013 - Getting Started With GTO Podcast

WhoanowWhoanow Red Chipper Posts: 19 ✭✭
edited January 2016 in Podcast, Articles, Freebies
Someone please clarify this for me. I thought that a GTO strategy was one that would win against any other strategy except another GTO one. That is, it is a singular strategy that does not change based on the opponent. If your opponent is too loose, it wins. If your opponent is too tight, it wins. Listening to this podcast, it seems the procedure is to know your range on certain street, say the river, make some assumptions to determine your opponents range, and solve for the optimal solution. I've recently been reading Will Tipton's book, and he goes through this same procedure. This implies that the optimal strategy is dependent on your opponents range. If you make different assumptions about your opponents range, then you get a different solution. That is where I'm confused. If you're playing a GTO strategy, your opponent's range doesn't matter. True or false? Please explain.

Comments

  • MikeMike Red Chipper Posts: 1
    Great episode!! Just to clarify, Mike, around the 35 minute mark when you discuss the BB/100 improvement you have calculated in certain scenarios, is that improvement solely due to following the pure GTO solution, or is that from using the GTO solution to personally create a stronger exploitative strategy?
  • The MuleThe Mule Red Chipper Posts: 790 ✭✭✭
    Great question Whoanow, I've been pondering the same thing.

    If you take two extreme cases:
    Opponent 1 plays any two cards;
    Opponent 2 plays only pocket aces.

    If you play heads up using the the same "GTO" strategy against both players, then it seems reasonable to expect to win comfortably against opponent 1, but surely you will lose against opponent 2 - in any given hand. But overall, and I'm just speculating at this but it seems reasonable, you should still win against opponent 2 because they will be playing so few hands.

    So my thought is, a GTO strategy should have a positive expectation against any opponent in the long run, but not necessarily in each hand that is actually played out.

    The extension of this is, I think a GTO strategy must actually require a given pre flop range ? frequency ? from each position, for both betting and calling. You can't just take your usual opening range and play GTO post flop and be guaranteed to win.

    Also interesting, how would a GTO strategy perform multiway with different player types - if you were playing 3 handed with opponents 1 and 2 for example, is your GTO strategy still guaranteed to win ?


  • Mike POKERINABOXMike POKERINABOX Red Chipper Posts: 14 ✭✭
    Whoanow wrote: »
    Someone please clarify this for me. I thought that a GTO strategy was one that would win against any other strategy except another GTO one. That is, it is a singular strategy that does not change based on the opponent. If your opponent is too loose, it wins. If your opponent is too tight, it wins. Listening to this podcast, it seems the procedure is to know your range on certain street, say the river, make some assumptions to determine your opponents range, and solve for the optimal solution. I've recently been reading Will Tipton's book, and he goes through this same procedure. This implies that the optimal strategy is dependent on your opponents range. If you make different assumptions about your opponents range, then you get a different solution. That is where I'm confused. If you're playing a GTO strategy, your opponent's range doesn't matter. True or false? Please explain.

    A GTO strategy doesn't ensure you'll win. It just means it's your best choice if your opponent is able to change his strategy based on what you have chosen (the whole idea of iteration).

    Also, your opponent's range DOES matter. The only way you can solve for the GTO solution is by making assumptions about his range.

    The important thing is what your opponent decides to do with the hands in his range (i.e. his chosen strategy). If you don't expect your opponent to change his strategy based on what you do (i.e. a recreational player), then this is where you get to solve for maximum exploitation using the GTO strategy as your starting point.
  • Mike POKERINABOXMike POKERINABOX Red Chipper Posts: 14 ✭✭
    Mike wrote: »
    Great episode!! Just to clarify, Mike, around the 35 minute mark when you discuss the BB/100 improvement you have calculated in certain scenarios, is that improvement solely due to following the pure GTO solution, or is that from using the GTO solution to personally create a stronger exploitative strategy?

    Thanks! Glad you enjoyed it.
    Answered this in the comments to the video.
  • Mike POKERINABOXMike POKERINABOX Red Chipper Posts: 14 ✭✭
    colldav wrote: »
    Great question Whoanow, I've been pondering the same thing.

    If you take two extreme cases:
    Opponent 1 plays any two cards;
    Opponent 2 plays only pocket aces.

    If you play heads up using the the same "GTO" strategy against both players, then it seems reasonable to expect to win comfortably against opponent 1, but surely you will lose against opponent 2 - in any given hand. But overall, and I'm just speculating at this but it seems reasonable, you should still win against opponent 2 because they will be playing so few hands.

    So my thought is, a GTO strategy should have a positive expectation against any opponent in the long run, but not necessarily in each hand that is actually played out.

    The extension of this is, I think a GTO strategy must actually require a given pre flop range ? frequency ? from each position, for both betting and calling. You can't just take your usual opening range and play GTO post flop and be guaranteed to win.

    Also interesting, how would a GTO strategy perform multiway with different player types - if you were playing 3 handed with opponents 1 and 2 for example, is your GTO strategy still guaranteed to win ?


    You're on the right track.

    The best way to think about this is as PAIRS of strategies (one for each player). You have a set of conditions (hand ranges, flop, pot size, stack sizes, and betting options) and this create the options for both players. Game Theory Optimization (GTO) is the act of solving for the strategy pair such that neither player can change their strategy and gain an edge.

    If you alter any of the conditions (e.g. hand ranges), then you're changing the scenario and this will likely change the optimal strategy pair.

    Also, these techniques are only possible for heads-up scenarios (e.g. on the flop). The models break-down when you add more players.
  • JfdiwinsJfdiwins Red Chipper Posts: 52
    Hi Mike,
    This sounds interesting but I am struggling to understand how it would work in practice.
    Please could you provide a hand example of how GTO would work?
    Thanks
  • Mike POKERINABOXMike POKERINABOX Red Chipper Posts: 14 ✭✭
    Jfdiwins wrote: »
    Hi Mike,
    This sounds interesting but I am struggling to understand how it would work in practice.
    Please could you provide a hand example of how GTO would work?
    Thanks

    Certainly. I'll give you 2 examples of varying levels of difficulty.

    Simple Example - Bluff Catching River

    Assume 100bb effective stacks.
    You call from the big blind with :Kc:Ts against a Small Blind 3bb open.

    The flop is :Kd:5c:4c
    Villain bets half pot and you call.

    The turn is :9h
    Villain bets half pot and you call

    The river is :Jc
    Villain bets half pot. What do you do?

    Without any other information, the best way to solve this is using GTO principles. I'll be teaching this in the GTO webinar.

    Advanced Example - Flop Exploitation

    Assume 100bb effective stacks.

    A solid reg opens from the High Jack to 3bb.
    You call with a standard, tight-ish range on the button.

    The flop is :Ah:Ad:4s
    Villain checks.

    What is your villain's optimal checking frequency in this spot?
    What do you think his ACTUAL checking frequency and range are?
    Is this exploitable, and if so how do we exploit?

    This example comes directly from the webinar where I'll be demo'ing GTO tools that make solving for this stuff surprisingly easy.

    This example is a very common spot that people get wrong, which means it is a big opportunity for us to profit against the entire field! The more of these you add to your arsenal, the higher your win-rate climbs.

Leave a Comment

BoldItalicStrikethroughOrdered listUnordered list
Emoji
Image
Align leftAlign centerAlign rightToggle HTML viewToggle full pageToggle lights
Drop image/file