AK is crappy multi-way all-in hand...
 

In turbos you end up in quite a few situations with 2 people all in ahead of you (or you call an all-in and someone else comes in behind you).

I'd always figured AK wasn't that bad in a 3-way pot. If I dominate one Ace and I'm a coinflip with a small pair, that's not so bad, right? well, I just ran the numbers and to my suprise, AK barely wins 1/3rd of the time. You might as well fold.

cards %win %lose %tie EV
3c 3h 45.04 54.40 0.56 0.452
As Kh 32.17 64.22 3.60 0.339
Qs Ah 19.18 77.21 3.60 0.209

Even against 2 smaller pairs, AK is only about 38% (which is certainly worth doing, but it's not as close to a coinflip as I would have thought).

Anyway, I figured if this was a suprise to me, it might be to others as well.

Edit: AKs gets up to around 38% as well.

Re: AK is crappy multi-way all-in hand...
 

38% is just over two to one to win, no? So with two all ins ahead of you, you get just the right pot odds vs these two hands. Only Kings and Aces have you smashed though, I think.

I think the question is though, do you want to take the ~1% (or whatever) edge with pot odds or wait till the maniacs bust each other out and attempt to outplay them later.

Re: AK is crappy multi-way all-in hand...
 

[ QUOTE ]

I think the question is though, do you want to take the ~1% (or whatever) edge with pot odds or wait till the maniacs bust each other out and attempt to outplay them later.

[/ QUOTE ]

big stack maniacs are hard to beat with few chips in a turbo tournament...sometimes ya just gotta cross your fingers =\

-eroc

Re: AK is crappy multi-way all-in hand...
 

Well, calling all-in for 1% seems poor. Shoving and getting called, you'll take it since you might have just stolen the pot.

If 2 people are all-in in front of me (and have me covered) then I can bet one probably has a pair and I should fold and wait for a better than 1% edge.

Re: AK is crappy multi-way all-in hand...
 

99.5% of players shouldn't pass up a 0.6% EV edge.

Try inputting a plausible range for your opponents 3-way, you start to get a better picture of how 3-way situations work.

Finding a specific 3-way scenario in where AK isn't a 1:1 in a situation where it only needs 2:1 on average is not going to make a plausible case.

Re: AK is crappy multi-way all-in hand...
 

Yes I'd have to call unless I knew one had AA or KK, for the reasons posted by eroc and Fortuna above.

Re: AK is crappy multi-way all-in hand...
 

[ QUOTE ]
Yes I'd have to call unless I knew one had AA or KK.

[/ QUOTE ]

You're not grasping the point of the numbers in the OP. ANY pair is bad for you in a 3-way all-in, not just KK/AA. You're crushed against AA/KK of course, but any pair puts you in a near break even spot. Better to wait and try for an all-in with an advantage.

Re: AK is crappy multi-way all-in hand...
 

Lots of good analysis -- let's take it one step further.

This is why it's hard these days in turbos to ring up an ROI of more than about 20% against non-idiotic opponents. (I miss the pre-Goodlatte days.) You don't have a ton of time to wait for perfect situations. You gotta take small edges like this, or end up too far behind to catch the big stacks without a miracle.

You'll make money pushing over the long haul. You just will make it more slowly and with more annoying bustouts than you'd like.

Re: AK is crappy multi-way all-in hand...
 

This thread makes me sad.

I hope that when I open it again tomorrow someone will have pointed out some of the obvious flaws in the above posts.

-Steve

Re: AK is crappy multi-way all-in hand...
 

[ QUOTE ]
99.5% of players shouldn't pass up a 0.6% EV edge.

Try inputting a plausible range for your opponents 3-way, you start to get a better picture of how 3-way situations work.

Finding a specific 3-way scenario in where AK isn't a 1:1 in a situation where it only needs 2:1 on average is not going to make a plausible case.

[/ QUOTE ]

Well, I put plausible ranges at 22+ and A2+ for starters. You throw in a few JQ type hands and it doesn't really get any better.

cards %win %lose %tie EV
Ad Kh 34.93 64.83 0.24 0.350
5c 5d 30.94 68.82 0.24 0.310
Qs Js 33.89 65.87 0.24 0.340

If you can put NEITHER player on a pair, it gets good. But one all in and one call? Awfully good chance someone has a pair.

AK is going to be a bit better overall as you might catch QJ shoving, called by AT, then you have AK. That's the dream scenario though. Most of the 3-way all-ins I see have at least one person with a pair.

Re: AK is crappy multi-way all-in hand...
 

Yup. Incremental edge * volume = profit. Variance's the price.

And that's something every tournament player must get used to if he or she has even a rational chance of becoming a successful player.

bogey,

You're defining a Hero with a skill edge over his opposition that rarely, if ever exists in tournaments with 300+ runners. Once you start trying to find better spots to double up with a bigger edge...

The bigger the field, the more you hurt your EV in the long term. And your edge would probably have to increase exponentially to overcome field sizes.

And practically anybody that thinks or assumes they do have that edge doesn't. Such a player doesn't exist.

Of course there are scenarios where you can fold AK, and no, you don't need to put your opponents on KK+ either to make your case. (FWIW, you're making a situational argument. Perhaps if you defined a specific situation, you'll make your point a little clearer?)

On pure maths, you have 33.9% EV, and when you examine 'stove calcs, you look at the EV metric, not the win % metric to assess what your equity is in the hand.

While it's true these days opponents have gotten better and this phenomenon has filtered down to the micros somewhat, there are still concepts and applications of those concepts they are clueless about and even solid, profitable players don't extract enough value from.

And that is superaggression and near-eidetic knowledge of ICM calcs in pushbotting phases of turbos.

Re: AK is crappy multi-way all-in hand...
 

[ QUOTE ]
38% is just over two to one to win, no? So with two all ins ahead of you, you get just the right pot odds vs these two hands. Only Kings and Aces have you smashed though, I think.

I think the question is though, do you want to take the ~1% (or whatever) edge with pot odds or wait till the maniacs bust each other out and attempt to outplay them later.

[/ QUOTE ]
Say you have 1000 chips and 2 allins in front of you, so you have the option of calling and if you win you get 3100, and your 38% to win. That gives 0.38x3100 = 1178, or a 18% increase in equity. 18>>>>1. You've got to be careful to not underrate your edge.
Also, comparing AK v QJ v AT or AK v AQ v 77 or etc isn't very helpful. Use pokerstove & ranges, for example:
equity win tie pots won pots tied
Hand 0: 26.776% 25.10% 01.68% 6493088 434363.67 { 55+, A2s+, K8s+, Q9s+, J9s+, T9s, 98s, A7o+, A5o, KTo+, QTo+, JTo }
Hand 1: 34.710% 30.98% 03.74% 8013844 966308.33 { 66+, A9s+, KJs+, QJs, ATo+ }
Hand 2: 38.514% 33.97% 04.54% 8788941 1175385.00 { AKo }

Re: AK is crappy multi-way all-in hand...
 

Pokerfarian makes an important point about not using hot and cold calculations against hand but rather using pokerstove ranges.

I think that it would benefit this analysis for someone to give a hypothetical example of a time where they might be calling all in with AK.

Will someone please post the stack sizes/blinds/antes from a point late in a turbo and create hypothetical action in front of them where they are the hero with AK. I think this will help show why the edge with AK is so great.

Re: AK is crappy multi-way all-in hand...
 

[ QUOTE ]
99.5% of players shouldn't fold AK PF....ever

[/ QUOTE ]

Minor FYP.

Re: AK is crappy multi-way all-in hand...
 

[ QUOTE ]
...there are still concepts and applications of those concepts they are clueless about and even solid, profitable players don't extract enough value from.

And that is superaggression and near-eidetic knowledge of ICM calcs in pushbotting phases of turbos.

[/ QUOTE ]

eidetic: marked by or involving extraordinarily accurate and vivid recall especially of visual images <an eidetic memory>

I had to look it up, so I thought I'd share!

Re: AK is crappy multi-way all-in hand...
 

[ QUOTE ]
[ QUOTE ]
99.5% of players shouldn't pass up a 0.6% EV edge.

Try inputting a plausible range for your opponents 3-way, you start to get a better picture of how 3-way situations work.

Finding a specific 3-way scenario in where AK isn't a 1:1 in a situation where it only needs 2:1 on average is not going to make a plausible case.

[/ QUOTE ]

Well, I put plausible ranges at 22+ and A2+ for starters. You throw in a few JQ type hands and it doesn't really get any better.

cards %win %lose %tie EV
Ad Kh 34.93 64.83 0.24 0.350
5c 5d 30.94 68.82 0.24 0.310
Qs Js 33.89 65.87 0.24 0.340

If you can put NEITHER player on a pair, it gets good. But one all in and one call? Awfully good chance someone has a pair.

AK is going to be a bit better overall as you might catch QJ shoving, called by AT, then you have AK. That's the dream scenario though. Most of the 3-way all-ins I see have at least one person with a pair.

[/ QUOTE ]

What if each of them has a pair?

Re: AK is crappy multi-way all-in hand...
 

PLAUSIBLE RANGES.

I.E.: First person jamming looser, second caller calling tighter, you have AK/AKo.

I'm running the pokerstove, but apparently it's quite a long calculation. I'll post it when it's done.

Re: AK is crappy multi-way all-in hand...
 

[ QUOTE ]
PLAUSIBLE RANGES.

I.E.: First person jamming looser, second caller calling tighter, you have AK/AKo.

I'm running the pokerstove, but apparently it's quite a long calculation. I'll post it when it's done.

[/ QUOTE ]

Just do Monte Carlo rather than all possible combinations.

Re: AK is crappy multi-way all-in hand...
 

What's the difference?

Re: AK is crappy multi-way all-in hand...
 

equity win tie pots won pots tied
Hand 0: 27.692% 26.34% 01.35% 143705872296 7375060046.00 { 22+, A4s+, K7s+, Q9s+, J9s+, T9s, A8o+, KTo+, QTo+, JTo }
Hand 1: 33.278% 30.56% 02.72% 166720018776 14836624904.00 { 55+, ATs+, KQs, ATo+, KQo }
Hand 2: 39.031% 35.64% 03.39% 194466677388 18477770654.00 { AKs, AKo }


---

So...yeah...AK is pretty damn good multi-way.

Re: AK is crappy multi-way all-in hand...
 

[ QUOTE ]
What's the difference?

[/ QUOTE ]

Not much, only with the monte carlo is just runs a whole bunch of simulations and records teh results. You can stop it whenever you want. For when every possible combination is too much.

Monte carlo is often more accurate in those situations becaus I think it runs the full combinations in a particular order. So if you stop the combination method after like 1 million hands, you aren't sure which ones you are getting. With the monte carlo, they are randomly selected and simulated.

At least I think.

Re: AK is crappy multi-way all-in hand...
 

OH GOD, FOR THE LOVE OF HUMANITY, STOP WITH THE AK THEORY FOLD POSTS!!! UGGGGGGGGGGGGGGGGGGGGGGGH. JUST...UGGGGGGH.

When I have AK, which hasn't happened a whole hell of a [censored] lot in the past month, I'm thinking: "how am I going to get my money in in good sized pots for FE preflop," or "how much am I going to RR," or "am I deep enough (read: never in online tournaments) to make a call here and play; or "WTF am I going to do against a 4-bet." I'm not thinking about how crappy this hand is multi-way and whether I'm going to make the umpteenth post this month about what folding AK because it may/may not have burnt me in a tournament last night. [censored].

Also, before posting your [censored] theory "Fold AK" posts (and I say this to help, please, I really do mean it, as violent as I sound, I do it with love), please read the 2+2 anthologies. Players who are better than you or I have posted on this subject; Woodguy made a nice post about small edges. If anything, read that post and try to counterargue it with a lot of thought; don't just post that you think AK is crappy in multi-way pots so you can get it in some person's head who is trying to improve and screw him up later on tonight.

Barry

Re: AK is crappy multi-way all-in hand...
 

i think you are forgetting the overlay of blinds and antes. plus the %of time ur against no pair is definitely about 15-20%. the % against kk aa is about 5%. its a positive ev play and a good way to amass alot of chips

Re: AK is crappy multi-way all-in hand...
 

[ QUOTE ]
don't just post that you think AK is crappy in multi-way pots so you can get it in some person's head who is trying to improve and screw him up later on tonight.

[/ QUOTE ]

I'm gonna fold AK preflop and send you the HH now, just to piss you off [img]/images/graemlins/grin.gif[/img]

Re: AK is crappy multi-way all-in hand...
 

[ QUOTE ]
[ QUOTE ]
don't just post that you think AK is crappy in multi-way pots so you can get it in some person's head who is trying to improve and screw him up later on tonight.

[/ QUOTE ]

I'm gonna fold AK preflop and send you the HH now, just to piss you off [img]/images/graemlins/grin.gif[/img]

[/ QUOTE ]


With the mood he's been in the last few days, you'll totally be in for a letter bomb at that point.

Re: AK is crappy multi-way all-in hand...
 

[ QUOTE ]
i think you are forgetting the overlay of blinds and antes.

[/ QUOTE ]

It should not have taken so long for someone to mention this.

The OP was talking about AK in the late stages of a turbo and in these cases the overlay is HUGE.

Also, in the late stages of a turbo everyone will have a low M so the first person's shoving range should be enormous and therefore the second person's isolation range should be very wide too.

AK usually has 38%-40% equity in a pot where they only need 25%-30% for a call to be correct.

This is an enormous edge.

When the OP was plugging in AK v 55 v AQ he was severely missing the points posted above and it is a shame that it took everyone so long to realize this.

Re: AK is crappy multi-way all-in hand...
 

[ QUOTE ]
i think you are forgetting the overlay of blinds and antes. plus the %of time ur against no pair is definitely about 15-20%. the % against kk aa is about 5%. its a positive ev play and a good way to amass alot of chips

[/ QUOTE ]

Yes, this is the point that people above you were missing.

Example: You have 10,000 chips and get into a 3 way allin with two people that have you covered. None of the 3 allin players is in the blinds. Blinds are 1000/2000 ante 100.

There will be 10000+10000+10000+3900 = 33900 chips in the main pot. You contributed 10000 of them, so you only need 29.5% equity for this to be profitable. If you actually have 36% equity you have a significant edge, especially this far into a turbo.

Re: AK is crappy multi-way all-in hand...
 

This thread comes just in time for me. My game has clearly improved the last few months since reading and beginning to understand how aggression and pushing on edges can help me become a better player.

Last night I was on a similar OP situation and... I pushed, it was just too good to let it go. So, if I made a donk move please tell.

$2.20/1545 players, 380 remains, ITM from 297. Big stacks are abusing the table and the game has turned a little wild. I limped bc I was pretty sure a raise was coming and wanted to rrai.

I'm including results this time, just to illustrate some possible ranges as the thread talks about that.

PokerStars No-Limit Hold'em Tourney, Big Blind is t200 (9 handed) internettexasholdem.com

CO (t3210)
Button (t4625)
SB (t8300)
BB (t2300)
Hero (t2570)
UTG+1 (t2187)
MP1 (t11910)
MP2 (t6720)
MP3 (t6150)

Preflop: Hero is UTG with K[img]/images/graemlins/spade.gif[/img], A[img]/images/graemlins/spade.gif[/img].
Hero calls t200, <font color="#666666">1 fold</font>, MP1 calls t200, <font color="#666666">2 folds</font>, <font color="#CC3333">CO raises to t3210</font>, <font color="#666666">1 fold</font>, SB calls t3110, <font color="#666666">1 fold</font>, Hero calls t2370 (All-In), MP1 folds.

Flop: (t9390) 5[img]/images/graemlins/club.gif[/img], 8[img]/images/graemlins/club.gif[/img], T[img]/images/graemlins/club.gif[/img] <font color="#0000FF">(3 players, 1 all-in)</font>

Turn: (t9390) 2[img]/images/graemlins/club.gif[/img] <font color="#0000FF">(3 players, 1 all-in)</font>

River: (t9390) 6[img]/images/graemlins/spade.gif[/img] <font color="#0000FF">(3 players, 1 all-in)</font>

Final Pot: t9390

Results in white below: <font color="#FFFFFF">
SB has 4c Ac (flush, ace high).
Hero has Ks As (high card, ace).
CO has Kc Ad (flush, king high).
Outcome: SB wins t9390. </font>