Thursday, January 28, 2016

Why and how to learn GTO -- The GTO Learning Path

One of the most common questions that I get from players who are looking to get started with learning GTO and integrating GTO concepts is "where do I start?".  Over the last year or two I've put out so many videos / blog posts / strategy packs, etc that it can be difficult to know where the jump in, whether you are completely new to GTO or whether you already have a strong fundamental understanding of GTO concepts.

At the bottom of this post I'm going to put links to all of our relevant GTO articles, videos, strategy packs, quizzes and brainteasers so that hopefully this post can be a one-stop place to learn everything you need to do to master GTO concepts and apply them to your game.

But before I lay out the curriculum I want to start by answering one of the most common and fundamental questions that every player should ask themselves before learning GTO theory and strategy.

Why should I learn GTO?

The core answer is simple very simple, to win more money!  How exactly understanding GTO theory does this however is more complex and often very misunderstood so I want to directly state the core values of learning GTO strategy and dispel some of the misinformation that is out there.  Perhaps the biggest misconception is the idea that the goal of learning GTO is so that we can play the GTO strategy 100% of the time and abandon exploitative play completely.  As we'll see below that is of course ridiculous, and in fact, an understanding of GTO is an invaluable tool in identifying and exploiting leaks in our opponents play.

At the poker table we are going to encounter a variety of players of various skill levels who play a wide variety of strategies.  At various times we will be playing in a situation where we have no read whatsoever on our opponents, or against a player who we know is stronger than us.  At other times we will be playing against weaker players who have clearly defined strategy errors that we believe they are unlikely to correct and that we can exploit.  

Our ability to win in the long run requires that we can both play fundamentally strong poker that is difficult for our opponents to exploit in situations where we are readless or against strong opposition, while also being able to quickly identify and attack leaks in our opponents' strategies in ways that do not open us up to counter exploitation.  Both of these goals are nearly impossible to achieve without a strong understanding of GTO strategy.

Lets begin by considering how to identify and attack leaks in our opponents.  Lets imagine that you are facing an opponent in a 100bb cash game who raised UTG to 3x and you flatted from the button and the other players folded.  He c-bets 80% of the pot on a AT3r flop.  You look at your HUD and note that this players OOP c-bet frequency in single raised pots is 62% and that he opens UTG about 17%.  How should we react to his c-bet with various parts of our range?  

It turns out that without a reasonable knowledge of GTO strategy the question above is very difficult to answer and in fact that for years "standard" play approached this situation incorrectly.  We first might want to ask, is a 62% out of position c-bet frequency too high (and thus should be exploited by calling or raising a much higher % than usual and by betting near 100% after he checks a flop) or to low which of course would be exploited differently..  This is impossible to know without knowing what the correct, unexploitable c-bet frequency is in out of position pots.

With GTO computation we can actually directly asses what an appropriate c-bet frequency on this board is (about 15%) and browse the entire GTO strategy and compare it to how we think our opponent plays.  You can see an example of that here:  We can then lock in our opponents incorrect c-betting strategy and optimize our play computationally against that.

For a long time many players, who at the time were considered quite good,  c-bet a very high frequency here.  They got away with it because no one had the tools or knowledge to correctly identify and exploit the leak.

However, we now know that there are very few boards where c-betting more than 50% of the time as an UTG raiser vs a BTN call is GTO, so c-betting 62% on average is way too high.  This player is c-betting far too aggressively so we should attack his c-bet.  If we have enough of a read to know that this play might c-bet 62% on this specific flop texture we could use a minimally exploitative strategy calculation (which you can learn about below) to determine that even if our opponent plays the turn and river perfectly we can probably raise this bet with as much of 20% (~3x as often as we would raise against a GTO c-bettor) of our range and we should only fold the about a third of our range, netting an EV gain of about half a big blind in a single hand.  

If we believe our opponent at least recognizes that AT3r is a board that should be c-bet from OOP  less often than an average board we might exercise a bit more caution but we can again estimate his c-betting range and determine that we should raise/call far more than usual.  

Without knowing what the GTO c-betting strategy is like is this spot we cannot identify strategy errors.  We can note that some players c-bet in a spot more or less than the population average but we have no way to be sure if that is a mistake or actually play that is better than the population and history shows that significant population level strategy errors can exist in "standard" play for years before they are exploited and fixed.

Similarly consider the situation where we open a 17% UTG range and get flatted by an unknown player in the BTN (or a very tough player).  Since GTO play will grind out a steady profit against sub-optimal players and because we have no specific read on the exact leaks in the BTNs game, our best bet in this spot is generally going to be to c-bet according to GTO.  By having a fundamentally solid and well structured c-betting range in this spot we will increase our win-rate against unknowns and prevent stronger players from exploiting errors in our play.  We can neither prevent leaks from appearing in our own game, nor can we systematically identify and exploit leaks in other players games without knowing what exactly "leak free" GTO play looks like.

So hopefully that gives you guys motivation to put in the study time and become GTO experts.  Here's how you can do it.

GTO Resources / Learning Path

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.