ICM - a query

[david3103]

Is it reasonable to assume that because Player A has 50% of the chips at the Final Table he has a 50% chance of winning the tournament?

At it's most simple - taking a table of 8 players each of whom have 12.5% of the chips ICM would make each of them equally likely to win.

But what if the final table was Longy, Royal Flush, The Tank, Phil Ivey etc
plus me

Even if I went to the table as a massive chip leader, would ICM be accurately reflecting my chances of winning?

If I read it right, ICM helps me to make better decisions if, and only if,  I have a good read on the other players ranges.
Given the standard of play at my regular games I'm off to study the relative values of my range against random cards.
And to brush up my mental arithmetic

[thetank]

But what if the final table was Longy, Royal Flush, The Tank, Phil Ivey etc

lol, spot the difference. Flushy's never that good.




As for the question of ICM accurately reflecting your chances of winning. It's a model that works for relatively short stacked poker situations such as sit n gos where the average stack when short handed is usually less than 15 big blinds.

If players are all relatively deep then it's no longer about ICM. More options exist than folding and shoving and the Phil Iveys of this world will have a chance to bring their edge to bear as they play poker on the flops, the turns and and the rivers.

ICM analysis tools such as SnG powertools and SnGWiz will flag a hand if players are more than 10BB deep. A message comes up saying that you might want to consider tactical options.
 

If the stacks are shorter at the final table, say the average stack is 10BBs, and you utilize ICM well, Phil Ivey will have very little edge on you. ICM in this case then would reflect your chances of winning.

If the average stack is 60BBs, not so much. I'll see you by the cashiers desk coz I'm getting 8th.

[AndrewT]

ICM analysis tools such as SnG powertools and SnGWiz will flag a hand if players are more than 10BB deep. A message comes up saying that you might want to consider tactical options.

Does the message say 'Unlucky pushbot - you'll have to play actual poker.'

[thetank]

pretty much.

[Longy]

Yes ICM assumes everyone is equally skilled, 10bbs deep there isn't much edge to be gained by anyone therefore ICM is fits quite nicely with this.

In some spots it is worth turning down some reasonably marginal spots like +.1% (% of the prizepool) as your opponents will make enough mistakes to make it such that you don't need to push this spot. Especially where my read is on a players range is so so. To be honest i push every edge i can get these days.

[david3103]

So
I can work out my 'Chip Value' and that of others in any given hand to a fair degree of accuracy (easiest in a 10 seater with 1,000 starting stack, but do-able in other combinations)
I can study the relative value of the ranges for each player via Poker Stove
All I have to do then is stick to a limited range of hands myself and I'll be able to make the right moves more often?
hang on a sec - I just need a pencil and paper and .... bugger the time's run out!

Seriously - maybe a spread sheet would work to get the numbers in my head....

If I stick with 1010+, A10s+, AQo+ and run each against ATC, and then run each against ranges from 35% upwards

does this make sense?

or am i over-valuing this theory like a man with A7 on an KQ7 flop?

[Longy]

So
I can work out my 'Chip Value' and that of others in any given hand to a fair degree of accuracy (easiest in a 10 seater with 1,000 starting stack, but do-able in other combinations)
I can study the relative value of the ranges for each player via Poker Stove
All I have to do then is stick to a limited range of hands myself and I'll be able to make the right moves more often?
hang on a sec - I just need a pencil and paper and .... bugger the time's run out!

Seriously - maybe a spread sheet would work to get the numbers in my head....

If I stick with 1010+, A10s+, AQo+ and run each against ATC, and then run each against ranges from 35% upwards

does this make sense?

or am i over-valuing this theory like a man with A7 on an KQ7 flop?

That is why we use ICM software like sngwiz which does all the maths for us, you can study 100's of different situations a day with software like this with a couple of clicks. I aim to write that post at some point today or tomorrow.

[thetank]

In the meantime, SnGWiz has a 30 day free trial and some excellent video tutorials on how to use it.

Check it ooooot.