Tuesday, October 19, 2010

ICM and the Previous Post

Well, I have done the ICM calculations. Now let’s walk through them. Firstly we need to assign a raising range for the big stack. He is pretty standard so let us say:
  • 25% holding Ax or Kx making me a 73/27 favorite
  • 25% holding a pocket pair making me a 43/57 underdog
  • 50% holding ATC, other than the ones above, making me a 65/35 favorite
  • (Numbers above come directly from Pokerstove)

Secondly, let us assume that he will call my all in reraise 100% of the time (and that the bb folds 100% of the time). Thus, we can do the math and see that I’ll win the hand 62% of the time.

Now we can take our stack sizes for winning, losing and folding and put them in an ICM calculator.
  • FOLD: If I fold, I’ll have 1275 chips and have an ICM of 17.3%
  • LOSE: If I get all in and lose, I’ll have no chips and have an ICM of 0%
  • WIN: If I get all in and win, I’ll have 2950 chips and have an ICM of 26,0%
  • (Numbers above come directly from www.chillin441.com)

Finally, we can combine our ICM values with the probability of them occurring.
Winning(62% x 26%) + losing(38% x 0%) = ICM 16.1%

Folding has an ICM of 17.3 and is more profitable than reraising all in which has an ICM of 16.1
In other words folding AK in this spot is slightly better than reraising all in!

Actually, this is so close, that any changes to the assumtions at the start may very well make this slightly profitable, for example there is always a tiny chance that he will fold to my reraise.

Roland GTX


  1. MrE, here is the rest of the hand. I shoved and the big stack called showing 55. The board ran AKxxx and I doubled up with two pair. The very next hand the short stack moved all in utg. I reraised all in with 99. The big stacks folded and my pocket pair held up putting me itm. I got heads up vs the original big stack guy. Once again he raised preflop with A10 and I reraised all in with AQ. He called and the board ran 1010xxx. So I ended 2nd with one bounty.

  2. Interesting stuff. I didn't get a chance to post below before you gave the rest of it.....I was guessing he was holding a pp below 8 and of course calls you 100% of the time. Did you run any numbers if the other shortstack shoves and do you have a range on him as well? I'm with MrE in any case, we are playing to win and not just crawl ITM.

  3. Yup, very interesting indeed. AS u all know, I'm not an expert at all, but I really wonder whether the ICM figures actually are any good here at all. I don't doubt that the figures are correct, but if you look at the situation from a slightly different angle, I am not sure whether folding is the better option here. With 100/200 and 25 in ante, every round on a 4 player table cost you 400 chips, which means that you're out of the game after 3 rounds, or 12 hands for that matter. You do have the exact figures, Roland, but the probality to get a better hand during these 12 hands, is rather low, isn't it? I know that you would have to shove latest when you're down to 500-800 chips, so the actual number of hands is getting ever less.

  4. There is a ton to be said for ICM regarding single table tourneys. I'll my thoughts on ICM and this hand in particular tomorrow. Right now Im busy playing:)

  5. Interesting thread. This one looks to be so close ICM-wise that it's hard to go wrong either way. So it basically depends whether you'd rather try to squeak into the money or go for the gold. In a low buy-in SNG, I'm with MrE and MrSmith (and apparently Roland based on how the hand played out), might as well go for the bigger payout.

    Roland, you may remember a related bubble example from Moshman's book. You have t5000 in the BB, the button has t1500 and the SB t2000. Blinds at 200/400+25. UTG/CO shoves with t5000. You put him on any ace, pocket pair or two facecards. Moshman recommends folding any hand but KK or AA. The key difference, of course, is that you are now one of the two big stacks instead of one of the two small stacks in which case calling with AKo would be more clearly EV-.

  6. I agree that this particular case is borderline. However, the general principles of ICM are definitely worth the effort to learn. As benkogambit points out, Moshman talks about it quite a bit when looking at bubble play. It might be counter intuitive at times, but it is definitely an eye opener!

  7. Did you set the date for next tourny E.?