Is Advanced Game Theory Giving HoldEm a Resurgence? (1 Viewer)

justsomedude

Straight Flush
Joined
Nov 10, 2014
Messages
9,868
Reaction score
17,629
Location
The Black Hills
It seems like all I've heard from a lot of the top players, successful high stakes players, and regular tournament winners/placers, is that they attribute their recent success to taking a new approach to holdem rooted in advanced "game theory" and returning to "the lab." This seems to be bringing some buzz back to holdem, which (at least in my opinion) had fallen behind in recent years to some of the more "fun" mixed games.

Is this "heavy study" approach and "new" strategy responsible for the semi-resurgence of holdem?
Are people reacting to it, trying to test their skills in the now "more advanced" holdem waters?
Is this why some of the well known staples of holdem - who disappeared for a few years - are back in full force on the holdem scene?
Is it really changing the way the game is played - and perhaps creating a "new era" for holdem?
Or are "game theory" and "lab" just buzzwords creating a bit of newfound hype to the same old game?

Curious to what the PCF crew thinks about all this...
 
Hmmmm...interesting question. First off, I read about game theory being applied to holdem 5 or 6 years ago, so IMHO, it is not new and not the driver of recent increased interest. I will also note that my perspective is likely swayed by the fact that all we play is holdem...no mixed games. So interest in holdem never slowed with my group.

As far as world wide interest, e.g. the WSOP Main Event, I think the increased participation has more to do with (1) the change in starting stacks and (2) the change in payouts. Over the past couple of years they moved from 25k to 50k starting stacks, and instead of paying the top 1000 players they paid 15% of the pool this year (1184 players).

So, while advanced game theory might be a big part of a number of players' strategies today, I'm not so sure it is a driver of incresed interest. Just my 2 cents (:
 
I also think the stronger economy has a lot to do with the recent resurgence in poker in general. People have more disposable income to spend now.

I too have read and studied game theory years ago, so to say it is something new if not accurate.

And as far as our group Hold’em is still not the most popular for cash games, PLO is by far. Tournaments, it is still Hold’em.
 
I’m curious about all this. I’ve seen it referred to as GTO or game theory optimal, right? I don’t know much about it, though I’m sure I should. I feel like game theory has always been a part of poker at some level - I think maybe stu ungar was one of the earlier pros to take advantage of it, though maybe not in the formal way it’s being studied now.
I guess I’m addressing your question specifically, I’d respond with this question - how popular is it really, among lot be players? I’ve assumed it was only popular among a small portion of the top pros. If that’s the case, I don’t see how it could be affecting the popularity of hold em, since the vast majority of Holdem players are recreational?
I dunno. I’ll be following with curiosity.
 
I know game theory is not new, but I can't help think there is some new strategy being used with how much everyone is talking about it. Maybe I should have been more clear on that in my OP...

Negreanu talked about all his time "in the lab" prior to the SHRB, and jabbed Doug Polk about it with his "more study is better" comment. And now Bonomo is openly talking about how he studied religiously prior to the OneDrop, and even spent his breaks "studying". This seems a bit different from the old "yeah - I work with a coach" type comments from Jacobsen and other players, from just a few years ago.

Maybe I'm just reading into these new buzzwords of "lab" and "study." But are the pros really applying some new poker strategy? Or is it just same old - same old poker?
 
Not to derail, because this is a great topic, but when I hear someone going into the lab, THiS is what comes into my brain...


My brain is a weird place.. Obv
 
Last edited:
I don’t even know what you’re talking about. Any good google links to game theory that you speak of?
 
Playing GTO is as the name says: theory. You still need to make a lot of guesstimates in live play to apply it as well as possible. If 2 computers were programmed to play each other perfectly it would show that they should come out almost perfectly even in money as it creates a zero sum

. It was proven to work in the Computer vs Pros heads up Limit challenge. In which the Computer trounced all of the pros.

The best/easiest application of it comes in tournament play when to push/fold hands as your stack gets shorter and what position you are in. Since you only have 1 decision to make. It gets much more complicated as you have to make decisions on multiple streets and if there are multiple opponents.

The basic premis of GTO is that no matter what your opponent does you aren’t making a mistake. But it isn’t always the most profitable way to play a hand since most opponents aren’t playing anywhere near optimal Poker themselves.

Often you will see discussions of playing GTO vs exploitative poker.
 
Generally speaking I think these are just buzzwords. The vast majority of players have no first hand knowledge of GTO, solvers, or game theory in general. Probably only a handful of high level pros spend much time "in the lab" working on this stuff.
 
Generally speaking I think these are just buzzwords. The vast majority of players have no first hand knowledge of GTO, solvers, or game theory in general. Probably only a handful of high level pros spend much time "in the lab" working on this stuff.
Exactly what I've been thinking. I'm not sure there are any books on the topic - I imagine whatever materials exist are locked away in german labs or possibly available via "pay for my videos" coaching.
It all seems very mysterious and unattainable super high level stuff - that's my impression anyway. But if we explore that assumption a little further, I'd think it would discourage interest in hold em, rather than increase it. If the game is nearly solved but that level of expertise is possesed by a few - why would I want to compete with them? Screw it, I'll play PLO8.
 
Exactly what I've been thinking. I'm not sure there are any books on the topic - I imagine whatever materials exist are locked away in german labs or possibly available via "pay for my videos" coaching.
It all seems very mysterious and unattainable super high level stuff - that's my impression anyway. But if we explore that assumption a little further, I'd think it would discourage interest in hold em, rather than increase it. If the game is nearly solved but that level of expertise is possesed by a few - why would I want to compete with them? Screw it, I'll play PLO8.

I was playing at Caesers back in June and the dealer and one player were talking about how close a spot was. One guy was all in and ended up losing. So after the hand the dealer has this other player look it up on his phone to see what his exact odds were and see if he made a good play.

I was really annoyed by this. My point being...what makes poker good is drinking, laughing, and gambling. When you have people looking up odds on their phone that is simply not good for the game. I am not flying to Vegas and playing in casino to have people sitting there crunching the odds on their phones.

Furthermore, to your point...I agree high level analysis is bad for the game. I have lost almost all interest in playing online because these guys all use software and analysis tools that I am just not willing to invest in (time or money).
 
GTO means Game Theory Optimal. There are fancy ways to describe it with regard to game conditions and strategies and who know who is playing what strategies with what cards, but at its simplest, if I am playing GTO and you're not, I make money.

Unfortunately, in an average game, GTO is quite useless. 99.9% of players do not play GTO or even know what it is, and at that point playing exploitative poker is most profitable. Exploitative poker is profiting from another player's mistakes.

So high level pros playing each other play GTO, and if you are better at it than the other guy, you make money. But when a non-GTO player comes along, you can still use GTO concepts but you are better off finding that player's weaknesses and exploiting them.

It's a dry read, and not too useful for play, and it helps to know terms like Bayesian, but The Mathematics of Poker by Chen and Ankenman is the best introductory book on the subject that I've found.
 
Furthermore, to your point...I agree high level analysis is bad for the game. I have lost almost all interest in playing online because these guys all use software and analysis tools that I am just not willing to invest in (time or money).
Right? Having a constantly updating stream of data, quantifying every decision that every person makes - that's not poker, that's nerdball, or some other game. And I suspect that this GTO grew out of all that data.
 
Exactly what I've been thinking. I'm not sure there are any books on the topic - I imagine whatever materials exist are locked away in german labs or possibly available via "pay for my videos" coaching.
It all seems very mysterious and unattainable super high level stuff - that's my impression anyway. But if we explore that assumption a little further, I'd think it would discourage interest in hold em, rather than increase it. If the game is nearly solved but that level of expertise is possesed by a few - why would I want to compete with them? Screw it, I'll play PLO8.

The funny thing is PLO8 is a more solved game than Hold’em!!! It is rare that a PLO8 game last more than a few hours as the players that know how to play crush the rest of the players so fast the games ends.

OFC became a solved game and had to change to be playable.

Hold’em is far from “solved”. Going in “the lab” is really all about building ranges for opponents and making the right decision with your hand based on those ranges. They are still estimates of both the ranges and the player in that spot. Both of which can be way off at times.
 
Right? Having a constantly updating stream of data, quantifying every decision that every person makes - that's not poker, that's nerdball, or some other game. And I suspect that this GTO grew out of all that data.

Years ago when I logged many hours online I did run a HUD and Poker Tracker which logs all your hands and it was fun. Yeah I agree it was nerd ball, it is not really the poker game we know and love. I like the data analysis side of it and I find all that really interesting. I am just not that smart so I don't get the benefit from it that some of these really bright kids do. I never really got my whole setup dialed in like some of the top players back in the day.

The good news is even if online poker dies there are always fun and juicy live games that can pop up anywhere, in LV or in your hometown. I don't think live poker is going anywhere.
 
Generally speaking I think these are just buzzwords. The vast majority of players have no first hand knowledge of GTO, solvers, or game theory in general. Probably only a handful of high level pros spend much time "in the lab" working on this stuff.
I think it is way more than a handful of pros. I myself have starting using ProPokerTools (online equity calculator) to build more intuition about equities and PioSOLVER (GTO solver) to get exposure to optimal play in certain spots. The latter especially is totally irrelevant for the games that I play in but I enjoy the mental exercise nonetheless. Another GTO adjustment that more players are making is the use of blockers to control bluff-calling frequencies.

Good starting resource would certainly be Upswing. Bart Hanson of CrushLikePoker has also been incorporating GTO elements into his strategy podcasts, although he still is way more opponent and stakes-based exploitative. The CLP Top Section podcast is apparently very GTO-heavy but my subscription level is not high enough to listen to it :)
 
Furthermore, to your point...I agree high level analysis is bad for the game. I have lost almost all interest in playing online because these guys all use software and analysis tools that I am just not willing to invest in (time or money).
Right? Having a constantly updating stream of data, quantifying every decision that every person makes - that's not poker, that's nerdball, or some other game. And I suspect that this GTO grew out of all that data.
The funny thing is PLO8 is a more solved game than Hold’em!!! It is rare that a PLO8 game last more than a few hours as the players that know how to play crush the rest of the players so fast the games ends.
Those guys don’t show up at the $2/$4 table at my cardroom!
 
My game is waaaaaaaaay to recreational for GTO to work. Hell, basic weak/strong tells don't work because people think their 10 high flush is the nuts even with a paired board.

Phrases like ”chip odds” get thrown around in my donkey corral.
 
My game is waaaaaaaaay to recreational for GTO to work. Hell, basic weak/strong tells don't work because people think their 10 high flush is the nuts even with a paired board.

Phrases like ”chip odds” get thrown around in my donkey corral.

Perfect. This is how poker should be. I love it when people throw around words like "pot odds". They know just enough to be dangerous but don't fully grasp these concepts. Its more like well I kinda wanna call and there are several other people in...ok I call.

"Donkey Corral" is a great name for a custom set/poker room...lets go!!!
 
Those guys don’t show up at the $2/$4 table at my cardroom!

$2/4 PLO8 or LIMIT O8?

Limit O8 is just a bunch crusty old guys bitching about life waiting for their turn to die.

I’ve never seen PLO8 with $2/4 blinds, usually is it the same as PLO, $1/2, $1/3 or $2/5.
 
What ste the best books on GTO?
Generally speaking I think these are just buzzwords. The vast majority of players have no first hand knowledge of GTO, solvers, or game theory in general. Probably only a handful of high level pros spend much time "in the lab" working on this stuff.

Matthew Janda’s two books; Applications of no limit holdem and No limit holdem for advanced players.
First one is steeped in GTO theories and the second book follows up with the first with solvers and AI (pokersnowie).
 
Generally speaking I think these are just buzzwords. The vast majority of players have no first hand knowledge of GTO, solvers, or game theory in general. Probably only a handful of high level pros spend much time "in the lab" working on this stuff.
Exactly what I've been thinking. I'm not sure there are any books on the topic - I imagine whatever materials exist are locked away in german labs or possibly available via "pay for my videos" coaching.
It all seems very mysterious and unattainable super high level stuff - that's my impression anyway. But if we explore that assumption a little further, I'd think it would discourage interest in hold em, rather than increase it. If the game is nearly solved but that level of expertise is possesed by a few - why would I want to compete with them? Screw it, I'll play PLO8.

As @navels pointed out, there are a lot more people studying GTO than just the high level pro's. I'd wager at least 85% of anyone playing 2/5+ for a living is studying GTO concepts, even if they don't necessarily study in that context.

For example, hand ranges and balancing are GTO concepts that are widely discussed, studied, and implemented. I think that the best strategy is to implement GTO concepts into your game, but also understand when to play exploitative poker to maximize your profit.
 
Perfect. This is how poker should be. I love it when people throw around words like "pot odds". They know just enough to be dangerous but don't fully grasp these concepts. Its more like well I kinda wanna call and there are several other people in...ok I call.

"Donkey Corral" is a great name for a custom set/poker room...lets go!!!
I know right! I might mock up design this weekend. I need a cash set. But my wife is complaining she ”needs” a new bathroom.
 
Hopefully I can chime in with some insight.

Over the last few years GTO has become increasingly prevalent in poker, especially online and at mid/high stakes.

Essentially, poker players are using software (Equilab and PIOsolver are the big two) to run complex simulations of poker hands in order to determine the amount of equity a player has in certain spots when facing a certain range of hands.

Sound complex? Basically, poker players use a computer program to “solve” a poker hand, providing invaluable, accurate information about the correct way to play a hand.

Why is this a big deal? Due to these programs we are learning that some of our previously held beliefs about poker strategy are wrong. Very wrong. Some situations that would have been an automatic fold for most of us, should instead be a call, or even a 3-bet.

Ok, but how is that GTO? Let me use a hand as an example.

Villain is on the BU with: XX
Hero is UTG: with Ah10h

Board: Ac5cQs (pot $200)

Villain goes all in for $200

As it stands, our pots odds are 2:1 or 33%

If villain’s range includes 66% value hands (hands that beat ours) and 33% bluffs, he is said to be playing GTO, as in his strategy is indifferent to our actions. He wins $200 if we always call, he wins $200 if we always fold.

If we only call when we have 33% equity in the hand, then we have an EV of 0. Nobody plays poker to have an expected value of 0, so we adjust our ratios of bluffs and value hands according to our opponents. This can be very profitable, but you need a considerable history with the player or a strategy based on strong evidence of their play.

Most players do this intuitively without realizing it:

“Mike likes to keep players honest on the river, so I need to have the goods if I bet on the river”

“John always plays with scared money, I should be bluffing him more often”

GTO is a good guideline for poker players, especially as a tool for teaching players to have balanced ranges. Everyone knows or plays with a self proclaimed “TAG” that will only play the super premiums. Mr. TAG has an unbalanced range of hands, and as a result has become extremely easy to exploit. If we know Mr. TAG will only play 3% of hands, we can simply hammer him on unfavorable flops, fold to overt aggression because he does not have enough bluffs in his range, and 3 bet out of position to diminish his positional advantage.

What does this have to do with the state of poker? IMO, it’s bad. The game has only gotten tougher and tougher. Information and strategy that took years and years of constant play can be learned in a month of diligent studying. Online coaching resources are getting better and better. NLHE is getting closer and closer to becoming a solved game. People like me spout off unintelligible nonsense about poker strategy on web forums and at the table, scaring off potential customers. (For the record, I NEVER talk about poker strategy when playing with others)

Fortunately, I don’t think live play will ever die off completely, but I definitely think that PLO or short deck hold em might take over as the most popular game in the future. One because the higher variance will allow less skilled players to win more frequently, and two PLO is far more complex than NLHE and the quality of coaching/resources have not caught up to that of hold em.

Alright, Thanks for reading.

Edit: Added a sentence, fixed some typos, corrected numbers about value hands/bluffs.
 
Last edited:
After reading this thread, I now know why I suck at poker.
Don't feel bad. The only location in a casino where I ever walk away a winner is the machine that requires my bank card and a 4 digit pin. and even that bastard machine denies me more often than not.

But I'll always have my other chips at home.
 
All this does take away from the charm of poker but it is the reality of the time we live in I guess.
Haven’t they been effectively doing this in sports for years? I remember reading an article over 10 years ago about all the statistical analysis that goes on to prepare for a game in college and pro football. Things like analysis of what types of plays a coach has called on every 3rd and 2 on the 27 yards line etc
 

Create an account or login to comment

You must be a member in order to leave a comment

Create account

Create an account and join our community. It's easy!

Log in

Already have an account? Log in here.

Back
Top Bottom