ICM Thoughts

[Unarmed]

DISCLAIMER: I'm not an ICM nerd.

I'm going to be talking about the ICM EV of an exact situation, static EV if you will. I'm not interested in talking about true overarching EV, the type that includes stuff like how pushing and winning makes me feel happy, which makes me play better, so I should include that impact in my calculations.

Anyway...

When calculating the ICM EV of a push, there are two value components:
1) The percentage of times everyone folds
2) The percentage of times you are called and win

I think we're all sort of aware of one side of this. Like, if I have KTs in the SB and 14BB, its a clear push by EV. However, I normally pass on it because I don't want to risk busting at that stage. Put another way, a ton of that situations EV is coming from 2) and with a decent stack, I don't really like to put myself at risk of busting.

I'm more interested in the other side of the equation. Say I'm a shortish stack 6 or 7 handed. Now I don't really mind busting, I'm way more interested in doubling up off someone. Therefore, when faced with an equal EV situation, say 74o in the SB, and KQs UTG (just assume you have calling ranges locked and the EV of the two is equal), I should NOT be indifferent between the two. I should be happier to take the KQs because a greater % of its EV comes from its ability to double through someone. I *think* this can be expanded to say that I'd rather take the KQs with 0.2% EV over the 74o and 0.3% EV, and if that's true perhaps I should be willing to take KQs and slight -EV for the chance to double through?

I guess what I'm trying to say is that there are two components of the ICM EV numbers we get when analyzing pushes, and that we should be aware of that, at the very least. Also, it is intirely possible that this post makes zero sense. /images/graemlins/grin.gif

are you saying that its right to pass up on these overlooked +ev situations, or that we need to take advantage of them?

[axeshigh]

I don't play by ICM exactly either. I try to identify the situations in which I need to double up, and those in which it would be more profitable just to steal the blinds, and try to play accordingly. Good post.

I am not an ICM nerd either. I think you're right given the situation you explain. When it's 6 or 7 handed doubling up gains you more equity than when the table is 4 handed. At the same time, if you can be *certain* about everyone elses calling range and you still have a negative EV than you should fold because ICM takes into account the "handedness" (to use Eastbay's term) of the table.

On the other hand, I am not sure that SNGPT takes into account your position relative to the blinds. So, using your example, it might be correct to bush with KQs under the gun even if it's negative EV because the model MIGHT (i don't know for sure) not take into account that you'll be posting the BB the next hand.

[eastbay]

[ QUOTE ]
Therefore, when faced with an equal EV situation, say 74o in the SB, and KQs UTG (just assume you have calling ranges locked and the EV of the two is equal), I should NOT be indifferent between the two. I should be happier to take the KQs because a greater % of its EV comes from its ability to double through someone.


[/ QUOTE ]

Huh? Explain please.

eastbay

[Apathy]

[ QUOTE ]
[ QUOTE ]
Therefore, when faced with an equal EV situation, say 74o in the SB, and KQs UTG (just assume you have calling ranges locked and the EV of the two is equal), I should NOT be indifferent between the two. I should be happier to take the KQs because a greater % of its EV comes from its ability to double through someone.


[/ QUOTE ]

Huh? Explain please.

eastbay

[/ QUOTE ]

At this stage of the tournament he is not looking to steal the blinds and maintain his stack he's looking for the best gamble to double it up, so he prefers a hand with greater showdown value at the expense of a loss of folding equity.

I'll give my own thoughts later.

[pergesu]

I think what he's saying is that when you have a very small stack, pushing and picking up 1.5 BB isn't going to be incredibly beneficial to you. You would much prefer to double up than pick up the blinds.

The idea then is that when you push from the SB with a bad hand, most of your EV comes from the likelihood that your opponent will fold. When you push from UTG, most of your EV comes from your ability to double up, which would give you a better chance to win the thing and steal more blinds.

Here's my issue with this. Unarmed calls it "static EV" or something like that, which I interpretted to mean the value of that single hand. If that's the case, then I think this makes no sense, because a +.3% edge is better than a +.2% one. If you look at it in terms of the rest of the tourney though, he may be on to something because the times you double up you're able to create many more +EV situations, or at least wait for some more.

[eastbay]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
Therefore, when faced with an equal EV situation, say 74o in the SB, and KQs UTG (just assume you have calling ranges locked and the EV of the two is equal), I should NOT be indifferent between the two. I should be happier to take the KQs because a greater % of its EV comes from its ability to double through someone.


[/ QUOTE ]

Huh? Explain please.

eastbay

[/ QUOTE ]

At this stage of the tournament he is not looking to steal the blinds and maintain his stack he's looking for the best gamble to double it up, so he prefers a hand with greater showdown value at the expense of a loss of folding equity.

I'll give my own thoughts later.

[/ QUOTE ]

The only way I can make any sense of this is to assume that the implication here is really that short stacks are overvalued by ICM (therefore the so-called "same EV" isn't really). I think that's a debatable question, but I think this is a strange way to frame that question.

Or maybe I'm not getting the point yet, it's hard to tell.

eastbay

[stupidsucker]

I think maybe he means

a .5% gained equity for hand A with an X% busting possibility is differnt then a .5% gained equity when the % chance of busting is higher...You might be gaining the same value in ICM theory, but because your risk is greater(busting more often) your actual value is higher to fold.

for example..lets use a ridiculous hypothetical. Lets say there is a lotto. you pick 1 number between 1-1,000,000 and if your number is picked you win $1.5 million. If every entry costs $1, but you can only play once a day it just isnt worth it. in the short run without being lucky. In poker you can pass on high bust + EV situations and pick up another +EV situation with lower variance.

Maybe I am waaaaaay off.

[eastbay]

[ QUOTE ]
I think maybe he means

a .5% gained equity for hand A with an X% busting possibility is differnt then a .5% gained equity when the % chance of busting is higher...You might be gaining the same value in ICM theory, but because your risk is greater(busting more often) your actual value is higher to fold.


[/ QUOTE ]

He seems to be saying the opposite - that you should prefer the higher risk play.

eastbay

[ewing55]

Are you saying you would rather take the KQ hand because it is riskier? Meaning you are much more likely to get called and when you do you either double up or go home. Whereas the 74o will most likely just add slightly to your stack?

I made a similiar post a while ago about how I'd rather push my really small stack with a decent hand against a bunch of caller instead of push my really small stack against 1 caller with AA. That way I either multiply my stack or go home, because doubling up doesn't really help much with a really small stack.

The best reply was to the effect of: "You don't get these choices. You get dealt what you get dealt and you have to decide whether or not to play the hand/situation you are dealt or not."

In theory it is nice to argue these points and it may be helpful in thinking about poker in general, but in the game your only decision is whether/how to play the hand you are dealt.

Of course I could be wrong and probably am. /images/graemlins/laugh.gif

---------------Jeff

[Unarmed]

[ QUOTE ]
a .5% gained equity for hand A with an X% busting possibility is differnt then a .5% gained equity when the % chance of busting is higher...You might be gaining the same value in ICM theory, but because your risk is greater(busting more often) your actual value is higher to fold.


[/ QUOTE ]

This is pretty much what I'm saying. I have two situations that net me 0.3% static ICM EV (like I put them in SNGPT and they both come out 0.3%) Sometimes I'll prefer to take the situation where I'm getting more of my EV through the strength of my cards, sometimes I'll rather the situation where I'm getting more through FE.

There may be little to know practical significance of this, and I want to go surfing so I'm in no mood to search for it. /images/graemlins/grin.gif

[eastbay]

[ QUOTE ]
[ QUOTE ]
a .5% gained equity for hand A with an X% busting possibility is differnt then a .5% gained equity when the % chance of busting is higher...You might be gaining the same value in ICM theory, but because your risk is greater(busting more often) your actual value is higher to fold.


[/ QUOTE ]

This is pretty much what I'm saying. I have two situations that net me 0.3% static ICM EV (like I put them in SNGPT and they both come out 0.3%) Sometimes I'll prefer to take the situation where I'm getting more of my EV through the strength of my cards, sometimes I'll rather the situation where I'm getting more through FE.

There may be little to know practical significance of this, and I want to go surfing so I'm in no mood to search for it. /images/graemlins/grin.gif

[/ QUOTE ]

You seem to come to the opposite conclusion, though. You're preferring the higher risk play for the same change in EV.

eastbay

[Unarmed]

Well, by saying that I sometimes want the option which gives me more EV through the strength of my cards, that implies that I wish to take the high risk option. I'd accept the higher risk if my placing in the tournament is such that I feel I need to take a gamble in order to make my situation profitable. Is that mental?

[llabb]

No, this is not mental, this is smart poker depending on the play-style of the table. In lower buy-ins, it's probably preferable to take the 74 push rather than the KQ, because you simply want to survive. The bubble will be soft, and if you survive until it is shorter-handed, you will have more +EV opportunities, you can successfully steal more on the bubble, and perhaps fold/steal your way into a bigger stack and ITM without a showdown.

In the mid-to-high buy-in levels, the bubble is better, and you will have fewer and less +EV pushing situations. It is more difficult to make it ITM as a short stack, since you have less FE and fewer stealing opportunities. Unarmed is saying that he feels as a shortie he needs to double-up in order to make it ITM.

Because this is his assessment of the overall tournament situation (which I agree with), it is +"tournament EV" to prefer the KQ over the 74. They may have equal "hand EV" for the chips results at the end of that hand and that hand only, but one situation is clearly preferable over the other in the tournament as a whole.

Eastbay, I think this means you are correct, that this whole post is a conclusion derived from the assumption that short-stacks are overvalued by ICM. The way tournaments play, I personally believe this is true for the mid-to-higher level buy-ins, but not for the lower-level buy-ins, for the reasons stated above.

At the $55's, where I currently play, I agree with Unarmed, I prefer the KQ UTG push over the 74 SB push.

[llabb]

One other thing, you were wondering about the practical application of this. I believe this means at the mid-to-high buy-ins, you should not pass up these UTG opportunities. Such as your KQ example, or Degen's A9 example a few days ago. A good reason for this is that the 74 example is less likely to come about, since it is less likely to be folded around to you.

However, you should in fact pass on these opportunities at the lower buy-ins, from what I have read by other posters. I think Lorinda has said to pass on these UTG marginal opportunities, in part because there is a greater edge later in the lower buy-ins. It is more likely to be folded around to you, and there are greater +EV opportunities closer to the bubble, whereas there are fewer ar the higher buy-ins. This is perhaps why Lorinda can successfully be tighter than other 2+2'ers at different buy-in levels, who need to be more liberal with their pushing ranges and marginal +EV opportunities.

First off, good post. I think discussions about misinterpreting/overlooking mathematical/ICM/PT hands should be implemented far more than they have been. But Unarmed, i think you are overestimating your "doubling up" equity.

For instance, KQo vs A2o=74o vs A2. Taking that into consideration, I think you are wrong. But, I would like some further discussion about this. I am still unsure.

Suge

[axeshigh]

[ QUOTE ]


For instance, KQo vs A2o=74o vs A2. Taking that into consideration, I think you are wrong. But, I would like some further discussion about this. I am still unsure.

Suge

[/ QUOTE ]

Well, sometimes he calls with TT or 77 and you look really stupid with 74o. (yes, obvious, but what is your point anyway?)

[WebGuySteve]

DISCLAIMER: Not only am I not an ICM nerd, I really don't even know the math behind it /images/graemlins/wink.gif

Let me give my shot at putting this into my own words:

He is saying that when there are many people left with large blinds and everyone has about the same amount of chips, it is FAR better to gain a large stack than take the 300 chips in blinds. If he is pushing KQ from UTG he is very likely to be called, however, he has a decent chance of winning. This is where, say, .2% EV comes into play. If you push 74o from the SB, sure, it's +EV, say, .3%, however, this EV comes from the fact that he's folding a lot. You're going to take both of them, however, you prefer the .2% of the KQ because a large stack at this point in the SnG gives you a much better opportunity to put the pressure on the smaller stacks.

Also, when there are 6-7 people left, stealing the blinds isn't going to really give you a better chance of making it into the money than if there are 4 people left.

By the way, I agree with this completely, and it is how much of my thought process is done when playing.

[quote it is FAR better to gain a large stack than take the 300 chips in blinds. If he is pushing KQ from UTG he is very likely to be called, however, he has a decent chance of winning.

[/ QUOTE ]

why does he have a "decent chance of winning" if he is called...as opposed to 74o vs a calling hand...like i said before, 74o=KQo vs A2...And i am using A2 bc i consider that the worst hand he would call with.

[ QUOTE ]
[ QUOTE ]
it is FAR better to gain a large stack than take the 300 chips in blinds. If he is pushing KQ from UTG he is very likely to be called, however, he has a decent chance of winning.

[/ QUOTE ]

why does he have a "decent chance of winning" if he is called...as opposed to 74o vs a calling hand...like i said before, 74o=KQo vs A2...And i am using A2 bc i consider that the worst hand he would call with.

[/ QUOTE ]

How about 44-JJ, Qx, Kx?

And maybe Unarmed is right, but the first thing that came to my mind was: You can't choose your cards!

When you have 5BB in the SB with a hand like 74o and it's +EV to push, you should definately do it, because you are not guaranteed to get another +EV situation this orbit, and if you go through the blinds you'll only have 3.5 BB's. I think stealing the blinds with a short stack does make a difference, for this reason.

[adanthar]

KQs is sort of a bad example. A better one would be pushing something like 87s vs. calling an all in with 44 while shortstacked. You can easily come up with hypoes in which pushing 87s UTG 5 handed and calling 44 from the BB 5 handed would both be +.3% EV, but the second one would be preferable IRL because of the Gigablock idea where doubling up gives you a bigger stack which gives you FE, lets you control the table more, whatever. ICM chronically underestimates the value of a big stack so it makes sense, anyway.