GTO vs Exploititive play

John A

John A

Poker Zion Coach
Silver Level
Joined
Apr 12, 2012
Total posts
6,492
Awards
3
Chips
37
Maybe someone can help me with something I may be missing. Why would you want to learn GTO play at nano/micro or small stakes? I'm not even sure how much it would really apply at mid stakes in the games I play. I mean, it does against certain regs, but still I play mostly exploitative. I'm asking that because of learning tools that try and teach GTO in no-limit holdem. I tried to convince myself it might be a good base to use and operate from, but as a learning model it's severally flawed and you'd end up leaving tons of money on the table if you applied GTO strategy to these games.

Am I missing something?
 
D

DunningKruger

Legend
Silver Level
Joined
Feb 7, 2013
Total posts
1,030
Chips
0
No I don't think you're missing anything because I've been saying the same thing for years. I mean substantial losers at those limits might stand to benefit from it somewhat but exploitative play is approximately 785,300 times more effective in uNL and SSNL games. An understanding of game theory doesn't hurt and a rough facsimile of "GTO" play can be handy to fall back on in MSNL games against superior opponents (or if you're writing scripts for bots lol), but even in those games deviating from GTO is critical in targeting whatever weak link(s) the game is built around.
 
R

rumsey182

Rock Star
Silver Level
Joined
Sep 27, 2013
Total posts
432
Chips
0
it is not that you have to play in a GTO unexploitable way while learning it, just knowing what is or isnt exploitable and learning the math and frequencies is always helpful

what better way to spot a leak
 
John A

John A

Poker Zion Coach
Silver Level
Joined
Apr 12, 2012
Total posts
6,492
Awards
3
Chips
37
it is not that you have to play in a GTO unexploitable way while learning it, just knowing what is or isnt exploitable and learning the math and frequencies is always helpful

what better way to spot a leak

Maybe you can explain better, because I'm still not sold. I think learning combos, understanding equity, and opponent tendencies is far more important than learning GTO play. There's no way in hell I'm teaching a nano-small stakes player about GTO until they understand hand ranges, combo counting, equity, and opponent player/types.

Maybe you can expand and make me a believer though.
 
L

loomis311

Rock Star
Silver Level
Joined
Oct 2, 2013
Total posts
173
Chips
0
I agree with you John. I think GTO in a micro stakes game will never earn the same monies as exploitative play. I believe exploitative play is more for a higher stakes game.
 
L

loomis311

Rock Star
Silver Level
Joined
Oct 2, 2013
Total posts
173
Chips
0
I agree with you John. I think GTO in a micro stakes game will never earn the same monies as exploitative play. I believe exploitative play is more for a higher stakes game.
 
Arjonius

Arjonius

Legend
Silver Level
Joined
Jun 8, 2005
Total posts
3,167
Chips
0
Maybe you can explain better, because I'm still not sold. I think learning combos, understanding equity, and opponent tendencies is far more important than learning GTO play. There's no way in hell I'm teaching a nano-small stakes player about GTO until they understand hand ranges, combo counting, equity, and opponent player/types.

Maybe you can expand and make me a believer though.
It seems to me that while learning GTO play isn't useless, it's also not a top priority at the micros where not that many players will recognize your exploitable tendencies and not all of those will adjust effectively to take advantage of them. And even when you run into a couple, how often will / can they exploit you enough so you're -EV overall?
 
John A

John A

Poker Zion Coach
Silver Level
Joined
Apr 12, 2012
Total posts
6,492
Awards
3
Chips
37
It seems to me that while learning GTO play isn't useless, it's also not a top priority at the micros where not that many players will recognize your exploitable tendencies and not all of those will adjust effectively to take advantage of them. And even when you run into a couple, how often will / can they exploit you enough so you're -EV overall?

I want to agree with you. I'm in between saying that it's useless or has some minor benefit. I've thought about this a lot over the years, and I always come back to the same place.

I know there's a math movement of backgammon and other players trying to focus on GTO, or nash equilibrium solutions as they should more accuretly be called, to no-limit holdem. I completely understand how they apply to more closed systems like backgammon, I'm just not completely convinced they will ever apply to no-limit holdem. I've had debates about this with some of the top players in the world, and while I think it's an interesting study on some levels, I'm not sure of the practical application to a game that includes such a high level of psychology like no-limit holdem.

In games that psychology has a very minor influence, like backgammon or chess, I get it. GTO is what you should be striving for.

Like I said, I'm completely open to be convinced otherwise.
 
Arjonius

Arjonius

Legend
Silver Level
Joined
Jun 8, 2005
Total posts
3,167
Chips
0
I want to agree with you. I'm in between saying that it's useless or has some minor benefit. I've thought about this a lot over the years, and I always come back to the same place.

I know there's a math movement of backgammon and other players trying to focus on GTO, or nash equilibrium solutions as they should more accuretly be called, to no-limit holdem. I completely understand how they apply to more closed systems like backgammon, I'm just not completely convinced they will ever apply to no-limit holdem. I've had debates about this with some of the top players in the world, and while I think it's an interesting study on some levels, I'm not sure of the practical application to a game that includes such a high level of psychology like no-limit holdem.

In games that psychology has a very minor influence, like backgammon or chess, I get it. GTO is what you should be striving for.

Like I said, I'm completely open to be convinced otherwise.
As a low-level recreational player, I haven't put much thought into this and certainly haven't discussed it with top players. So, my opinion is pretty much "quick and dirty" and just my own.

So, if there are solid arguments as to why I should re-think this, I'd certainly be interested to see them since re-thinking is a form of learning / improving.

That said, although I have no empirical evidence, when I consider the balance between how many players have highly exploitable flaws plus how often those situations come up vs. how often and how much I'd actually be exploited if I play in a way that is theoretically more exploitable, my admittedly purely subjective feeling is that it's pretty hard to imagine GTO actually being more profitable at the micros.

Btw, can you point me toward anywhere I can see pro-GTO arguments?
 
B

baudib1

Legend
Silver Level
Joined
Dec 2, 2008
Total posts
6,635
Chips
0
The bottom line is that if you are playing exploitatively in order to profit off the vast majority of players at your stake, by definition you can be exploited.

It is virtually impossible to play GTO because forming a GTO range in one spot depends greatly on knowing precisely how villains will respond, and it's not just on one street but all streets.

However, having a solid understanding of GTO concepts helps greatly in building starting points for, say, preflop ranges, c/c ranges, etc.

If you are playing exploitatively, then guaranteed your frequencies are way off -- in ways that will make you money against a ton of players, but in ways that can easily be attacked by thinking players pretty quickly. For instance, if you build a range where you are 3-betting AJ+/KJ+/QQ+ vs. BTN openers, then your range is totally unbalanced and totally sucks on any Broadway board.

Similarly, if you are always C-R your nut flush draws + overs, then your range is capped on so so many runouts that your c/c range is going to suck. And if your c/c range sucks, you are letting the BTN exploit you because he can then cbet ATC.
 
duggs

duggs

Killing me softly
Silver Level
Joined
Jul 28, 2011
Total posts
9,512
Awards
2
Chips
0
Im not even sure how one would only learn to apply GTO without understanding the game and its ideas. Like GTO makes huge assumptions about others playing, but it doesn't mean its not a nice baseline, it also removes a lot of plays that are fundamentally unsound.

But really our aim as players should be to play optimally, i.e. to play as optimally as possible given our opponents strategy and beliefs. is we can make exploitive adjustments, awesome! but if our opponent has the potential to make exploitive plays or we are unsure about how our range looks in a particular spot, balancing our range to protect ourselves seems like a good idea.

Also I think that GTO principles are pretty handy to use as reference points, i.e. Knowing what an optimal push/fold decision would be v nash callers, and that allows us to make more accurate adjustment in either direction. Or facing a river bet and knowing that the optimal bluffing freq is what evs % and we can then make inferences about whether we expect less or more than that in factoring to calling.

More over I'm not convinced that the two concepts are mutually exclusive, I don't see why learning some GT principles wouldn't round out a baseline understanding of the game.
 
LD1977

LD1977

Legend
Silver Level
Joined
Dec 22, 2012
Total posts
3,091
Chips
0
"GTO makes huge assumptions about others playing" - Incorrect, it is basically a defensive approach that can't be exploited by any strategy so it doesn't assume anything.

GTO play is good as default and after you get reads on opponents you can exploit then when chance presents itself.
 
duggs

duggs

Killing me softly
Silver Level
Joined
Jul 28, 2011
Total posts
9,512
Awards
2
Chips
0
"GTO makes huge assumptions about others playing" - Incorrect, it is basically a defensive approach that can't be exploited by any strategy so it doesn't assume anything.

GTO play is good as default and after you get reads on opponents you can exploit then when chance presents itself.

GTO assumes rationality, ie that your opponents are rational, ie that they play optimally. That's a pretty huge assumption to make at any stakes and you would be better served with a baseline exploitive strategy v the representative agent, but as you move up in stakes the rep strategy converges to GTO. Thus GTO becomes your best response.
 
H

HooDooKoo

Visionary
Platinum Level
Joined
Aug 24, 2013
Total posts
985
Chips
0
Maybe someone can help me with something I may be missing. Why would you want to learn GTO play at nano/micro or small stakes? I'm not even sure how much it would really apply at mid stakes in the games I play. I mean, it does against certain regs, but still I play mostly exploitative. I'm asking that because of learning tools that try and teach GTO in no-limit holdem. I tried to convince myself it might be a good base to use and operate from, but as a learning model it's severally flawed and you'd end up leaving tons of money on the table if you applied GTO strategy to these games.

Am I missing something?

John --- I am a fairly accomplished amateur player (approaching $200K profit since 2002) and I am now primarily a HU NL Tourney specialist. My approach to the game in the beginning was not an intuitive one. Rather, it was very mathematically based (as I'm a reasonably bright guy with a Masters in Applied Stats). I started out playing limit hold'em and was very successful because I am patient/disciplined enough to use hand selection to my advantage and I had no trouble folding hands I opened if I missed and got played back at. After a while, though, I got bored with limit (even though I was making solid money playing it) and I ventured into NL play to mix things up.

As a math-based player coming from limit, I was (unsurprisingly) a rock. I fared well against calling stations and maniacs, but got exploited by good LAGs. In an effort to improve my game against LAGs, I decided to learn to play HU. Learning to be a successful HU player was great for my full-table NL game (six-max or 10) and reduced how often I was exploited. In fact, I enjoyed the action so much that HU NL tourney play became (and remains) my go-to game.

Having said all this, my poker "home base" has been and always will be math- based (as opposed to intuitive, tell-based, or exploitative). When I heard about GTO poker, I was VERY excited because I was looking for ways to further reduce the "exploitatability" of my play. So I devoured "The Mathematics of Poker" by Bill Chen and Jerrod Ankenman and "Expert Heads Up No Limit Hold'Em" by Will Tipton. I sometimes wonder why I bothered with "The Mathematics of Poker" because I had already come to the conclusion that you (and others) have --- that GTO poker is, IMO, only nominally useful. Despite my reservations, I read "The Mathematics of Poker" from cover to cover, being sure I understood the concepts and the math. When I finished it, I still felt exactly the same way about GTO poker that I had after reading Tipton's book (which I read first because the math is somewhat simpler than the math n Chen/Ankenman's book) --- GTO poker is theoretically interesting but has, IMO, limited practical value.

The reason that I think the value of GTO poker is so limited is as follows: the equity calculations are incredibly dependent on ACCURATELY ranging your opponent's hand --- a skill that is widely employed but, IMO, rarely executed well. While the statement above devalues the practical application of GTO poker, it also devalues one of the central tenets of your poker analysis tools (based on what I've seen from you here at CC) --- that of hand-ranging.

Please don't misunderstand what I'm saying, as I think that hand-ranging is a valuable and useful skill in many situations/contexts. Having said that, I also think that basing poker decisions (fold/call/raise) based on equity calculations that are based on hand-ranging is taking things too far for most players --- including the large majority of winning players. I say this for two reasons: 1. Few people can accurately perform equity calculations at the table (even after hard-core use of pokerstove or any other equity calculation tool); and, more importantly, 2. Few people's hand-range assignment skills are good enough that their equity calculations are particularly useful. Online players with good analytical skills and a significant amount of HUD experience have a shot at reasonable hand-ranging against players they have a lot of data on. In all other cases, though, I question most people's ability to accurately/effectively hand-range. There are clearly some people that can do it accurately, and those people don't need GTO play or equity calculators to make good decisions at the poker table. For everyone else, I question the value of hand-ranging.

Please do not interpret my statements above as an attack on you or your teaching methods. I'm certain that you're both a better player than I am and a better teacher than I am. Furthermore, based on what I've read from you, I have reason to believe that your hand-ranging skills are very good. As a result, GTO poker would likely help you against players you have little data on/are unfamiliar with. Once you have enough data on your opponents, though, I agree with the central theme in this thread --- that exploitative is much more effective/profitable than GTO play if you have properly identified your opponent's tendencies. Until that time, however, GTO play is a reasonable approach for those people with proven hand-ranging skills. For most everyone else, I think it's much ado about nothing.

Best of luck.

-HooDooKoo

P.S. I wrote the missive about my background to open this post so that you would understand that I'm a mathematically inclined player with an real math education, as I think that background is important in evaluating the usefulness of my post.
 
John A

John A

Poker Zion Coach
Silver Level
Joined
Apr 12, 2012
Total posts
6,492
Awards
3
Chips
37
As a low-level recreational player, I haven't put much thought into this and certainly haven't discussed it with top players. So, my opinion is pretty much "quick and dirty" and just my own.

So, if there are solid arguments as to why I should re-think this, I'd certainly be interested to see them since re-thinking is a form of learning / improving.

That said, although I have no empirical evidence, when I consider the balance between how many players have highly exploitable flaws plus how often those situations come up vs. how often and how much I'd actually be exploited if I play in a way that is theoretically more exploitable, my admittedly purely subjective feeling is that it's pretty hard to imagine GTO actually being more profitable at the micros.

Btw, can you point me toward anywhere I can see pro-GTO arguments?

All opinions are valid. I'm just saying that I've had quite a few discussions about this over the years with various top level pros.

Any ways, I appreciate your comment. I don't have any GTO post that I could point you towards off hand though.
 
John A

John A

Poker Zion Coach
Silver Level
Joined
Apr 12, 2012
Total posts
6,492
Awards
3
Chips
37
John --- I am a fairly accomplished amateur player (approaching $200K profit since 2002) and I am now primarily a HU NL Tourney specialist. My approach to the game in the beginning was not an intuitive one. Rather, it was very mathematically based (as I'm a reasonably bright guy with a Masters in Applied Stats). I started out playing limit hold'em and was very successful because I am patient/disciplined enough to use hand selection to my advantage and I had no trouble folding hands I opened if I missed and got played back at. After a while, though, I got bored with limit (even though I was making solid money playing it) and I ventured into NL play to mix things up.

As a math-based player coming from limit, I was (unsurprisingly) a rock. I fared well against calling stations and maniacs, but got exploited by good LAGs. In an effort to improve my game against LAGs, I decided to learn to play HU. Learning to be a successful HU player was great for my full-table NL game (six-max or 10) and reduced how often I was exploited. In fact, I enjoyed the action so much that HU NL tourney play became (and remains) my go-to game.

Having said all this, my poker "home base" has been and always will be math- based (as opposed to intuitive, tell-based, or exploitative). When I heard about GTO poker, I was VERY excited because I was looking for ways to further reduce the "exploitatability" of my play. So I devoured "The Mathematics of Poker" by Bill Chen and Jerrod Ankenman and "Expert Heads Up No Limit Hold'Em" by Will Tipton. I sometimes wonder why I bothered with "The Mathematics of Poker" because I had already come to the conclusion that you (and others) have --- that GTO poker is, IMO, only nominally useful. Despite my reservations, I read "The Mathematics of Poker" from cover to cover, being sure I understood the concepts and the math. When I finished it, I still felt exactly the same way about GTO poker that I had after reading Tipton's book (which I read first because the math is somewhat simpler than the math n Chen/Ankenman's book) --- GTO poker is theoretically interesting but has, IMO, limited practical value.

The reason that I think the value of GTO poker is so limited is as follows: the equity calculations are incredibly dependent on ACCURATELY ranging your opponent's hand --- a skill that is widely employed but, IMO, rarely executed well. While the statement above devalues the practical application of GTO poker, it also devalues one of the central tenets of your poker analysis tools (based on what I've seen from you here at CC) --- that of hand-ranging.

Please don't misunderstand what I'm saying, as I think that hand-ranging is a valuable and useful skill in many situations/contexts. Having said that, I also think that basing poker decisions (fold/call/raise) based on equity calculations that are based on hand-ranging is taking things too far for most players --- including the large majority of winning players. I say this for two reasons: 1. Few people can accurately perform equity calculations at the table (even after hard-core use of pokerstove or any other equity calculation tool); and, more importantly, 2. Few people's hand-range assignment skills are good enough that their equity calculations are particularly useful. Online players with good analytical skills and a significant amount of HUD experience have a shot at reasonable hand-ranging against players they have a lot of data on. In all other cases, though, I question most people's ability to accurately/effectively hand-range. There are clearly some people that can do it accurately, and those people don't need GTO play or equity calculators to make good decisions at the poker table. For everyone else, I question the value of hand-ranging.

Please do not interpret my statements above as an attack on you or your teaching methods. I'm certain that you're both a better player than I am and a better teacher than I am. Furthermore, based on what I've read from you, I have reason to believe that your hand-ranging skills are very good. As a result, GTO poker would likely help you against players you have little data on/are unfamiliar with. Once you have enough data on your opponents, though, I agree with the central theme in this thread --- that exploitative is much more effective/profitable than GTO play if you have properly identified your opponent's tendencies. Until that time, however, GTO play is a reasonable approach for those people with proven hand-ranging skills. For most everyone else, I think it's much ado about nothing.

Best of luck.

-HooDooKoo

P.S. I wrote the missive about my background to open this post so that you would understand that I'm a mathematically inclined player with an real math education, as I think that background is important in evaluating the usefulness of my post.

I'm not sure why I would take that as an attack. It sounds like we basically agree. I think you may be confusing hand ranging skills and only applying them to GTO theory, which I'm not.

I do disagree that equity and hand ranges are not practical. It's funny you bring this up though because my wife and I were talking about "thin slicing" the other day and it's applications to poker (and other fields). She's reading the book Blink, which I read many years ago, so she had her head in how decisions are processed and made.

It was shortly after reading this book awhile back that I started having the idea about an equity trainer, and the power of pattern recognition in thin slicing situations that aren't computable at the poker table. I came up with the Ace Poker Drills Equity Trainer for this, which I think was quite an ingenious idea (pats self on back lol). However, I think it's too much to try and explain to the average Joe and why it will actually help you in game time decisions (it was difficult marketing the product, and I thought it might be so). It's had some success, but not as much as it should imho. :)

Any ways, my point is that most top level pros can put people on hand ranges pretty accurately on the fly. The better ones can understand their approximate equity in a given situation (this takes a lot of practice of course). Not sure if I need to state this, but I'm in no way saying playing high level poker is limited to excelling in these areas. For top online pros though, it carries a lot of weight.

In general I'm saying that these skills will get you a lot further than trying to apply GTO theory because 1) GTO will not translate into the highest EV for small stakes and under games, and most mid stake games, 2) You'd get more benefit from studying other skills before even worrying about GTO.

Just my opinion though.
 
DrazaFFT

DrazaFFT

public static void
Bronze Level
Joined
Mar 9, 2013
Total posts
6,188
Chips
0
This is actually a great tread that can give inputs to us beginners how two successful players have different approach, i have to say that because im still playing without hud that i need to play GTO and imo reading about game theory and learning deeper into the game really gives satisfaction itself beside improving , im currently reading sklansky and really enjoying doing it, i see that it wouldn't change my game much but im loving it...
I have read Jonh A book and after it i realize even more how playing without hud and without thinking about ranges can limit your profit and moving up stakes and can make your decisions difficult in some position that if you knew ranges you could have exploit it for more profit. Am i right btw? You can't think think about ranges if you don't have stats on your opponent right?

And another thing, didt want to PM you John, you are probably covered wit PMs, thank you very much for the opportunity to download and read your book, the part of taking notes is what i grasped immediately and start using it, and knowing that author is right there and having study group tread is great i still cant involve in it because playing without hud limits my ability to exploit the most what your book and tread can offer, any way without making this to long, THAN YOU!!!
 
R

rumsey182

Rock Star
Silver Level
Joined
Sep 27, 2013
Total posts
432
Chips
0
GTO assumes rationality, ie that your opponents are rational, ie that they play optimally. That's a pretty huge assumption to make at any stakes and you would be better served with a baseline exploitive strategy v the representative agent, but as you move up in stakes the rep strategy converges to GTO. Thus GTO becomes your best response.
this is wrong, it doesn't assume rationality bc then it could be exploited by being irrational

exploitative play generally does bc your assuming someone unknown plays like most people and or guessing your reads and assumptions are right
 
B

baudib1

Legend
Silver Level
Joined
Dec 2, 2008
Total posts
6,635
Chips
0
GTO had nothing to do with ranging your opponents accurately. In fact, it is closer to the opposite: knowing your own range and this playing in a manner that is indifferent to your opponent's action.
 
C

cotta777

Visionary
Silver Level
Joined
Dec 27, 2012
Total posts
868
Chips
0
You should play GTO when you dont have enough good information on opponents, and exploitable situations will come naturally against bad players playing to many hands easily dominated. so you will naturally make profit to some extent there.

My approach is to play solid while im running good and picking up playable hands getting good spots, I only bluff when I have too and I take into account my image and track my hands per orbit to ensure Im not over active at the table, this allows me bluff more effectively get credit for my raises,

Alternitively I might play over active deliberately if the players aren't inteligent enough to notice I switch gears, this is wear usually I can stack my opponent in a cash game if they are donkish
 
Top