I am trying to figure out how to appropriately divide my range on a wet flop. Since there are many possible ways to build a balanced range, I just want to know if my approach is reasonable, and what improvements you might make to the ranges I’ve constructed.

Situation
Effective stacks are 100 bbs. You raised to 3.5 bbs from CO and are called by BB. SPR after flop is 13 and BB checks to you on the flop. The flop is K 5 3.

Assumptions:

1. Villain plays well.
2. Equities for reasonable starting hands against the 10% of hands with the most equity on this flop generally range between 30%-70%, so you have no pure bluffs.
3. Want to c-bet about 60%.
4. Villain will continue against a c-bet with at least 70% of his range.
5. With polarized ranges, we would want to bet close to full pot, but with merged ranges we will want to bet smaller. Most ranges on this flop have decent equity so ranges will mostly be merged rather than polarized. Much of villain’s range will continue whether you chose to bet full pot or half pot. The main benefit of a bet will be to deny villain’s equity with marginal hands that he will fold even to a small bet. I chose a bet size of ½ pot.
6. If you bet ½ pot and get check-raised, you need 46% equity to stack off.
7. If villain check-raises full-pot, he risks 18.75 to win 11.25, so we need to continue with at least 40% of our betting range to prevent him from check-raising profitably with any 4 cards.
8. We should never slow-play on a dynamic flop like this.
9. We need to check behind with our weakest hands as well as some of our stronger hands to protect our checking range. (Good candidates might include hands with potential that don’t want to get check-raised like Middle set-No low Draw, or Bare Nut Low Draw, or Naked NFD)
10. Bet/folding hands would include our weaker betting hands that have poor nuttiness.

Methodology
In general, this is a better than average flop for strong starting hands. I spent some time with Pro Poker Tools to get a feel for which hands perform best against other strong hands.

1. I ordered categories of hands based on how well they performed versus the top 10% of PPT starting hands.
2. I re-ordered the hand categories based on how well they performed versus the 10% of hands that performed best against PPT’s 10% starting hands. The second step was a fair amount of work and the results are not that much different than just doing step1, so it may not have been worth it, at least for this type of flop.

Observations:

• All of the top 1% performing hands have some kind of low draw.
• The only hands to make it into the top 12% without some kind of low draw are top set and set + FD.
• The very best hand (Top set + NLD) has 77% equity against other top 10% hands.
• The worst hand in the top 10% (FD + OESD) has 48% equity vs other top 10% hands.
• A 50th percentile hand (Top two with no low draw) has 31% equity vs a top 10% hand.
• The worst hand in the top 70% (Bad LD w/ no High) has 26% equity vs a top 10% hand.

Ranges: Flop K 5 3

Action	Hand Rankings	Example Hands	Eq vs Top 10%
Bet/Raise	Top 1-12%	Set + Low Draw	77% - 46%
		Top Set
		NFD + LD
		FD + 2nd Nut LD
Bet/Call	Most of Top	Top Pr + Bare NLD	46%-35%
	12%-40%	FD + Bad LD
		Top Two + Bad LD
		NFD+Pr-No LD
Bet/Fold	Most of Top	FD+pr-No LD	35%-29%
	40%-60%	Top Two – No LD
Check Behind	Worst 40%	Bare LD- No High	29%-9%
		Aces – No LD
		Bottom Two – No LD
	A few hands	Mid Set – No LD (Rank 13)
	in top 60%	Bare Nut Low Draw (Rank 32)
		Naked NFD – No LD (Rank 45)

Questions:
I know this is a lot to think about so feel free to focus your criticisms on only part of this post. In particular, I am interested in any or all comments on the following:

• Does my methodology make sense?
• I used the equity versus the 10% of hands with the best equity on this flop. Would some other percentage have been better (say top 20%)?
• What is the best bet size to use?
• If you chose to bet full pot, how would that change your strategy? I assume you would c-bet with fewer hands.
• Should we have a bet-folding range on a wet flop like this?
• How much different is this strategy than the way people play in real life?

PM send

So you're down to two-handed after the flop. Right?

Why did you choose 60%? Wouldn't it depend on your opponent?

Again, wouldn't it depend on your opponent?

I think polarized ranges is more of a Texas hold 'em concept than an Omaha-8 concept.

Interesting.

Yes. Very interesting.

I think it's opponent dependent... but generally probably top 20% is probably more realistic for most full tables.

I think it's opponent dependent.

I think you're making your strategy too card dependent, rather than opponent dependent. Maybe that's the way to do it, but it's not my way. I tend to make my strategy more opponent dependent than card dependent. (I know you didn't want to hear that, but it's pertinent to my response).

I think it's opponent and history dependent.

That's something worth thinking about. (Alas, I don't have an answer).

Buzz

Right.

I don’t have a very good reason. It seemed reasonable and I thought a 60% c-betting frequency was probably what good players might use in a heads-up pot.

Well, I’m assuming we’re playing a good villain and if we c-bet ½ pot, then we're getting 2 to 1 odds on our bet and he has to continue with at least 67% of his hands to keep us from having a profitable bet with any 4-cards. Since our bets won’t have zero equity when called he will actually need to continue with something more than 67%.

You’re probably right that it applies more to holdem than to O-8, because you might be going high or low or both. But I think it probably applies at least somewhat in Omaha, especially on the river or when no low is possible.

Yes, you’re right of course. If I knew enough about villain, I could use exploitative bet sizes or frequencies, and that would certainly be a more profitable way to play. But I’m trying to come to some general conclusions of how you should play in a given situation versus a very good player or when you don’t have any reads. Also, if you have a good understanding of how you should play versus a very good player, it makes it easier to adjust your play based on opponent tendencies.

Some very interesting stuff here. Thanks for sharing.

Since BB in particular should be defending pretty wide (67%?), and he's continuing against your 60% c-bet 70% of the time, I think 10% seems too narrow. Maybe 25%?


I would c-bet with different sizing - something like pot top and bottom of c-bet range, and then 1/2 to 2/3 in the middle. I might occassionally bet smaller when I smash the flop (KK2sAs) if villain likes to check-raise weakish c-bets. So I wouldn't necessarily vary my c-bet%, just my c-bet sizing.


I think so but what would be an example of 'FD+pr-no LD' that you would bet/fold? Having a hard time trying to correlate to your PF raising range, unless your raising range is uber-wide from the CO.

Thanks. Buzz suggested I use a higher percentage too. If you want to see how you are doing against the best hands on the flop, using a higher percentage (at least 20%) would be better. My thinking was that the times it mattered most would be when you got all-in and when the money goes all-in you are more likely to be looking at a 10% hand. But I’m reconsidering that.

That makes sense, especially in a real-life scenario. I was thinking an observant villain might be able exploit a bet sizing tell if you based your sizing on your hand strength, but maybe we could balance it along the lines of what you suggest. However, unless it materially improves our results, I prefer sticking with a single bet size because it simplifies an already complicated strategy.


Maybe a hand like AdTd9s3s which is a top 9% starting hand according to PPT, but it does poorly on this flop especially against a strong check-raising range.

I was thinking higher percentage in terms of looking at what/when to c-bet, but agree that when it matters most a 10% hand is good to use.

In answer to one of your other questions, single bet-sizing is pretty common at the lower stakes on Bovada I play at, so your strategy does not seem all that different than real life to me (though no way that 98% of Bovada players have done anything close to this research). Someone like amok would probably advocate different bet sizing, and it would obviously be required against him or even a less-competent but observant villain.

Since you're using 10% and villain won't always have a 10% hand, I think pot would be the default because if you're this good equity-wise against 10% you are probably even better against his actual hand. So building the pot isn't terrible, and of course you should get greater fold equity as well.

For some reason I was realizing a 3 counterfeited your low draw but not that it qualified as a pair. Can and would bet/fold this hand, so thanks for the example. But 20% still seems like a high % in the bet/fold category. Do you mind telling me the actual number of hands that equals (if you still have your calcs)?

I don't follow. (I don't see how Villain is forced to continue with at least 67% of his hands).

Let's make the pot "P." Then our half pot sized raise is "0.5P."

If Villain re-raises, he may re-raise to 2.5P. (From P+0.5P +0.5P+2P). In that case, calling Villain's pot sized re-raise costs us 2P, making our total contribution this betting round 2.5P. (Or we could fold to Villain's pot sized re-raise).

I don't follow why Villain is forced to continue with at least 67% of his hands.

Again, I'm not following. Perhaps you could explain where the 67% came from.

Good point. Opponents will often bet a scare card on the river... or opponents putting Hero on a particular draw or draws will often bet a brick for those draws on the river.

That's a very reasonable point of view, and of course we all probably do about the same thing, without formalizing it as well as you have done.

You have some interesting ideas... food for thought. Thanks for sharing.

I know you want feedback, but if you want an expert opinion, you may have to pay for it. There are many better Omaha-8 players than I am, some of them post on this forum and some of them are interested in coaching or helping others for a fee. We don't allow advertising for coaching or helping others on this forum, I'm not interested in coaching myself, and my policy is not to recommend anyone as a coach.

If you're interested in paying for an expert opinion, I suggest you read posts on this forum and come to your own conclusions regarding who would be most helpful to you.

Beware of getting scammed.

And of course you're welcome to continue posting your ideas here.

Buzz

There are 9099 hands out of a possible 211876 hands that fit into the FD + pr – No LD category. Here’s the PPT sysntax if you want to try it:
(ss!As : (K,5,3)) : (*!{[A,2,4,6,7,8][A,2,4,6,7,8]})
However, the harder part is eliminating hands that have already been included in a higher ranked category, and after doing that, there are only 6361 hands remaining in that category.

There are several other hand categories included in the top 40% to 60%. For instance the category,
Middle Pocket Pair w/ 3rd NLD or worse contains 16968 hands but after eliminating the duplicates there are only 8822 remaining. These hands have 29% equity vs. my top 10% hands for this flop.
Here’s the PPT syntax for the entire group.
(QQ-66) : ({[A,2,4,6,7,8][A,2,4,6,7,8]}!(A2,A4))

Yeah, I probably didn’t explain that very well. You’re looking at it after the villain raises our c-bet. I’m looking at it, at the point we make our c-bet and before he acts.

Imagine that we c-bet ½ Pot with 100% of the range that we got to the flop with and that villain only calls or raises 50% of the time. Also imagine we tried to steal the blinds with a hand of 2222, which has close to zero equity.

Amt we win when Villain folds = P
Amt we lose when Villain calls or raises is the amt of our c-bet = ½ P
Villain Fold Pct = 50%
Assume our Equity when villain calls or raises = 0. In reality it is something greater than zero, so this is the worst possible case for us.

When we c-bet with 100% of our hands, our equity is:
50% x P – 50% x ½ P = .25 P

Villain cannot let us automatically win .25 P when we c-bet, else we have a very easy strategy of always raising preflop and always betting ½ pot on the flop and realizing .25 P profit every hand we play heads up against him. He must call our ½ Pot c-bet at least 2/3 of the time to prevent this simple strategy from succeeding.


There are many better than me too, and I’ve read enough of your posts to know that you’re one of them.

Thanks for the advice. I’m considering the expert-opinion option, but in the meantime there are plenty of smart and capable people on the forum who might be willing to share some of their thoughts.

flop is K,5,3.
OK, we can make 211876 different four-card hands from the 49 cards that remain. (math is 49 choose 4 = 211876). OK.

Since 20% of 211876 = about 42 thousand, I don't follow the significance of the 9099 or 6361. (Sorry... I'm trying to follow your logic).

I presume FD = flush draw,
pr = pair, and
- No LD = low draw.

Flush draw
Heads-up any flush draw would be nice. With more opponents I'd feel safer with nut or 2nd nut flush draw.

pair
I'm wondering why anyone wants a hand with any pair.
A hand held pair that makes a set, for example KK**, 55**, or 33**, would be good, but any other hand held pair, QQ**, for example, is behind four random cards.
3*** and 5*** make pairs with the board, but are about even with four random cards.
K*** would be good against one opponent with a random hand.

Not quite clear to me what the pair is.

low draw
It's not immediately obvious to me why you'd write "- No LD" rather than simply + LD. (Doesn't matter).

But unless we manage to get all-in on the second betting round, the betting is not over. There are two more betting rounds yet to go. Seems to me you're ignoring them. Or am I missing something.

You're welcome.

We hope so. Interesting to read your ideas, but you might do better getting expert responses if you make shorter posts.

Buzz

for clarity, you are proposing that you open-raise from CO with 100% of dealt hands?

consequently, you are building a c-bet range of 60% of the 100% of dealt hands?


and perhaps of less consequence, do you imagine your 'worthy' big blind to have just a single defending range(he never 3-bets from the bb), or does he have a 3-bet range and a calling range?

Right, there are about 42,000 hands ranked between the top 40% and top 60% of all hands. Within these 42,000 are a bunch of different hand categories, for instance FD + pair/No Low Draw which contains 6361 of those 42,000 hands. If you added the number of hands for each category in the 40% to 60% range, it would sum to about 42,000
Correct.
Well, a pair in O-8 isn’t a hand you aspire to make, but if you happen to have one, it is better than nothing as sometimes it wins high or improves to two pair which wins high. I’m not trying to say these are good hands, but they have to be included somewhere. An example of a hand fitting into this category is AdTd9s3s, which is not a good hand on this flop. In fact, it ranks worse than 48% of the other hands you could have on this flop.

I categorized pocket pairs differently than hands that make pairs with the board. Any way you choose to group the hands is a little bit arbitrary, but I tried to group hands with similar characteristics and that most people would probably play similarly together in the same category.
Yes, but if he folds there is no more betting. And that is the case, I’m talking about. If he folds more than 33% of the time we automatically profit. If he continues by either calling or raising, we can always just fold the first time we face a bet from him, and we will lose no more money than the ½ pot c-bet we invested on the flop. In reality our expectation is better than that because, sometimes we will win even if he calls, and sometimes our hand will be extremely strong and able to profitably call a raise or get all-in.

So I am ignoring future betting rounds on purpose, because you are never forced to put more money in the pot after you c-bet, and even if you never put another dollar in the pot after you c-bet, you will still come out ahead if he folds more than 33% of the time to your c-bet.

Excellent points. My postflop sims didn’t take into account the preflop ranges of the players. I hadn’t thought it would make that much difference in approximating postflop ranges, but the more I think about it, the more I think that’s a major flaw that I need to go back and correct. Thanks for pointing it out.

i also noticed, in terms of characteristics of your 'hand categories' you leave out connectedness.


and really leaving the forest for the trees,
i would think the range is slightly off/needs a little more added--not flush draw, not trips and not pair of kings(the 4th card is possible when you include not trips)
(QQ-66) : ({[A,2,4,6,7,8][A,2,4,6,7,8]}!(A2,A4,k,ss,rrr))

Actually, I think I do handle connectedness somewhat, but I didn’t provide enough examples of the hand categories to make that evident. However, I think I need to do a more thorough job of it and the example you provide, suggests a better way to do it. Also, had you not provided your suggested PPT syntax, I would have thought I had already accounted for all 4 cards with: (QQ-66) : ({[A,2,4,6,7,8][A,2,4,6,7,8]}, and not realized I needed to add a “no trips” constraint. Tracking down a bug like that might have taken me a long time.

Thanks very much for taking the time to review my approach carefully enough to identify some major flaws and weaknesses. It’s been very helpful.

"SADS" pm?

(strategy advice discussion suppression)