Thursday, July 26, 2007

Master of the Four Way all In

Since Marsh asked me, I felt a duty and an honor to post on his blog. You can always learn something about this game, and Marsh helped me with a couple tricks early in my poker career.

I thought I would comment on my wife’s play recently, as she has had not one but two four way all ins in her last 2 sessions of casino poker. In all the time I have played, both live and on line, I can’t recall a single four way all in, other than my wife’s. I have seen sets over sets, royal flushes, quads losing to a straight flush (on line), a famous one outer, but never a four way all in. Here is a description of the hands, and then some strategies if you would like to try to duplicate this feat.

Playing in Palm Desert on a loose table with some tourists, locals, and rich people, MB (my wife) has four callers in front of her and looks down at AQ os. Unconventional, I admit, but she min-raises the 5 way limped pot. The flop comes Kc, Jc, 10s. She checks, seat 3, a construction guy who plays aggressively bets $8 into a $20 pot, seat 5 raises to $28, seat 6 folds, seat 9 calls, and MB raises to $60 leaving $28 in chips behind her. Seat 3 looks at MB, says I bet you have AQ and then goes all in for about $75, seat 5 goes all in as well but is covered by seat 3, and seat 9 goes all in too for his last $10. MB then goes all, having everyone covered and capping the action. The next 2 cards blank, seat 3 now famous for not giving anyone odds to suck out, flips over K5 os, seat 5 flips over 9c10c, and seat 9 flips over 8,9 os. MB rakes the monster.

3 months later, playing in London at a $50 pound max table with 1 pound, 1 pound blinds, MB has 3 callers in front of her and looks down at AA in the big blind. Unconventional again, she min raises to 2 pounds. Everyone calls. Since I was not at the table, the replay may be a bit off but it went something like this. The flop comes Q, blank, blank with 2 clubs. MB checks, Seat 4 bets, Seat 6 calls, Seat 8 calls, and MB raises. Everyone calls. The next card is an Ace, not a club. MB goes all in, and seat 4, 6, and 8 all call. Players show AQ for 2 pair drawing dead, the 8 and 3 of clubs, and KJ (the name of the cardroom is the gutshot). The next card blanks and MB rakes another monster.

How did she do it:

• Maximizes value out of premium starting hands. Most all experienced players would have put in more substantial raises preflop. It is tough to min raise with big starting hands, but you are still the most favored hand with AA vs. any other possibilities of starting hands. So even if you are in a four way pot, you may have less than a 50% chance of winning the pot, but will have a greater likelihood of winning against any other starting hand. Min raising with AA in a 4 handed pot is shocking to most players, so few will put you on that hand.
• Look like you don’t know what you are doing when you bet. MB is a very experienced player but she does have an unsure look about her when she bets. Players often think she is a donk and call her down, then get surprised when they lose all of their chips. If you don’t think you can master the donk look when you bet, then make a donk play on purpose for a small amount of chips early in a multi hour session, it may pay off for you beautifully later.
• Play at tables without deep stacks. MB bought into tables where the maximum buyin was 50 times the blind. Obviously, this creates more desperation plays and more donkish behavior.
• Don’t be afraid to be out of position in a multi way pot. Some may disagree here, but I don’t find as much of a disadvantage to out of position play with multiple players involved. You are first to act, but by checking, someone will typically bet top pair, middle pair, or a draw if it is a multiway pot, and then you can make your decision being last to act. Flopping the nuts makes these decisions much easier.

I realize you will have some losses on big pots with this strategy, but if you can win a multi-way pot for every pot you lose, I think the strategy can work and create a positive EV.

11 comments:

Sushi Cowboy said...

(with an "I can't believe Jason actually said that" shaking of the head in disbelief) ummmmm...got it. Thanks.

Ryan said...

Welcome, Jason! Thanks for contributing to the blog, that's awesome! Now I'm going to tear you apart! (See Marsh’s post on giving and receiving criticism for my constitutional right to rip you a new one.)

Let's take it from your "How she did it" section, since everything before that is basically hand recap stuff. How she did it, according to Jason:

"Maximizes value out of premium starting hands."

Misplaying a good hand preflop, flopping the nuts, and having them hold up against a ton of draws, is not "maximizing value." Misplaying the best starting hand possible preflop, turning the nuts, and having them hold up against a draw is not "maximizing value." It's getting lucky.

"Most all experienced players would have put in more substantial raises preflop. It is tough to min raise with big starting hands, but you are still the most favored hand with AA vs. any other possibilities of starting hands. So even if you are in a four way pot, you may have less than a 50% chance of winning the pot, but will have a greater likelihood of winning against any other starting hand."

The irony that you are a banker continues to deepen, Jason. You are now more likely to lose this hand than to win it--with a hand that you probably won't be able to get away from when you are beat--and the fact that you are a favorite over any individual hand is utterly meaningless.

"Min raising with AA in a 4 handed pot is shocking to most players, so few will put you on that hand."

You will never get me to believe that min-raising with AA in the BB with three callers is *ever* correct. It's true, guys...if I min-raise from the BB in that scenario, you can take AA right out of my range, because it will never happen. "They'll never put you on it" is a terible reason to severely cripple your EV with the best starting hand in the game. Yes, you have to mix up your play in hold'em, but there are ample ways to mix up your play while adhering to a solid base game, which includes never making this play.

"Look like you don’t know what you are doing when you bet."

Generating an image at a new table when you intend to play for an extended session is a good idea. There are many images you can generate, and there are no bad images, only images one fails to capitalize on.

"Play at tables without deep stacks. MB bought into tables where the maximum buyin was 50 times the blind. Obviously, this creates more desperation plays and more donkish behavior."

If people are more desperate with their stacks, then it makes the lightly-played nature of these hands even worse. You should put in the big raise and hope someone makes the desperation call or moves in. At a table you call "desperate," you should push reasonably hard with your premium hands, not slowplay them.

"Don’t be afraid to be out of position in a multi way pot."

In general, the whole problem with being out of position is thatyou don't know if it's going to be a multi-way pot! The rest of the table hasn't acted yet! In the case of aces from the BB, I would agree that once you are in a multi-way pot, being out of position is not as big a drawback as it is heads up, but you are almost completely discounting it. Being out of position still sucks, just not as badly as it sucks heads up.

The problem is, you are out of position with a hand you can't easily get away from! Position isn't hugely important when you are playing a small suited connector in a mult-way, because you are going to flop a monster, a draw to a monster, or you are going to go away. How to play the flop becomes an easy decision from any position.

"Flopping the nuts makes these decisions much easier."

We agree on this point. Flopping the nuts is excellent poker strategy.

"I realize you will have some losses on big pots with this strategy, but if you can win a multi-way pot for every pot you lose, I think the strategy can work and create a positive EV."

That's a big "if," and even if it comes to pass, +EV doesn't mean maximum EV.

I appreciated the comments about generating an image at a new table when you intend to play for an extended session, but I will never seek to generate a multiway pot with a premium hand out of position via weak, "sweetener" raises (or, in your AA vs. AA story from Wednesday night, smooth-calling a reraise with AA with multiple players to act in between you and the reraiser). It's just bad poker, not clever poker.

Watch all three seasons of High Stakes Poker, and find me one example of a pro playing a premium hand this way. You will never find it, because, despite being extremely creative, tricky, deceptive players, they would never, under any circumstance, cripple their EV with a premium pair like that.

I warned Jason in an email exchange that I was going to post a critical response, and his reply, besides saying “go for it,” included the statement: "Maybe it's just me, but I have never been happy with many players response, 'I raised preflop, took down the pot and did not get my Aces cracked.'"

You speak as though the only possible result of actually raising with your aces is that everyone will fold and you will take down the pot right there. Were people folding a lot to raises preflop? Was there actual blind stealing happening at that table? You describe the table as being full of bad players, so why would you think that a standard raise is going to chase away all your customers?

The good news for MB is that she still didn't play either of these hands as badly as you played AA UTG in pot limit by raising the pot, getting a bunch of callers, having the button reraise the pot, and you smooth calling (ultimately getting sucked out on by one of the subsequent callers, I might add). I can't think of a hand that I've witnessed or heard of where aces were played more dismally than that. Just. Awful.

Again, welcome to the blog, Jason!

jsola said...

I agree with most of what Ryan said, but this part really sticks out as the crux of why you don't have to play aces like this:

"You speak as though the only possible result of actually raising with your aces is that everyone will fold and you will take down the pot right there. Were people folding a lot to raises preflop? Was there actual blind stealing happening at that table? You describe the table as being full of bad players, so why would you think that a standard raise is going to chase away all your customers?"

I've played at a lot of d0nk-heavy tables, and time after time they have shown me that they do not like to fold. It doesn't matter if you sat down 6 hours ago and haven't played a single hand, they're too dumb to notice. If you have pocket aces, and 5 people have limped ahead of you, raise the shit out of it, and hammer like 90% of the flops that you see. People will call you down for a whole stack and show you top pair no kicker and you'll win and say "what the fuck" because they're just that bad.

In the hand examples you gave, I think preflop is atrocious, but post flop is just fine. She managed to get her money in as quickly as possible with the best hand, although the check raise is a bit risky for my liking.

As far as minraising and then flopping the nuts in a multiway pot, it's just not going to happen often enough to make up for the loss of preflop EV by not isolating out a single opponent. Technically you could use the strategy you described with any two cards (flopping two pair with 64o is just as good as two pair with AQ, and it's less likely to have hit someone else!), so why give up all that juicy preflop value?

Austin said...

Today during the lunch game I managed to get AA and decided to play it totally differently than I've ever played it before to try out this slow playing the nuts idea. I was about mid-table and limped in by calling the blinds since there were only 3 or so other people to go and I was the first to call. Table ends up with four people and the flop was 2, 8, 10. I figure, sweet I'm way ahead of most people. Ends up that Martin had 8,10, cracks my aces and I lose the majority of my stack.

Granted, things could have been different and he could have only hit top pair instead of two pair, but why take the chance? It would have been much better to raise it up and perhaps get one caller (and get Martin the hell out of there with his 8,10) than have what happened happen. I'd much rather have a heads up where you win a smaller pot than getting lucky with a monster pot with several people in it, because really, when that many people are calling, your chances of winning go WAY down even though you do have the best hand before the flop. Once that flop comes, the entire game could change and it certainly did in my case.

I can remember another time when I had aces a week or two ago UTG and raised it up to 7 having everyone else fold after that and me taking down the blinds. I'd MUCH rather have that happen 100% of the time than potentially getting the AA cracked because I was trying to be sneaky and get lucky. Suffice it to say, I'll never slow play the AA again and experienced first hand why you really don't want to do it either.

jtrey333 said...

I now have both ends of my risk/reward scale filled in... on one end, we have weak/tight Martindows 3.1 and on the other end, we have the Jason/MB scale. Everyone falls in between.

Ryan said...

The ironic thing is, both models play hands the same way, but one model calls it being cautious and the other one calls it slowplaying...

Sushi Cowboy said...

And just to be clear, raising pre-flop with Aces will not necessarily get me off of my 8T (especially if it is suited) but it will help to define the hands you are up against. If you raise big and get one caller, then the flop comes out KQJ, there can easily be a set out against you. Pre-flop raise won't help to make your AAs bulletproof but it will narrow the range of hands out there so you can make a better determination if you are beat or not.

Ryan said...

Or, let's say it's 1-2 game, and a raise to 8 will *never* get Martin off T8o in the BB, he'll call 6 more 100% of the time (not far from the truth).

Let's further say that Martin is not going to continue on that flop unless he hits top pair or an open-ender, and will check/fold otherwise.

When you bet him off his missed flop (the vast majority of the time), if you've raised him to 8 preflop, you are getting 6 "extra" chips off him. When you have limped, you get no extra chips off him.

Which scenario when you bet him off a missed flop do you think best helps pay for the times where he hits T8x, you end up all in, and it holds up?

The lesson here is that making that preflop raise is not simply about chasing away a bad hand versus letting them see a flop, it's about making a bad hand *pay* to try and crack you, such that when they finally do, you've already paid for that beat with their money, 6 chips at a time...

jason said...

A couple points of clarification. You would have to run some mathematical models (with lots of educated guesswork) to figure out if, over time, the value of winning multi way pots vs. the losses associated with playing big cards that you can't get away from.

In the case of the hand where MB turned trip Aces, given the flop and turn, MB's play was clearly much more profitable than had she put in a big raise preflop. Her odds of winning are greater than 50% against the entire table post flop, and greater than 75% post turn.

If she raises big preflop, AQ likely calls, KJ likely folds, and 8,3 of clubs likely folds. The flop comes Q, blank blank with 2 clubs, and MB likely doubles up.

Given the flop of Q, blank blank with 2 clubs, the entire outs of the table are 2 for the person with AQ, with redraw outs for AA should a Q hit. 9 for the person with 2 clubs, with virtually no redraw outs for the AA, and almost none for KJ, except for runner runner straight. So at this point MB is still about a 56% favorite to win, computer models will give you a better number but this is close.

On the turn with the Ace, AQ is drawing dead, KJ has 4 outs, and 8, 3 of clubs has about 7 outs, depending on how many pairs are available for the full house. MB is now about a 78% favorite to win vs the entire table. I would be happy to take these chances to win a 4 way pot. You need to be prepared to lay down aces when the flop, turn and betting patterns don't cooperate. MB is very good at making laydowns, which is why this strategy seems to work for her.

Yes, she was out of position in this hand but she was pretty much assured, given the min raise, that all players would call. You are correct, heads up out of position sucks but it is not as big a deal in a multiway pot.

I must have not told the story right about the ACE ACE vs. ACE ACE vs. 7,8 os. I raise in early position to the pot, calling station calls, small blind reraises, I smooth call to induce calling station to call. If I raise, I basically have to go all in ( we are all playing with about 100-120 pounds and the preflop raise is now 36 pounds). I figure even calling station will fold to a raise of 72-100 pounds preflop when he is only in for my intial raise of 7 pounds I believe. Plus, I don't know that small blind raiser has AA, he may have QQ,KK, or AK. In fact, it is highly unlikely that he has AA. The flop comes 9,8 blank, AA goes all in, I call, and calling station calls. Small blind raiser then asks me " Do you have a big pair", I say "yes do you have a set of 9's" he says "No but you are behind." The identical hands of preflop raiser and myself hold up and ACE ACE vs. ACE ACE split the pot.

Sushi Cowboy said...

So in the London AA vs AA hand are you saying that it is better to lose your whole stack instead of splitting pre-flop action with the other AA?

And your description of MB's AA in that particular hand may be accurate but you need to consider all of the other possible flops like when someone flops two pair or trips and MB doesn't catch a two outer. There will probably be much more variance when playing AA versus three other hands. Overall though, I think you will end up with less EV multi-handed because you will not have defined your opponents' hands well enough and you'll have no idea if you are ahead of TPTK or way behind two pair so you'll end up calling down a lot of hands you shouldn't.

Still, I can't claim to have two different four way all-ins so I encourage MB (or anyone for that matter) to min-raise with AA as much as they want.

Ryan said...

"You would have to run some mathematical models (with lots of educated guesswork) to figure out, over time, the value of winning multi way pots vs. the losses associated with playing big cards that you can't get away from.

In the case of the hand where MB turned trip Aces, given the flop and turn, MB's play was clearly much more profitable than had she put in a big raise preflop."


The results of this one hand are not particularly relavent to the question of whether the approach as a whole has long-term merit. Yes, it’s a "complicated mathematical model with lots of guesswork" to prove scientifically that playing AA this way is a long-term mistake, but screw that. I will stick with my “High Stakes Poker” point: show me a pro who would play AA that way preflop under any reasonable circumstance. You won’t find one.

Now, you can suggest that you have stumbled upon an approach to maximizing your return on premium hands that has somehow eluded the collective minds of the greatest players the game has ever seen, but I’m going to stick with the premise that they are right: it is flat-out incorrect in the long term to play AA preflop this way, because the best players in the world never do it.

If you need a complicated mathematical model to be convinced of this, well, I'm not going to produce one. So, by all means, keep minraising preflop monsters in a deliberate effort to build large, multiway pots, then justify it as a high-risk, high-gain style choice. I encourage that mentality in all my opponents...

As for the London AA vs. AA hand...I apologize if I butchered the details, but regardless, a raise was called for. The reraise was the perfect result for you, opening a door that you then failed to walk through.