Maybe Cali meant that the ev of the player that is doing the exploiting will be closer to the ev of the equilibrium strategy, which I think is true.

To attempt to exploit and arrive at NE would mean you failed, or that your opponent is just as good as you are? GTO would be -EV shrinking to zero as equilibrium is established. Exploit would be +/- EV when one exploiter is better than the other. "Better exploit" is always relative and "perfect GTO" is always indifferent. I dont see NE as a goal of exploit.


Sent from my iPhone using Tapatalk

The latter is true and the former is false.

Perhaps the confusion is rooted in the old view of gto being unexploitable. The best we can do within the confines of human language to describe what happens when an equilibrium strategy pair plays poker is to say that both strategies are maximally exploiting each other. Of course, this also may cause some confusion because some will say "but gto doesn't adjust thus it can't exploit." However, realizing that words put together to form sentences in human language is limited in the ability to describe the universe is of most importance. Maybe we will never be able to use words to completely correctly describe what happens when gto faces off against itself.

Despite all of that mumbo jumbo, I still think that the title of the thread here:
is just plain wrong.

Also.

The idea that gto is somehow a defensive strategy is misguiding at best. If this were true, then gto would not profit vs non equilibrium strategies. This may hold true in some games like rock paper scissors where gto breaks even vs any strategy. However, poker is not rock paper scissors. In real poker, when an equilibrium strategy faces a non equilibrium strategy, the non equilibrium strategy will have an expectation of less than or equal to zero in the long run. This means that the non equilibrium strategy loses ev and the equilibrium strategy gains ev. While it's true that this gain is less than the ev that may be captured by an exploitive strategy, this doesn't mean that gto is defensive.

Categorically disagree with lots of what you just said. But we can still be friends.


Sent from my iPhone using Tapatalk

Ok then, if this is true:

Then we can just bluffcatch at frequency based on pot odds and do whatever we want with non bluffcatching hands and we'll be playing gto. However this cannot possibly be the case.

----

If this is not true:

Then how do you suggest we define what happens when two equilibrium strategies face each other?

If it's two player zero sum then the players will always have equal distance from their NE EVs. For non zero sum games you can easily come up with examples where the exploiting player is further away from the NE ev.

Since the whole concept of GTO is pretty advanced, maybe we should abandon analogy altogether.

When I suggest GTO is a shield, that includes all the spots where exploit fails and allows autoprofit, for instance. The GTO will put extreme pressure on the exploit player in that instance, and that is not defensive. However, to adopt the strategy, overall, is a preemptive countermeasure to exploit. And exploit is how humans play poker, so to play equilibrium is to simply not play "poker" at all but to play a different game, at the same table with the same cards and chips as the "poker" players.

Thus, multiple GTO players are simply going through the steps of the GTO game, theoretically solved. Any exploit players would seek out other exploit players and play "poker".

To play equilibrium versus equilibrium would possibly serve to add decimal places to the solution. So it would then be nothing more than math for maths sake.


Sent from my iPhone using Tapatalk

Maybe to seek the purpose behind GTO can reveal the practical application of it, to human beings. Since we are human, we like to play poker, and we want to be successful at same. Why GTO?

Hmmm.

Well, practical applications at nose bleed stakes come to the imagination.
Say, you are king of the hill so to speak. You welcome anyone to play heads up at the highest stakes spread on your site, or in your casino.
What kinds of players will sit down and play, and how should you play them, not knowing anything about them?
How about the classic whale, a wealthy person who fancies poker and has played for very high stakes, often.
How about the young gun, who has read everything and understands it pretty well. Maybe he is bankrolled by someone else, who knows?
How about the best player in the next casino, come to test you out?

The point of GTO here is it will not matter. Sure if our hero knew all the backstory of each player and had a few thousand hands against each, he may or may not choose to exploit. But GTO provides him the option to not know the opponent, and further force the opponent to play at a very high level, to even be a threat, in the long run.

Also, same nose bleeds, approximate GTO is how humans can minimize losses to each other while waiting for, and then getting to know a new player to sit down and try to compete at the highest level.

Neither of these situations requires perfect GTO by humans, but allows humans to use GTO against both known and unknown opponents.

This does not guarantee these heroes a profit, it guarantees the heroes a minimal loss against unknown players at very high stakes. Once the unknown players reveal weaknesses, humans will exploit by altering or abandoning GTO until GTO might be needed again, against new unknowns.

So why would a low stakes scrublord even care about GTO?

Would be good practice to at least try to find some threshold hands/situations, maybe practice watching for autoprofit, maybe start to defend blinds universally for any stake instead of what the hud has to say about some rando players steal frequency.



Sent from my iPhone using Tapatalk

obviously gto and exploitative plays are not the same thing by definition. But good exploitative play should always be +ev than GTO, because it relies on additional information allowing you to deviate or not from GTO.

But what you are doing intuitively MAY be very good. If you think as an agressor & slowplay your reads get you higher EV then good for you.

But I guess you relying on GTO for bluffcatching may be very good too if you think your reads dont allow an exploitative play.

I think this is right on.

Do we think that a mix of streets playing exploit but wind up with a "GTO call" on the river is even +EV?

What I am thinking is that humans are best at starting the hand GTO and finishing as GTO, with room for mistakes (and EV) in the middle.


Sent from my iPhone using Tapatalk

The more I think about it, the more convinced I am that mixing strats, IE trying to arrive at MDF against river bluffs, having played the previous streets as exploit is -EV.

Decision 1. Is my opponent seeking pure NE?

If yes, then why the fk was I exploiting him.

If no, then why would GTO have MORE information to add on the river than I have gathered so far.

Decision 2. Does my opponent perceive me as narrow or wider range. Does he normally view me as value heavy or value light? Decide.

Decision 3. Put decision 2 info into the levels, but try not to overthink it, lol.

Decision 4. Sum up all these more accurately than villain, get +EV, fail and lose EV.


Sent from my iPhone using Tapatalk

I think Matt Berkey is an idiot. Or to put it nicely, very misguided.

He makes humongous novice errors that shouldn't be seeing the light of day at the stakes he plays.

He has no clue what he is talking about with gto.

GTO is very special, if you actually know true GTO in most, if not all, situations, you can beat 5/10+.

But I think bluff catching with a perfect GTO strategy is 0EV (same as folding), because the optimal bluff catching strategy is to prevent opponent from making successful bluffs.

Well, if both players are perfect, sure.

Behaving randomly, and playing GTO are not always the same thing. Balancing the value with the bluff, and tailoring this to your opponent, is what I believe most people are talking about, on the river.

If tailored to an imbalanced opponent, you are trusting that opponent to be consistent, which is a proven fact that humans are consistent and arrive to similar decisions repetitively, especially when under emotional stress or fatigue.

If you are facing a new opponent, who is well rested and emotionally centered, then you would best be served to seek pure GTO frequencies, knowing that any profit there comes from the opponent being farther from Nash E, basically less accurate GTO.


Sent from my iPhone using Tapatalk

MATT BERKEY

............CASHES...EARNINGS
WSOP.... 26.... $1,024,147
CIRCUIT... 1...... $179,119

Hendon Mob has him at over $3.1 million. Some idiot!

Jamie Gold
Total Live Earnings
$12,586,358


must be very smart if we use that as some kind of measurement

you should know better...

wp

True GTO-play in the literal sense is far too complex for humans and it is basically impossible to implement. Most people (dare i say all people???) will end up making more mistakes by trying to "play perfect GTO".

Really, you're just taking GTO principles that you know/think are good and applying them to your exploitative strategy.

Another thing to consider about purely exploitative play - If you see opportunities to exploit your opponents game, how do you know those opportunities will be available in the future? Any time you deviate from your best interpretation of a GTO strategy to exploit your opponent, you open yourself up to counter-exploitation. This may not matter against truly weak players, but if you're playing against a true unknown, it could definitely matter if he/she also has a strong understanding of the game.

Very well said!

Every year the top HUNLHE players beat the best bots, and according to the players, it is not really close.

I suspect that is because imperfect GTO is just a long walk that gets you back to bad poker. To seek equality from a human being is a very difficult task, since we are the most selfish creation in the universe.


Sent from my iPhone using Tapatalk

You are truly an amusing and amazing poster!

o7 !


Sent from my iPhone using Tapatalk

This thread is a trainwreck, obv.

Hmm. Obviously I missed that, earlier this year. Thanks!

Two things though,

The bot was not GTO, it was learning and adapting to each pro. Awesome!!!!

Each hand the stacks were reset to 200BB.
So, while that is not a violation of poker rules, it is not 100 percent poker, just yet. But almost!

I wonder if the pro's will adjust and try to reclaim the title next year. I doubt it though.

Cheers!


Sent from my iPhone using Tapatalk

No one claimed it was "GTO". It was learning but it was not adapting to the pros.

...
How much percent would you say it's poker? I think it's between 78-83% but can't decide for sure.