Why 3bet?

[Pokey]

[ QUOTE ]

See Pokey, that's the whole reason I'd prefer to call with AJ.


[/ QUOTE ]

OK, I think I'm doing a ridiculously bad job of explaining myself. Here's my opinion on this matter, but I warn you that I may range over several topics in this rambling diatribe.

Let's start with this: being out of position sucks hot sweaty monkey balls. You act with less information on every street. You have a much harder time manipulating the pot size. You cannot take free cards easily. You win less when your hand is good and you lose more when your hand is bad. It sucks completely. As a result, you should try very hard to play as few hands as possible while out of position.

That having been said, you will occasionally have to play a hand out of position or else you'll give away FAR too much value. If you fold AA in the big blind just because you'll be OOP, you've made a huge mistake (obviously). You will have to fold far more hands than usual, and doing so is the most +EV choice you can make (OK, the least -EV choice, but you get the general idea).

Unfortunately, your savvy opponents are aware of this fact. They have the nerve to try to abuse this fact by trying to steal your blinds (the bastads!). So we now have a tradeoff: wait for superpremium hands that can survive great heat even out of position or mix things up with lesser hands on occasion. If we only wait for superpremiums we'll be folding way too often, and a clever opponent will steal our blinds mercilessly, which can be extremely profitable for the enemy: if they steal the blinds just once in every five orbits this adds 0.75 PTBB to their winrate every thirty hands, or +2.5 PTBB/100 to their overall winrate. (No wonder they try it so often.) Just laying down and dying whenever our opponents try to steal becomes so profitable that they increase the frequency of their steal attempts, and then we're in even WORSE shape.

On the other hand, we can try to fight back on occasion. The downside is that we're going to be out of position all hand long and we're not in control of the betting. If we smooth-call we're likely to face a flop continuation bet; at that point, things get REALLY sticky for us. Assuming we're not totally thrilled with our hand (and if we're mixing it up that'll probably be the case at least 2/3rds of the time, and probably more often than that) then we've got two options: go for a check-raise (commiting a VERY large amount of money on a bluff) or just check-fold (walking away from 4xBB instead of the 1xBB that we would have lost if we'd just folded in the first place). The check-raise bluff is a hefty expense: if villain raised to 4xBB preflop then we've got about 8.5xBB in the pot now, and a standard c-bet will be 6xBB or so. If we make a real check-raise that will be to at least 18xBB, at which point we're committing almost a quarter of a stack on a hand that is dog crap -- we've created a big pot with a nothing hand while out of position and miles away from a showdown when our opponent has twice said that his hand has strength. That's ugliness incarnate.

So: I don't like autofolding to steals.

I also don't like calling and "playing poker" postflop because our expected benefit is so small and our potential costs are catastrophically high. We can't play for set value because we won't be able to get our opponent to commit 40xBB postflop with any sort of reliability, and we can't check-raise on a bluff very often because our opponent's hand will often be stronger than our own, and we're still two streets away from a showdown with no obvious folding equity and an opponent in position.

My choice is the three-bet. Heads-up and out of position, I want to at LEAST be entering the flop with control over the betting. Barring a specific read (e.g., "villain frequently three-barrel bluffs" or "villain gives up after c-betting" or "villain cannot fold a hand after raising preflop") I want to take control preflop. A raise will deny my opponent the really deep stacks that give him the maneuvering room to outplay me badly postflop, it wrestles betting control from my opponent forcing HIM to make a hand on the flop instead of the other way around, it disguises my hand, and it helps me to understand my opponent's hand better based on his response to my three-bet.

You raise an excellent point in saying that perhaps AJo isn't the best hand for this kind of move, and I agree. I'd vastly prefer to three-bet with hands that have some potential to flop something powerful: suited connectors, pairs, suited aces. Along with the natural powerhouse hands of QQ+ and AK, this puts you at 12.5% of your hands, so you'll be restealing one time out of eight. For most players, this is FAR more often than they'd usually consider three-betting OOP, but still infrequent enough that your opponent is going to have to give you some respect when you make your three-bet.

The difference between three-betting with AA and three-betting with 72o is actually somewhat small in terms of practical outcomes: the hand is rarely going to go to showdown, and you're very frequently going to win preflop or on the flop. Everything else being equal, the two will have different outcomes probably less than one time in ten. In fact, against an amnesiac who likes to steal the blinds you would probably have a +EV strategy if you three-bet 100% of your hands.

Of course, real-world opponents aren't accomodating enough to forget our previous actions between hands; as a result, we need to choose our battles carefully. We can't resteal too often or our opponents will (correctly) call down light, making even MORE profit from our mistakes. That's why we restrict the list somewhat. I'd rather three-bet with A3s or 44 or 87s than AJo because of all the reasons you mentioned above. However, I think that if you've come up with an arbitrary list of 10-15% of your hands that you wanted to three-bet as a resteal, and if AJo was one of them, then your strategy would be fine. In fact, you could even have the following three-bet plan:

QQ+, AK: always three-bet
Anything else: roll two dice and steal if they total nine.

Now you're stealing about 13.5% of the time and with highly unpredictable hands. While I'd PREFER that you steal with hands that had slightly higher odds of flopping well, the difference really won't be that significant, since you're not really going to showdown with them.

In short (too late) I'd want to three-bet rather than smooth-call to force my opponent to have a hand both preflop AND postflop, rather than smooth-call and only win when *I* have a strong hand. Showdown strength is nice, but folding equity is king in steal/resteal situations.

[DaycareInferno]

here's the thing though, pokey. even against villains with super wide stealing ranges, you can only 3bet them so much before things get out of control. since there are pretty big limits on your 3betting frequency, you're left to choose which hand types that you would rather roll with in those low showdown, 3betting scenarios. i would rather have a hand that either hits a homerun or is easy to dump. i'd rather not roll with a hand that runs a high risk of either being folded out by a worse hand, even when it hits, or being beaten by another hand, even when it hits.

in a vacuum, 3betting with AJ in these situations probably is best, but in a real game, 3betting with anything probably is best, so you're left with that choice of when to do it, and what to do it with. if i'm going to do it x% of the time, and its going to be roughly as effective nomatter what i do it with, then maybe i want to remove some hands from that range that can also show a profit by calling, in order to maximize my play as a whole.

[traz]

[ QUOTE ]

In short (too late) I'd want to three-bet rather than smooth-call to force my opponent to have a hand both preflop AND postflop, rather than smooth-call and only win when *I* have a strong hand. Showdown strength is nice, but folding equity is king in steal/resteal situations.

[/ QUOTE ]

Why are these the only 2 options? If you smoothcall but then put pressure on his hand postflop, isn't that kind of a combination? THAT'S the combo I like the most.

I advocate calling, but that's with the assumption that we don't cf every single flop we miss.

[traz]

also fwiw I don't think any option is significantly better than the other, and depending on your skillset there might be "best" choices for different players

[clowntable]

If we assume he only calls/plays back with QQ+/AK, that's 3*6=18 pairs and 16 AK combinations for a total of 34 combinations.
If we hold an A we remove 4 AK combinations and 3 AA combinations from his playback range. Instead of playing on with 34 hands he now plays on with only 27 hands.
That's > 20% less. And that's why we prefer having AJ over 72 for the 3 bet.

[DaycareInferno]

its also 20% less with an A2, which you will just be throwing away anyway. if you can call profitably with the AJ, then you can use the A2 instead of pitching it, while maintaining the same 3bet frequency.

[Leviathan101]

you also remove a jack from JJ reduces the chances further. playing back with AJ is better than 72 simply because you have an ace and a jack. and besides, when you cbet an ace high flop, you actually have to have an ace some of the time.

Pokey pretty much laid out the math on why to 3bet here, AJ may be practically no different from 72o postflop, but when it's proftiable to 3bet AJ why not do it?

[DaycareInferno]

[ QUOTE ]
you also remove a jack from JJ reduces the chances further. playing back with AJ is better than 72 simply because you have an ace and a jack. and besides, when you cbet an ace high flop, you actually have to have an ace some of the time.

Pokey pretty much laid out the math on why to 3bet here, AJ may be practically no different from 72o postflop, but when it's proftiable to 3bet AJ why not do it?

[/ QUOTE ]

because the combined profitability of AJ and xx is higher if you can profit from them both, as opposed to maximizing profit on the AJ and tossing xx.

[Chicago Twister]

[ QUOTE ]
I really don't like 3-betting with AJo. Your hand is good against his button opening range, yeah; but your hand is also [censored] against his 3-bet calling range. So you're turning your hand into a bluff, and bloating the pot if you do get called. I'd prefer to call preflop and c/r a lot of flops.

[/ QUOTE ]

Disagree. I prefer 3betting still because there is more to the hand than just his 3bet calling range. For one thing he's folding most of the time preflop trypically. Sure its a bluff, but its a bluff with no hand, so its not like youre turning AA in to a bluff. Him folding AT is not a bad thing here.

But also he is going to miss the flop 2/3 of the time, so when you follow through he still can't continue. So then you will often win wiith nothing, and sometimes get a better hand to fold (like 44 or AQ).

3betting is best, not because AJ is such a great hand, but because his range is so wide. Against a better opponent who knows what you're doing and will push back with calls & re-raises on the flop (sometimes also with nothing), 3betting is less viable. Against those opponents, I just fold pre.

[Unknown Soldier]

i havent read through, which could be quite ignorant. If you think thats the case I appologise but I'll give my opinion anyway.

3b AJ isnt great because he isn't calling with worse and it's strong enough to play. There's no point in turning it into a bluff because the hand has good enough value to call. I'd much rather 3b with A3o, and do so frequently.

Don't be afraid to make tough decisons. I have a feeling that all the arguments for 3b AJo will just come down to making it easier to play. As poker players we strive for the most +EV line, not the easiest. That's important.