3bet range preflop

Is a person with a 3bet range of say 11%, more likely to be 3betting a polarized range or a wider strong range? Is there a general preference for either strategy that you guys have?

depends on their positions and other factors

Yeah that's what I'm trying to get at. Do we expect people to use a mixed strategy based on table conditions, or a more fixed strategy of either polarized or not. I mean in personal experience from midstakes 6max cash.

That's probably impossible to generalize.

I'd imagine a 3b less polarized IP, prob 3b stuff like KT-KQ,QT-QJ type stuff IP if I felt a cold call wasn't profitable but a 3b was.

I have a tendency to 3bet such hands more out of position, and call IP. The reasoning is that they are good enough to play, but too hard to play out of position. This is especially true for being in the SB, where a call might invite the BB to come along too. I'm trying to revise this sort of thinking though.

Originally Posted by jackvance

I have a tendency to 3bet such hands more out of position, and call IP. The reasoning is that they are good enough to play, but too hard to play out of position. This is especially true for being in the SB, where a call might invite the BB to come along too. I'm trying to revise this sort of thinking though.

My logic is that if we call IP, and facing heat on these boards when we spike them, we're likely up against a stronger range (their EP range). By contrast if we call in the BB (or less often in the SB), we're keeping in a ton of dominate hands (high LP stealers will have all K2-K9 stuff, Q2-Q9 etc etc).

Also 3b IP with hands like QT, KT, might get folds of dominated hands like QJ/KQ etc.

Originally Posted by griffey24

3b IP with hands like QT, KT, might get folds of dominated hands like QJ/KQ etc.

These also block v 4bet range.

This is very interesting. But I have 2 reasonings for why I do this:

- There is simply way more spots to 3bet out of position than in. 1/3 you are in the blinds, then UTG, so only 3 spots give you the chance to 3bet, and depending on only 1, 2 and 3 out of 6 people, respectively. Shifting to more 3betting IP will just drop my VPIP a lot

- I just like to see flops IP, because it upgrades the value of my hand due to my position. I feel like if I 3bet more IP, people will fold. Ok this is fine if I think about it more, but to extract value sometimes you wanna see flops with people where they can make mistakes. And people OOP have it rougher.

I am considering if I should shift my ranges more towards what you are saying, which would make me tighter. In part because this could be a leak of mine that is causing big swings.

But if you are going to 3bet more IP, what does your call range look like IP?

Originally Posted by griffey24

That's probably impossible to generalize.

I'd imagine a 3b less polarized IP, prob 3b stuff like KT-KQ,QT-QJ type stuff IP if I felt a cold call wasn't profitable but a 3b was.

Wouldn't this technically still be referred to as a polarized range, since we would be 3betting the best hands that we felt a call wasn't profitable with? Whereas, with a depolarized/linear range, we are choosing to 3bet some hands we could profitably call with, expecting the 3bet to get more value due to villain calling with a worse range?

Shouldn't villain's tendency when facing 3bets play a large part in helping us choose which range to 3bet? With us 3betting a more polarized range against players that are playing primarily a 4bet/fold game when facing 3bets, and us 3betting a more depolarized/linear range against players who are primarily calling/folding when facing 3bets?

So that against most fish, it leaves us 3betting a depolarized range both IP/OOP. Placing hands like our better suited/offsuit broadways into our 3bet range. Unless he's flatting a lot of 3bets and folding frequently to cbets, in which case we could 3bet tons of hands profitably.

And against other players, such as regs, we typically have a polarized range IP, and depolarized OOP? Since a lot of regs only 4bet/fold 100bb deep OOP, but flat wider when they open from CO/BTN and get 3bet from the blinds.

So assuming not against a fish, IP against an EP/MP raise, our 3bet bluff range will consist of hands like offsuit broadways/A5o/K9s/T8s/etc that we can't profitably call with, but that hold blockers and/or are suited (in case he does call). Whereas, IP BTN v CO, we can more aptly call with some of our better offsuit broadways, shifting our 3bet bluff range downwards towards more suited gappers/Ax/Kx/Qx type hands.

And OOP, against a CO/BTN open, where we expect villain to flat more 3bets, we can start placing hands like suited broadways/good suited connectors and gappers into our 3betting range that play well postflop. While flatting those steals from the BB with offsuit broadways keeping in dominated hands.

Overall, OOP I probably don't 3bet much against EP/MP opens unless they are either folding quite often, or calling quite often. This might be a mistake (idk).

Assuming you guys have big enough hand samples in your database, I'd advocate running filters for different sets of hands when PFR = false, and cold call = true.

I did this a few weeks ago. It's surprising how many hands are just not profitably cold calling (even btn vs co, assuming blinds are aggro).

I think the generally accepted definition for polarized range is that part of your range is clearly for value and part of your range is clearly a bluff. 3betting KTo, I would be hard pressed to say this was for immediate value or not. KTo flops well, if I have position I can pot control, and I can get a fold with my 3b.

Nice tip Griffey, gonna do that after study. And I agree with your definition of polarized, although I really like the idea that p4 and stackes used, ie that you have blockers for top hands but you remain aware you're essentially doing a bluff raise. Makes me wonder what the statistical relevance of your blockers is in these spots, and how it weighs up against just picking an opportune time (ie with ATC).

Originally Posted by griffey24

Assuming you guys have big enough hand samples in your database, I'd advocate running filters for different sets of hands when PFR = false, and cold call = true.

I did this a few weeks ago. It's surprising how many hands are just not profitably cold calling (even btn vs co, assuming blinds are aggro).

I've ran those filters recently looking at the profitability of cold calling versus 3betting BTN v CO with various groups of hands (suited connectors, offsuit broadways, suited aces, etc). And in every case, 3betting was by far the most profitable with all the hands I felt were probably borderline.

Although, I think I play rather poorly postflop without the initiative.

Originally Posted by griffey24

I think the generally accepted definition for polarized range is that part of your range is clearly for value and part of your range is clearly a bluff. 3betting KTo, I would be hard pressed to say this was for immediate value or not. KTo flops well, if I have position I can pot control, and I can get a fold with my 3b.

I suppose that 3betting hands like KTo is going to leave a less polarized range than if you are 3betting worse hands like T6s, etc, which is essentially what is commonly referred to as polarized. But 3betting KTo against a player that is typically only continuing via a 4bet (most regs when they are OOP) seems like clearly a bluff. Since anytime he continues, we will be folding to his 4bet (unless we are opting to bluff shove some of those hands). So the only real difference between 3betting a hand like KTo and T6s there is blockers and the increase in pot equity the rare times they do flat.

If we feel we are in a situation where we can't flat hands like KTo, QJo, etc, why would we ever really be 3betting worse hands than those? I mean those offsuit broadway hands would be the best hands we would be folding, and thus seem to make the best sense to 3bet bluff with.

So it seems like the only time we are really 3betting a traditional polarized range, including the top of our range for value, then hands that are clearly bluffs [T7s, 94s, K4s, etc] would be if villain is folding to 3bets very frequently or we have a wide cold calling range in this scenario.

If it's in a spot where villain is more likely to call the 3bet, such as a fish anytime or regs IP LP v Blinds, then 3betting those KTo/KJs/QTo/etc type hands aren't as clearly bluffs. As we will get to realize the hand's pot equity more often when villain flats, rather than nearly always folding when villain continues. So more for thin value than anything.

Spent 2 hours in Poker Equilab to look at ranges and their equities. I had to think of this thread again, and another question. I have it stuck in my head that a 1:2 bluff:value hands ratio (so 1/3 bluffs) is optimal as an unexploitable strategy, am I remembering correctly or is this about bluffs? (where you bet pot I guess?) What is the mathematical reasoning for a bluff frequency? I find it hard to calculate because there are so many factors, like not just the equity of a hand but how they play on different kinds of flops, the power of position, adjusting to calling frequencies, adjusting to the raisesize-to-pot ratio, which even differs in the SB and the BB (suggesting higher raises in the blinds I think). But I'd love to hear how others look at this.

If you're closing the action in a hu pot on the river and the person before you bets the pot then you need to call with at least 33% of your range to not be exploitable. i.e. Alpha Value.

Bar that it sounds like rubbish and I see no reason why it'd hold true in terms of bluffing frequency, but I could obviously be very wrong.

I'm also not sure why you seem to want to polarize your range in what seems like every spot possible.

I'm thinking about it in terms of what an optimal strategy would look like, and why. Not to use a fixed range but to have an idea for where the optimum lies. For example, blinds 1/2 6max, UTG (30/25) opens to 6 and I reraise to 18. In terms of pot, I risk 18 to win 9. What is my optimal percentage of total garbage, from a mathematical perspective, in this spot?

It depends on his opening range, how much he folds to 3bets, how he plays post flop, his 4betting frequencies, how he views you and how much he's been adjusting to what you've been doing, how all the people left to act play, etc etc

3bet bluffing him at all could be completely not worth it in some circumstance. I don't think there's a number you're going to get out of this.

I seem to remember a post from many years ago from sauce (or some other pro) on 2+2 that a 1/3 garbage ratio is optimal. But my memory is fuzzy and this 1/3 could have been about how often you want to bluff compared to value bets in a certain spot, not sure. That's why I ask if anyone knows what the generally accepted ideas on this are (or if there are any, or just what the pros have to say about this).

http://en.donkr.com/forum/optimal-3-...-part-1-533561

I recommend you give that article a read. I need to do so again and take some notes on it myself.

To answer your question about value:bluff combos in 3bet ranges, I think the number you are looking for is 40% value (intending to shove over a 4bet, even 5b bluffs like A5s fall into this category) to 60% bluffs (folding to a 4bet). This is assuming that the villain is only going to play a 4bet/fold strategy when faced with the 3bet, such that you don't get to realize any of you bluff hand's pot equity (since you always fold when they continue).

So you determine what hands you can 3bet/shove for value (have greater than 50% equity against villains 4bet/call range). Then you add in 5b bluffs with hands like Axs/etc. Enough combos so that the worst hands in villain's 4bet/call range is breakeven. This gives you your value 3bet range (shoving over 4bet). For every combo in your value 3bet range, you want 1.5 combos in your bluff 3bet range (folding to 4bet) to reach the 40:60 value:bluff frequency.

Fwiw, this is assuming 3.5x open, 12x 3bet, and 27x 4bet. The frequencies change a little depending on sizing.

I think your 1/3rd bluffs kind of rule is more applicable for river bets, because these close the action.

So if you're betting pot, then villains calling odds are 2:1, so he needs to be good 33% to break-even on his call, so hero needs to have 33% bluffs to be indifferent if villain calls or folds.

But in your example pre-flop, if you 3b from 6 to 18, then you are risking 18 to win 9. So for immediate gains from your 3b , he needs to fold >66.6% of the time. But maybe he's horrible post flop and maybe he calls often pre but folds to a half pot cbet (cbet of 19.5 into 39) like 80% of the time. But a half pot cbet only has to work 33% of the time. So even if he's not folding >66% of the time pre, we're still making money on our cbet etc.

I don't think any bet before the river is quite as cut and dry.

Originally Posted by griffey24

I don't think any bet before the river is quite as cut and dry.

Even in the scenarios laid out in the article, there has to be assumptions that are always adhered to. Mainly that villain's only possible response to a 3bet is to fold or 4bet. When in reality, they have the option to call. Which would change the situation dramatically, as now our bluff hands would get to realize a portion of their pot equity, as well as we would get to utilize postflop fold equity, etc.

But I think for the most part regs tend to play a 4bet/fold game when OOP facing a 3bet (100bb deep). So I think it's potentially beneficial to strive for the 40:60 value:bluff frequency in spots that aren't sure of the exploitative strategy you should be using.

Originally Posted by Stacks

http://en.donkr.com/forum/optimal-3-...-part-1-533561

Great link Stacks. Everyone should read this article, it's very very good, and written in a clear/concise way for anyone thinking about implementing GTO into their game. Thanks for posting!