Level 5 pocket 10's / My first ICM analysis [Archive]

View Full Version : Level 5 pocket 10's / My first ICM analysis

michw

05-12-2005, 04:20 PM

Here's a hand I played yesterday that I'm now questioning. Pushing w/ 10's is automatic, but do I really want to call a big bet w/ 10's in this situation? I also think it would be an automatic call if I had say 800-1000 chips, but with 1600 chips maybe it's not. I also wonder if I would even be questioning the play had I won. I have no particular read on the pusher and it's a Party 30+3 tourney.

Party Poker No-Limit Hold'em Tourney, Big Blind is t200 (5 handed) converter (http://www.selachian.com/tools/bisonconverter/hhconverter.cgi)

BB (t2015)
UTG (t1800)
MP (t1435)
Hero (t1660)
SB (t1090)

Preflop: Hero is Button with T/images/graemlins/heart.gif, T/images/graemlins/spade.gif.
1 fold, MP in all in t1435, Hero calls t1435, 2 folds.

Flop: (t1735) J/images/graemlins/club.gif, K/images/graemlins/spade.gif, Q/images/graemlins/spade.gif (2 players)

Turn: (t1735) 4/images/graemlins/club.gif (2 players)

River: (t1735) Q/images/graemlins/heart.gif (2 players)

Final Pot: t1735

Results in white below: 
MP has Ah Kd (two pair, kings and queens).
Hero has Th Ts (two pair, queens and tens).
Outcome: MP wins t3170. 

Yesterday I spent some time reading about ICM calcuations and I think I now somewhat understand it's limitations and usefulness. Based on what I read, I believe an ICM analysis could properly be used on this hand to determine whether I made the correct call from a $EV perspective. I was tempted to purchase Eastbays tool (probably still will), but would like to be able to perform some rudimentary calculations on my own first. Below I attempt to do this while using Dethgrinds tutorial as my template.

The first problem I see is that my action does not conclude the hand. I'll have to assume the SB and BB won't wake up w/ pocket aces or anything like that.

1) If I fold I have 20.8% of the prize pool - 1660/8000
2) If I call and lose I have 2.8% of the prize pool - 225/8000
3) If I call and win I have 42.4% of the prize pool - 3395/8000
4) If I call and split I have 22.6% of the prize pool - 1810/8000

Because I have no read on the pusher I'll use the 'average' push range from Eastbays tool - 77+ and A9+ and I also threw in KQ. According to pokerstove...

I win 55%
I lose 44%
I split 1%

Calling is then worth .028*.44 + .424*.55 + .226*.01 = 24.8% of the prize pool. And since folding is worth 20.8% of the prize pool would you agree I made the correct decision?

Questions for the group
1) Is this proper usage of ICM?
2) Did I perform the calculations correctly and is this basically what Eastbays tool does only much faster and cleaner?
3) Based on this analysis would you agree that calling with the 10's was the correct decision?

Thanks in advance for any responses this post may generate. This is my first attempt at using ICM for determining whether I made a good decision and I want to be sure I'm not talking out of my ass here.

UMTerp

05-12-2005, 04:31 PM

[ QUOTE ]
1) If I fold I have 20.8% of the prize pool - 1660/8000
2) If I call and lose I have 2.8% of the prize pool - 225/8000
3) If I call and win I have 42.4% of the prize pool - 3395/8000
4) If I call and split I have 22.6% of the prize pool - 1810/8000

[/ QUOTE ]

You're close here, but the section I copied is wrong. You need to use the ICM calculator to convert the chip stacks into prize pool equity: http://www.bol.ucla.edu/~sharnett/ICM/ICM.html

It's easy to see the flaw in your calculations - what if somebody had 5000 of the 8000 chips? Would he have 62.5% of the prize pool equity then? Of course not, since first prize is only 50%.

I have a feeling that once you run the correct numbers, they'll be a lot closer than 24% to 20%, but calling may still barely be correct. I'd also loosen up the pushers range quite a bit.

BTW, it's commendable that you're making an effort to really understand what the ICM calculations do and how to do them longhand. Not many new players here have the discipline to do that, or care enough to.

KJ o

05-12-2005, 04:32 PM

What you have done is roughly correct (see above), and Eastbay's tool agrees with you that you should call.

His tool assumes that the blinds will fold 100% of the time. In reality, they only fold 97% of the time (or whatever), but when they do call, you are in bad shape, usually, so it's possible that some small value should be removed from the calculations.

My version of the tool has Average as 44+,A7s+,A9o+,KJs+. Can't remember if I changed that or if that's the default. Anyway, the tool lists a call with TT as +1.0%.

A call with 99 is -0.5%, so that's the worst PP where it is still correct to call.

michw

05-12-2005, 05:05 PM

[ QUOTE ]
It's easy to see the flaw in your calculations - what if somebody had 5000 of the 8000 chips? Would he have 62.5% of the prize pool equity then? Of course not, since first prize is only 50%.

[/ QUOTE ]

Oops, you're correct and it does make intuitive sense. After reviewing Dethgrinds tutorial one more time I see I missed this step. I'm now going to have to figure out how this conversion is performed but recrunching the numbers yields:

1) If I fold I have 20.8% of the prize pool
2) If I call and lose I have 3.8% of the prize pool
3) If I call and win I have 34.3% of the prize pool
4) If I call and split I have 22.1% of the prize pool

Calling is now worth .038*.44 + .343*.55 + .221*.01 = 20.8%!!! Exactly the same as folding. In that case, assuming some type of skill advantage and that the blinds have yet to act I think it's a fold and I made a mistake by calling. Maybe this wasn't entirely results oriented thinking on my part after losing the hand.

UMTerp

05-12-2005, 05:13 PM

Yeah, you got it now. In reality, I think the pusher's range is probably a little looser than you originally gave him credit for, and your call was correct, albeit barely. It's a lot closer than one would think though, and this exercise drives home the point of why it's such an advantage to be the aggressor when the blinds get big. Of course, at the lower levels, most players don't know that it's "right" to fold 66 or 77 here, so that needs to be taken into account as well.

[ QUOTE ]
I'm now going to have to figure out how this conversion is performed

[/ QUOTE ]

That's not something you'd want to do longhand, except if it's headsup, when the ICM is pretty much rendered useless. I'm pretty sure it'd be some sort of higher-order equation you'd need to solve.

michw

05-12-2005, 05:13 PM

[ QUOTE ]
My version of the tool has Average as 44+,A7s+,A9o+,KJs+. Can't remember if I changed that or if that's the default. Anyway, the tool lists a call with TT as +1.0%

[/ QUOTE ]

I had a flaw in my initial calculations. After recrunching the numbers it showed folding to be about equal to calling. However, if the villians push requirements are relaxed, TT figures to be a favorite against most of these hands, probably making calling to be slightly correct as Eastbays tool showed.