Icm

    • Philipvw
      Philipvw
      Bronze
      Joined: 28.05.2007 Posts: 578
      A3o in BU, 3 players, 100-200 blinds, stack 1400 is push and
      A5o in BU, 3 players, 200-400 blinds, stack 2040 isn't.

      Why is that? I was thinking about Sklansky's M and i thought it was a push fo both.
  • 13 replies
    • helemaalnicks
      helemaalnicks
      Bronze
      Joined: 21.09.2007 Posts: 7,195
      because the FE is higher in example 1?
    • Unam
      Unam
      Moderator
      Moderator
      Joined: 17.08.2006 Posts: 8,999
      Originally posted by Philipvw
      A3o in BU, 3 players, 100-200 blinds, stack 1400 is push and
      A5o in BU, 3 players, 200-400 blinds, stack 2040 isn't.

      Why is that? I was thinking about Sklansky's M and i thought it was a push fo both.
      I can't answer this, unless you give me 2 sample hands, I need the stacks of all players, not only the ones that are still to act, I have to know how many players are left and what the payout structure of the SNG or MTT is.
      All this has a great influence on our ICM decisions.

      So give me more information and I will gladly try to answer your questions.
    • DoigteurFou
      DoigteurFou
      Bronze
      Joined: 16.08.2007 Posts: 1,528
      Just a small question about the usability of icm: how can it be used when you're at the table? Don't tell me you calculate everything in your mind? I spent 2 hours yesterday to find approximations to be able to calculate in mind (and in time) but all i found was'nt usable. Is icm an artifact for players who play on their pcs with their calculator on another window, or can it be used by players?
    • timukasr
      timukasr
      Bronze
      Joined: 26.05.2007 Posts: 1,820
      If you practice in ICM trainer then you remember diffrent situations and you know about with what ranges you have to push, call etc.
    • DoigteurFou
      DoigteurFou
      Bronze
      Joined: 16.08.2007 Posts: 1,528
      ok. I don't think i'll train in since i don't feel it's worth the time given the huge number of hours to cover situations with a model based on big approximations, and which does'nt cover position of blinds...
    • helemaalnicks
      helemaalnicks
      Bronze
      Joined: 21.09.2007 Posts: 7,195
      Originally posted by DoigteurFou
      ok. I don't think i'll train in since i don't feel it's worth the time given the huge number of hours to cover situations with a model based on big approximations, and which does'nt cover position of blinds...
      it does cover position of the blinds, and it isn't based on approximations, but on statistics that derive from a sample that is (close to) infinite.
    • chenny8888
      chenny8888
      Bronze
      Joined: 03.10.2007 Posts: 19,324
      the ICM trainer is REALLY useful. the only complaint i could really have about it is that it gives too many straightforward questions.

      while the actual answers are based on a long series of calculations, it does not actually take long (maybe 1 hour) to very vaguely familiarise yourself with ICM and use it correctly more than 90% of the time. very very cheap price to pay, IMO.
    • Tjablonsky
      Tjablonsky
      Bronze
      Joined: 30.09.2007 Posts: 86
      where can I find formulas to calculate my odds for pushing or folding? I am training now, but I react on situations by filling..
    • chenny8888
      chenny8888
      Bronze
      Joined: 03.10.2007 Posts: 19,324
      if you go into the ICM trainer there is a calculator tab where you can input situations and it will then calculate the range you need to push, call or overcall.

      sadly only applies for a standard sng where blind levels rise moderately fast and top 3 positions are paid. but for MTT i apply a similar strategy when it comes to high blinds (although some people will suggest a slightly looser approach)
    • Tjablonsky
      Tjablonsky
      Bronze
      Joined: 30.09.2007 Posts: 86
      OK I see that, but how do I calculate this range by myself, where can I read that?
    • timukasr
      timukasr
      Bronze
      Joined: 26.05.2007 Posts: 1,820
      You can read it on a golden article :D
    • Mugge88
      Mugge88
      Bronze
      Joined: 29.10.2006 Posts: 510
      You shouldn't try to calculate the ranges yourself, since the calculations are very long, but generally you calculate the $EV for each move any players at the table can make and then you make them fit each other..

      e.g (NOT correct values, and not enough data, but just to show the point):

      SB pushes any two because he knows that BB only calls with top 30%..
      But now if BB knows that SB pushes any two he can perhabs call with top 45% to make his decision +$EV..
      If SB now knows that BB knows that SB pushes any two and that BB has changed his range, he can perhabs only push with top 80% profitably.
      Now BB knows that SB pushes 80% and he should change his range again.. etc etc.

      At some point the call/push ranges will not change anymore and the Equilibrium has been found.. The program calculates this for you.. When there are more players in the hand it get's very, very complex to calculate and I won't be able to do it... This is where the program get's very handy

      Best regards,
      Mugge
    • chenny8888
      chenny8888
      Bronze
      Joined: 03.10.2007 Posts: 19,324
      it's not that hard if you know how to use excel and spreadsheets and iterations to find calculations. i think with my engineering maths modelling skills i could do it...

      that said, i will not (ever) bother explaining to someone else how to do it. mugge's definitely on the right track though