No announcement yet.

Exploiting Villains Who Do Not Understand ICM

  • Filter
  • Time
  • Show
Clear All
new posts

  • Exploiting Villains Who Do Not Understand ICM

    In poker the way we generally make money is not by making amazing plays but by our opponents making mistakes.

    Towards the end of tournaments we need to consider ICM pressure when creating our ranges.

    Often this simply means we need to be calling all-in with a tighter range than usual.

    If an opponent was not aware of this and was calling down with their usual range then clearly this is a mistake. Our opponents making mistakes a good outcome for us and is something that we should be looking to induce.

    The problem here is that the opponent's mistake is also increasing the risk exerted against our own tournament life. Is this technically reversing the ICM pressure back onto us?

    Should we still be shoving with the same range against this opponent to allow them to make a mistake? Or do we need to tighten up from a theoretical GTO range as our opponent's mistake means our shove is now a potential -ve ICM_$EV decision?

  • #2
    I'd say that anything that is GTO (including ICM) are great to understand as the baseline, and you can and should veer from that to take advantage of known player mistakes.

    What I mean by that is that if you were playing against perfect robots or top level GTO pros, then sticking with the pure ICM theory is the best you could do.

    If however, you were playing with someone that isn't going to fold as often as the "should", then you shouldn't follow the GTO method that is assuming a higher level of fold equity than you actually have.

    Likewise, if someone isn't shoving as often as they "should", you shouldn't be calling with a range that is factoring in a wider shoving range than what you are facing.

    Understanding the baseline perfect play and also how that perfect play would change as your opponents move away from perfect play, is the best approach.


    • #3
      Originally posted by jeffreyalanleach View Post
      If however, you were playing with someone that isn't going to fold as often as the "should", then you shouldn't follow the GTO method that is assuming a higher level of fold equity than you actually have.
      Wouldn't this therefore mean that our opponent is no longer making a mistake as we are not playing the theoretical correct approach and is instead playing perfectly against our range? Purely from an ICM point of view.

      I understand that the idea of GTO from a non-ICM perspective, in a cash game for example, is that it's perfect and unexploitable way (in theory) to play against a balanced opponent but that we make more money by making exploitative plays targeting our opponents' who are not playing GTO themselves.

      I don't think I'm explaining myself very well...


      • #4
        It's a complex subject for sure and I don't pretend to be an expert on it, but I'll try to give an argument for what I said.

        If there are three people left in a tournament and one of them is a small stack, the biggest stack could be in a great situation to push wide on the medium stack. If for example the medium stack stands to gain $1000 with a double up but lose $10,000 with a loss, the medium stack should only call with hands that have 90% equity against the wide range of the big stack. So, they should pretty much fold everything when the big stack jams on them.

        But if the medium stack isn't thinking that way and is going to call you with any ace and any Broadway hands, the big stack better not be jamming 8 3 offsuit. In some cases you can get just as many folds with a raise as with a jam late in a tourney when people believe they are in a jam or fold situation and strict adherence to any theoretical system that assumes perfect play won't take that into account.


        • #5
          In general, when ICM considerations are in play, the people playing a given pot have big potential downsides and the people not playing the pot share a decent amount of the upside. This is both in the sense that chips lost are worse to a greater degree than chips earned are good. Whichever player in the pot loses chips, if they pass below any of the players that they were previously ahead of, that's a benefit to those players who weren't in the hand since they now have less ICM pressure due to there being more stacks shorter than them. And if any player busts that's obviously beneficial to everyone who didn't bust. The player who wins a given pot has the most upside, but you can't win a pot without putting yourself at risk for the downside obviously.

          So how do you exploit a player that doesn't understand ICM? You let them suicide themselves against someone else's stack. You don't raise their blinds purely due to ICM pressure situations (what jeffreyalanleach said). You had the right idea I think, tighten up against them, let someone else play cop. You shouldn't be too extreme with this, just don't be too creative against plays that won't feel the pressure you're applying.