GTO poker: MITpokerclass 2013 Lecture 3 feat. Bill Chen

Lecture has two parts, the material is quite mathematically intensive, however its a good introduction (and wake up call), to GTO strategy. The basic idea is somewhat simple, there is an ideal strategy (GTO = game theoretical optimum) in which your strategy cannot be exploited regardless of how villain responds. What is not talked about often in the media and commentator explanation of poker and poker strategy, is that given no other specific information, we should be striving to play a GTO strategy (whatever that might be).

GTO is difficult to define, as poker is a complex game, however certain specific situations are easier to define than others, and we can use those situations, in order to get a general understanding of what a GTO strategy might be in other more complex spots.

Often professionals with this knowledge are not running constant scenarios through their heads wondering what lines their opponents might take, but rather they are simply reverting to GTO strategy, and hoping to force their opponents to open up and make the mistakes themselves.

The idea of the video is if we get to the river and don`t like our GTO options, then we might re consider how we should play the turn, if the turn isn`t a favorable set up with respect to the hands we have available to us (because of our flop and preflop strategies), then we will revert to a flop analysis and so on.

I`ve watched these videos a few times, it was complete gibberish the first time, quite overwhelming, and basically disheartening...how ever after much study and consideration, this new understanding for the game has bred and new love for it...

Consider this topic as part of your study for the game, it is quite interesting and certainly the future of it:

https://www.youtube.com/watch?v=BuxCNZ0RVKA

Comments

  • will watch... but going into it I believe this GTO stuff is just a good way to win the minimum, consistently.
  • Seems to me his solution to the second problem is incorrect
  • Bfillmaff wrote: »
    will watch... but going into it I believe this GTO stuff is just a good way to win the minimum, consistently.

    In order to win the maximum, or in other words to "exploit" a player maximally, we need to be able to recognize them as not "un-exploitable" or not GTO.

    It needn't be very mathematical though, what good non mathematical players refer too as "good poker" is likely somewhat closer or closer than most to GTO.
    pkrfce9 wrote: »
    Seems to me his solution to the second problem is incorrect
    I couldn't exactly decide which your were referring too, time stamp?

    edit: actually i guess you mean the 2nd vid in whole!?
  • The 2nd problem in the 2nd video.

    tapatalk puts this here to annoy YOU
  • pkrfce9 wrote: »
    The 2nd problem in the 2nd video.

    tapatalk puts this here to annoy YOU
    I think it is the one where X picks up a card and wins if its not a club?

    x has 3 out 4 hands that will always win...

    if X bets 100%, 75% of the time he wins

    Y cannot profitably call.

    Is that the correct scenario?
  • I just assumed he watched no video and was messing with you. :p
  • Startles wrote: »
    In order to win the maximum, or in other words to "exploit" a player maximally, we need to be able to recognize them as not "un-exploitable" or not GTO.

    I don't have my head fully wrapped around this GTO stuff, so when Bill Chen put up a slide in video #1 titled "Optimal play vs. Exploitative play" I thought "OK, he's going to explain to me why his way is better..." but IMO he then failed to actually make a case for why optimal would be better than exploitative. He kind of laughs like "of course optimal is better" and moves on. Maybe I need to watch again but that's how I remember it.

    You make a valid point about identifying the GTO play and comparing it to villian's play to find the exploitable weakness. Is that extra step actually required though? At the end of the day we're talking about using the exploitative play to maximize EV, so how you get there isn't as important - in most games anyway.
    Startles wrote: »
    It needn't be very mathematical though, what good non mathematical players refer too as "good poker" is likely somewhat closer or closer than most to GTO.

    In a way I think you are right, but the term "good poker" has a lot of room for interpretation. Good poker to most would be an ABC strategy, designed to win slowly and consistently... but just as exploitable as anything else.

    You did say good players though, so I can't claim to know what I am talking about.
  • Bfillmaff wrote: »
    I don't have my head fully wrapped around this GTO stuff, so when Bill Chen put up a slide in video #1 titled "Optimal play vs. Exploitative play" I thought "OK, he's going to explain to me why his way is better..." but IMO he then failed to actually make a case for why optimal would be better than exploitative. He kind of laughs like "of course optimal is better" and moves on. Maybe I need to watch again but that's how I remember it.
    yes he presents key parts and doesn't fully put them together, honestly i don't think it was the series i was thinking about but still worth something. It's probably useful to just let the concepts float around and still just continue playing the game like they doesn't exist while you are at the table.
    You make a valid point about identifying the GTO play and comparing it to villian's play to find the exploitable weakness. Is that extra step actually required though? At the end of the day we're talking about using the exploitative play to maximize EV, so how you get there isn't as important - in most games anyway.
    Well the question is how would you get there without a proper bearing point? GTO is the balance point on the scale. In this context if we suggest villain is "loose" we really should mean "looser then GTO" etc. Kind of like calibrating a scale it is very useful to know were the balance point is. I think of it like a "libra" type scale, i don't know the name for it i am looking for though.

    We often think, "oh this play is bad Im going to "exploit" them" but it is really difficult to actually truly exploit them if we don't have a proper starting point for both evaluating their strategy (comparing it to gto), and also for "calculating" our adjustment.


    In a way I think you are right, but the term "good poker" has a lot of room for interpretation. Good poker to most would be an ABC strategy, designed to win slowly and consistently... but just as exploitable as anything else.
    Yes I always try to leave this "room". As time goes on and the game gets harder though, regardless, good poker will equal GTO more an more, this is something that again has room, but is quite helpful in our analysis of the game i think.
  • Startles wrote: »
    I think it is the one where X picks up a card and wins if its not a club?



    x has 3 out 4 hands that will always win...



    if X bets 100%, 75% of the time he wins



    Y cannot profitably call.



    Is that the correct scenario?







    yes but I think that reasoning is flawed.
  • pkrfce9 wrote: »
    yes but I think that reasoning is flawed.
    Is it .25 x 200 we win vs .75 x 100 we lose?

    then we are winning .25 x 200 = 50

    but on average losing .75 x 100 = 75

    So calling with the hopes player X has a non-club has a net result of .5 - .75 = -25

    I skimmed the video and not much practice with the math involved but that's how I understood it so far.
Sign In or Register to comment.