Thought this would be better than the General section, and I did some searching and couldn't find what I'm looking for, so here goes...

I've been reading about / understanding the Independent Chip Model, but I'm wondering why it's so favored over just Chip Equity and using your Pot Odds. I mean, if our Expected Value is positive, that would always result in your tournament equity being pushed up...so why not focus on just EV and not worry about your Tournament Equity, which is just more calculations? Are there situations where your EV might be negative, but your overall Tournament Equity is increased? I can't really think of a situation where it would be, unless you have a huge stack and the chance to knock a very small stack out will leave you in the money.

Or is the ICM useful more on bubble situations, where early on it's not relevant and EV is good enough?

I see so many examples of how it's calculated and what it means...but never when you should use it or an example of EV in chips vs tournament equity and the different results you get.

ICM is an ok model and is always better than cEV. There's other ways to model your equity that are better though.

As a general rule you just want to play a little tighter in sit and go's since there are consolation prizes.

Mainly comes into play on the bubble. For instance if utg has 70 chips, button has 1000 and sb has 1000, and only top two play, sb should be pushing extremely loose and bb should be calling extremely tight. In practice, most people don't actually play this way, but as a general push loose and call tight on the bubble since a rational player doesn't want to risk coming in last on a marginal hand even if it is probably ahead.

I'm not sure what's better than ICM other than Future game simulation. I suppose it's a mystery.

ICM is a basically an algorythm that converts your chip expected value (Cev) into $ expected value ($ev), based on chipstacks and payout-structure.

In a cashgame or winner takes it all tournament or Heads-up in any game, you use Cev. Every chip won is worth exactly as much as a lost chip. So as soon as a play nets a negative expectation compared to other plays, you dont do it. And vice-verca with winning. So you always go for the play that nets you the most chips or loses you the least if winning chips isnt possible.

In a tournament this is different. Because a chip won isnt worth as much as losing one. Take a normal 10+rake$ 9man Sit-and-go with a 50/30/20 payout. There are 90$ in the pot. First gets 45$, second 27$ and third 18$, so winning every chip means you increase your stack by 900%, but your invested $ only increase by 450%. You also can increase your investement by 270% eventho you only had 10% of your initial chipstack on the bubble...

Thats why we use the ICM algorythm to convert the chip value in $ value.

I give you an example. Live game, 3 players, 10$ buy-in, no rake, 1st and 2nd get 50% of the pricepool:

First hand, you start with 1.000 stacks blinds are 10/20. BTN folds, SB goes all-in and turns his AsKd over, you look down at your hand and see 7h7d. You have pretty much 55% to win.

So chip-ev is easy, folding loses you 20 chips, calling nets you 100 chips, thus calling has a positive expectation by 120 chips compared to folding. Easy call.

But we are in a tournament, your equity when calling and loosing is 0$. Your equity when calling and winning is 15$. 55% of 15$ is 8,25$. So your expected $ value when calling is 8,25$.

While folding leaves you with 980 which represents a stack worth ~9.83$. Thus calling loses you 1.58$ each time you call, eventho you end up with a 1120 stack.

There also spots where your $ev increases while your Cev decreases, but those are usually spots where other peoples are involved.