You're a Bunch of Factors

5000/10000 Blinds (8 handed)

CO (25k)
Button (20k)
SB (50k)
Hero (BB) (25k)
UTG (5k)
UTG+1 (85k)
MP1 (25k)
MP2 (25k)

Preflop: Hero is BB with 9, 4 off
UTG all in 5k, UTG+1 raises 35k, MP1 calls all in, MP2 calls all in, CO calls all in, Button calls all in, 1 fold, Hero ?

What would you do, take into account i will be SB next hand and have only 15k left. Another twist is that i am second on the leaderboard and SB is the person i third 50 points behind me. First was already out so we both had already taken over him.

Torak wrote:5000/10000 Blinds (8 handed)

CO (25k)
Button (20k)
SB (50k)
Hero (BB) (25k)
UTG (5k)
UTG+1 (85k)
MP1 (25k)
MP2 (25k)

Preflop: Hero is BB with 9, 4 off
UTG all in 5k, UTG+1 raises 35k, MP1 calls all in, MP2 calls all in, CO calls all in, Button calls all in, 1 fold, Hero ?

What would you do, take into account i will be SB next hand and have only 15k left. Another twist is that i am second on the leaderboard and SB is the person i third 50 points behind me. First was already out so we both had already taken over him.

Not playing for points - call
Playing for points - fold

Easy game.

yep, if you fold, then chances are that a few are going to get knocked out.

but if you want to win and win only, then gambool

I think the above situation is more interesting if you had something like AT suited or two sevens. Then what are we doing? Its still a huge gamble if we want to win the hand, but hands like these are much harder to fold.

maccatak11 wrote:I think the above situation is more interesting if you had something like AT suited or two sevens. Then what are we doing? Its still a huge gamble if we want to win the hand, but hands like these are much harder to fold.

Poker stove it mate against four random all ins. Equity won't change that much between 7,2 os and A,Ks.

I mean minimally to the extent that the pot odds offered will still be attractive.

The consideration to be made here is tournament equity, not pot odds.

Yeah i know thats what im getting at. People (including me) will happily fold 94 to get an extra round of the table and hopefully move up an extra spot or two in points, but i don't think i would ever fold AKs in the same spot, even though there is only a few percent in it, and the principal is essentially the same.

tourney equity would dictate a fold every time i would think.

not sure, but is there a way in calculating tourney equity for a win only, compared to a cash (or top 3 or whatever)?

garth, do you know how to easily calculate tourney equity ICM style? i haven't really looked into this much, just scratched the surface of ICM.

bennymacca wrote:tourney equity would dictate a fold every time i would think.

not sure, but is there a way in calculating tourney equity for a win only, compared to a cash (or top 3 or whatever)?

garth, do you know how to easily calculate tourney equity ICM style? i haven't really looked into this much, just scratched the surface of ICM.

Paging Benn Skender!!!


Oh, wrong forum. No don't use ICM much and normally have someone on Skype to throw some solutions at me!

My thinking was pretty much along those lines, does the fold have equity over the tournament life. 4th place i thought would be enough to hold off the other players for the venue, unless he won it. If he won he deserved it.

Evaluating after the hand, it was 5 handed and it got folded around to me where i have suited A-10, had to put the rest in and got called by a JQ.
Flop 2 3 J, Turn Q, River K
Went on to go heads up against the other player for venue and nightly winner, which i won.

bennymacca wrote:do you know how to easily calculate tourney equity ICM style?

The math is grade school level, but there are too many calculations to perform to make it viable to do so by hand.
For a standard STT:
Probability of 1st is simply your percentage of the chips in play - P(X wins) = X's stack / chips in play.
2nd is the total of your chip stack ratios, given every other player finishing first by removing the 1st place finisher current stack from the chip total, multiplied by the probability of that player coming 1st - P(X 2nd | Y 1st) = X's stack / (total chips - Y's stack) * P(Y wins) -
and so on for the other 2 players winning also, then sum these %s.
which for just a 4 players left top 3 pay situation is 3 subtractions, divisions and multiplications, then summation of 3 numbers (assuming you have worked out P(win) for each player already) - and multiply this by each of the 3 remaining players.

And then do similar for 3rd...

As you can see this all gets tedious and time consuming, hence programs written to do it quickly and easily. And when considering this, your own numbers are useless except in comparison/conjunction with the numbers for the other player(s) in the pot, so you do need to know them!

But once you have your probability of 1st, 2nd 3rd, you simply multiply these by the prizes, ie P(win) * $win + P(2nd) * $2nd + P(3rd) * $3rd and you have your overall $EV in the tournament (STT SnG).

Clearly doin this for a decent size MTT close to the bubble is even more complex (in terms of number of calculations - at a glance it looks like a problem polynomial in time complexity, atleast O(n^2) ). There have been some simplifications tried by some people, but am not sure how well they actually model the situation, ie Work out P(win) * $win, and divide other prizes equally. This clearly undervalues big stacks, but is also a hell of a lot simpler to calculate.

Here is an online ICM calculator for upto 10 players...

Edit: 1 interesting thing to note, if you play around with some figures, and then adjust for someone busting, is how the equities change for players NOT involved in the hand... They increase to, but the bigger the stack the less the increase. It seeems (just on casual inspection) that close to avg chips results in the largest increase when a player busts (of those NOT involved in the hand), which is good evidence for why medium stacks are more risk averse on the bubble than the big or small stacks... Ergo why bullying them works best if you are the big stack