seeing the flop with 66.5 bb in our stack and 67.5 bb in the pot (assuming all raises were pot sized)

What hands are we automatically stacking off with? Below are my rough thoughts >
1. Any set (do we make exception for A high board?
2. Any flush or straight
3. Any flush draw
4. Straight draw on rainbow non-A high board with at least one backdoor flush draw (4 outs for straight + 4 outs for set + 1.5 outs for backdoor = 9.5 outs total)

Thanks for feedback, this is really interesting topic to explore.

You don't really need to understand everything a solver does. In many instances it's not possible. Human brains have limitations.

If I had to guess, it's a combo of the low card hands that can defend against a 4b which plays poorly multiway and therefore doesn't do as well flatting, and so becomes better as a 3b.

I can only give you my opinion on this. If we have 66.5 bb in our stack and 67.5 in the pot, we're at SPR 1, so pretty much any reasonable connection with the flop at that point and we're shoving (or at least I am).

Let's say that your opponent has stats of something like 30/10/2 over a reasonable number of hands logged, we can you can almost be sure that his 3 bet is Aces or Kings but we put him on Aces.

With a flop of something like 3,6,8 rainbow or 3,6,8 or 3,6.9 with two of of our suit, that would be a stack off. What do you think of a flop with something like A,3,6 with 2 of your suit? Tough call maybe? If the flop is A,6,9 with two of a suit, that might be something that would give you thought. When we're double suited and two of our suit flops, that might be another scenario to go with it. On a flop of something like K,T, 9, even at SPR 1, you obviously can't continue and that's an easy fold.

As far as the non high A rainbow flop with a straight draw, set and BD flush draw, possibilities, we're pretty happy with that if we're at SPR 1, so it's a fairly easy shove, right?

Basically, if we're not reasonably hitting the flop, it's a fold and we still have our 66.5 BBs.

Thoughts?

further question: should we 3bet less with these hands if villain tends to call too wide? (which is common in micro/low stakes)

Looking at the situation if we 3-bet from the BB and just get called by UTG, we'll have a stack of 90 bb (assuming pot size raises and 100 bb starting stack) and a pot of 20 bb, so an SPR of 4.5.

To my newb eye, being out of position in a 3-bet pot with 6622 looks awful. Seems like we'll be flying blind if we flop a set and get any type of action. If opponent has a bigger set we are crushed, unless we have a flush draw to go along with it in which case we're 2-1 dog. If opponent has a wrap-draw or open-end + flush we're a small favorite or dog.

Here's some equities I ran for 6622 vs different hands on a QT6 flop (I didn't specify specific suits of flop cards or UTG, but I did specify 6622 was double-suited, I'm not claiming these equities prove anything definitively but think they are still useful to look at). I chose ones that could plausibly end up playing a big pot.

6622 v QTxx (top-2 pair)
64% - 36%

6622 v KJxx (oesd)
62% v 38%

6622 v AKJx (13-out wrap draw)
58% v 42%

6622 v QQxx/TTxx (top/middle set)
11% - 89%

6622 v KQJT (2-pair + oesd)
52% v 48%

6622 v KJ9x (17-out wrap draw)
52% v 48%

My very rough intuition is if an opponent will regularly stack off with naked top-2 pair or open-ended straight draw maybe we're not in such a bad spot. Admittedly we'd need much more sophisticated equity calculations that take into account the frequency of different flops with opponents' ranges.

Simply looking at when you flop a set is probably not how you want to look at a hand, especially looking at specific combos of your opponent's hands.

Since most of the strength of this hand comes from fishing from a set (I assume), I just wanted to take a look at plausible scenarios when we do hit a set. I figured bottom set with a straight draw on board would be one of the more common scenarios. I know it's not comprehensive, but I thought it would be instructive to look at scenarios where we hit what we're looking for.

Well okay, how are you missing out 1 pair + gutshot from the range you give your opponent?

What about the mere fact that you get to cbet and take it down on a lot of boards?

This is, as you say, very rough, but it's rough to the point where it's just noise and you're missing the point and spending energy, as I've said before, on the wrong things. You're theory crafting spots that are too big and where you have too little experience to know how to theory craft it. Start small and then piece together once you have enough small parts understood.

1. I'm starting small > I'm looking at a small slice of plausible scenarios and trying to think through how they play out.

2. I'm not theory-crafting. I'm not making any declarative statements that 3-betting is wrong. I'm trying to figure out where the value of 3-betting 6622ds from the BB against a UTG raiser is coming from.

3. When I wrote post I didn't think average opponent is stacking off with 1 pair + oesd against a preflop 3-bettor, I could be wrong.

4. I'm assuming a lot of the value from 6622ds comes from its potential to hit a set, so I'm thinking through scenarios where that happens.

5. Yes, the hand also gets strength from its double-suitedness and straight capabilities. Yes, there are many other possible flops. I can't include all of them in one post.

How would you suggest I "start small and then piece together once you have enough small parts understood"?

If you look at them to learn how to play them out, then yes, it's a small slice.

But if you're looking at them to try understand why you're 3betting it in the first place, then you're trying to understand an entire puzzle holding only a small piece.

What do you mean this isn't theory-crafting? If this isn't theory-crafting, what is? Essentially all strategy construction is theory crafting. The point is that theory will never be able to cover all bases, so good theory crafting knows its limitations and is able to approximate its margin of error based on it.

Not knowing the other pieces, you're not able to do that.

How? Do what coaches and/or solvers tell you in terms of general strategies and earlier street, and try to work your way back from the river or the last decision.

Once you have a good understanding the various situations you get into on the river (or say the flop if it's a 3b decisions) you can then work your way back to an earlier street and start piecing together working heuristics and overarching ideas. Right now you're trying to understand the entire picture when you've barely done part of a corner.

You will never understand things thoroughly at the very start. Lots of highly advanced disciplines have basic facts that would be highly counter intuitive to someone who knows nothing. Do you think it's possible for someone who's never been told to figure out that the earth is not flat for instance? This is humility.

Accept that that understanding is a progression and focus on what small nuggets of understanding is achievable, and for the rest, just go with what works or what solvers/coaches say until you're ready for the bigger pieces. Asking questions and being inquisitive is good, but learning takes drilling and time, and you have to ask the right questions at the right time. I've already said this multiple times, so this will be the last time. You are your own master.

So here's a good example. You realise when you 3bet, you're not just either stacking off or not stacking off on the flop right? Like, you can bet 50% pot or even 33% pot sometimes, and there are multiple streets to go. So you're literally only focusing on certain boards and looking only at flop stack off scenarios, and you're hoping to understand the preflop decision? So by only looking at stack off ranges you miss some of the most common occurrences that will happen after you 3bet and cbet. Not only that, pair + gutter also has a lot more combos than hands like "two pair oesd" and so on. Also, none of your categories have combo weightings, so some of those combos occur many many times more than the other combos, so even with all the numbers presented, how can you actually know what the average EV you have is? (Okay to be fair, you can input this stuff into PPT and get an answer, which would be a bit closer, but is still a bad approach. In fact I'm, surprised you didn't just plug this whole board and hand combo into PPT)

This is possibly not entirely obvious to you, but the fact that you say this is a HUGE indication that you're not ready to extrapolate heuristics yet.

Learning takes a leap of faith. This is why you have to choose sources of knowledge you trust. I'm not your coach, so you can believe whatever the hell you want, but I'm telling you you're trying to extrapolate with too little data. Half truths misused can be way worse than even flat out lies; it lulls you to a false sense of security in what conclusions you make. Just accept you don't have the whole picture.