Ok, I definitely accept what you are saying, and I do believe that theoretically there should be some (very large) discrete set of actions or responses that in aggregate would make up a gto strategy for no-limit.

I see that I was slightly off-base in how I was thinking about this. I read about 10% of the thread, and there seemed to be a number of posts that implied that this kind of strategy already existed, and would/could be used to essentially get behind the game. Like, "I'm gto so even if the opponent knows exactly what I'm doing I'm 100% unexploitable" and that sort of thing.

Anyway, thanks very much for the enlightening information. I will try to look into this further, and perhaps finally force myself to buy the Ankmenchen book, and hopefully to read it.

In terms of the overall thread, then, I guess we can say that (many of) the top players are doubtless informed by gto concepts, but that none of them is actually playing a gto "style" or "system."

When I read everyones posts its just text but when I read Ike's posts I hear his voice read it out loud in my head.

Excellent thread, thanks Mike and Adam for pointing this out in the 2+2 podcast!

and Ike,.. here's a suggestion for what to do with the pricemoney if durrrr takes you up on the bot bet:

LoL if there was a propbet with Durrr to play against a gto bot, Durrr would win in such a devastating way that if the game were being played with real cash the opponent would probably have to call 911.

Jesus. Think about it people. You know it's a gto bot and you know its strategy. Within like 100 hands its strategy would be figured out and totally exploitable.

I mean, is somebody actually saying that there is a gto heads up NLHE strategy that NO strategy can effectively counter? That the best possible outcome with it is a draw? Even if it is completely automated and insentient and can't adjust itself?

Really? is somebody saying that

The bet is for LHE

"I mean, is somebody actually saying that there is a gto heads up NLHE strategy that NO strategy can effectively counter? That the best possible outcome with it is a draw? Even if it is completely automated and insentient and can't adjust itself? "

Yes such a strategy exists but nobody knows what it is.

I don't think you understand what gto means.

You just figured this out?

Errrrr, when I say "exists" I mean it in the Aristotelian sense; in other words, is it manifest, with extension and identity.

I don't mean it in the sense of existing in some idealistic or theoretical sense, out there in the ether and waiting for us to discover it.

I mean does it exist and can some joker put it into a bot and beat Tom Dwan with it !! And I would take that bet that Dwan would piss all over what any body got right now

otherwise, why don't we just argue about whether Han Solo could beat Captain Kirk in a warp 3 space race

in the sense that there is actually a strategy already calculated for deep stacked nlhe that would crush durrr, then no, nobody is saying that. in the sense that it theoretically exists, then yes, everybody with a clue is saying that. the bet that was proposed was for flhe, and for that game there absolutely exists an already solved, completed automated and insentient strategy that cannot adjust itself that would crush durrr. everybody with a clue is saying that

Errr, thanks for the clue I suppose.

It seems like people in this thread are kind of seamlessly switching between assertions about limit and assertions about no limit, and from theoretical ideas of gto to statements like "I'm gto, bitches, and even if you know my strategy it doesn't matter."

Ok, I guess I don't know enough about advanced limit holdem strategy to troll *errrrr* make an educated comment about what you have said. However, what about the prop bet? Are you saying that nobody wants to challenge it with a prop bet? Or what do the really top limit players say about all this, do they concede that the gto bot is the best

There were jokers with an exact bot in mind offering to put of 7 figures against Tom. He was the one who offered LHE as a format. Srsbzness LHE players were excited to take his action.

You are one of the ones confused. The thread started with Tom offering LHE HU action vs. "gto bot". A bot was suggested and negotiations started. The top LHE players were on the side of the bot. Again, these guys were offering to put up millions that TD couldn't beat the bot. Whether that's a statement of the strength of current bots or their belief that Tom isn't a strong HULHE player is unknown. Nobody offered a top HU LHE player vs. bot bet.

Here's Tom offering to play a LHE bot.

Here's Ike saying "heck, yeah"

Durrrr obviously plays poker in the 5th dimension. A 3rd dimensional GTO Bot couldn't even comprehend what kind of strategies durrrr would be playing.

If we were to make the game pot limit hold em and assume 100BB starting stacks for each player and then further restrict the action when betting to say as an example:

1/4 pot bet
1/2 pot bet
3/4 pot bet
pot bet

Would this make it any more feasible to solve the game using current computing power?

Although this would not be directly applicable to actual games, we could maybe get an idea / extract some general ideas about how to play a balanced style?

It would make it a lot easier. Something like this has been and will certainly continue to be part of how the problem is approached in the near future, but even with these simplifications we're a long way away from approximating GTO play in HUNLHE. Even in HULHE we currently have to do things like pretend that you're limited to 1 bet and 1 raise per street and/or betting ends after the turn and/or using a modified deck of less than 52 cards to get very close to an approximation of the nash equilibrium.

Do we actually know the exact process to undertake to solve the game? Is it not a case of just setting the program to run and waiting for it to finish executing? Is there anyone out there trying this? I guess storage space is a problem as some-one else said earlier on. Using today's biggest computers assuming storage space is not an issue, how long would it take to solve?

There's been a lot written about this. If you're interested, the stuff on this page is a good place to start: http://poker.cs.ualberta.ca/publications.html

"The computation of the first complete approximations of game-theoretic optimal strategies for full-scale poker is addressed. Several abstraction techniques are combined to represent the game of 2-player Texas Hold'em, having size O(10^18), using closely related models each having size O(10^7). Despite the reduction in size by a factor of 100 billion, the resulting models retain the key properties and structure of the real game. Linear programming solutions to the abstracted game are used to create substantially improved poker-playing programs, able to defeat strong human players and be competitive against world-class opponents."

From the link

This is completely wrong man. Since we haven't actually solved NLHE yet we can't say a GTO strategy is +EV vs every other strategy. It's only guaranteed to be neutral EV. (realistically it's not even guaranteed to be neutral EV because of rake)

GTO strategies will be neutral EV against other GTO strategies, and neutral or positive EV against all other strategies (it's possible for specific non-GTO strategies to still be neutral EV vs. GTO strategy, but -EV vs. other strategies). And yes, all of these calculations are pre-rake; depending on the rake, GTO strategy could be -EV vs. a number of strategies (which would themselves also be -EV, as only the rake is making money).

So do we have any developments from durrr or has he still vanished from the thread/the bet?

GTO must be the strategy that phil ivey plays

No he plays the exact opposite, exploitative.