Quote:
Originally Posted by angeles
chess is actually super super simple compared to nlh (200bb deep or otherwise)
This is false. Chess is far more complex than poker. There have been threads on this topic on this forum in the past. I predicted years ago (with an old screenname I don't have access to) that computers would easily beat all the top human players in poker.
Quote:
Originally Posted by ShallowMind
Historically for unsolved games, human + computer is much better than computer or human. At least that's the current state of chess. I'd guess it's the same for NLHE, live computer assistance would give humans a definite advantage.
Source: https://www.ted.com/talks/shyam_sank...er_cooperation
There would be no advantage for "human + Libratus 1" vs. Libratus 2 in your example, because if libratus is playing GTO poker, there is no improvement that a human could provide. Take a look at AlphaGo, eg. The computer would give the probability that a human would make any particular move. One of the startling moves it played, it estimated that a human would play 1 in 10,000 times. So the computer is considering everything that a human would consider, and then makes its decision.
Quote:
Sorry, I understood very well what you said. But you would probably agree that using processed data (such HUDs) gives significant advantage to a player in human vs human play. In the case we discuss the bot had this info by default and humans did not.
Let's compare it with chess. When Kasparov played with Deep Blue he was allowed to use a computer to check debut info and probably some other stuff. Everybody was agree that Kasparov cannot and should not compete wit Deep Blue in memory capacity, e.g. holding database of debut option. They competed in the actual playing chess.
Similar to our case everybody would agree that even average PC could process and hold information much better than human brain. But this differs from the game itself using this processed information.
Otherwise - yes, it seems obviously now that the bot have huge advantage over the humans. I'm not trying to accuse it of cheating.
This is false. Kasparov did not have any access to a computer. Deep Blue not only had every one of Kasparov's games stored in its database, but also millions of games played by grandmasters; probably all significant games that had ever been played. That is why Kasparov's victory in 1996 was so extraordinary. Not only did Deep Blue have a huge database, but could calculate millions of moves ahead, while Kasparov only relied on his mind.
Between 1996 and 1997 when the computer won, IBM hired a chess grandmaster to work full-time with them. This was a huge factor in Deep Blue becoming much stronger. Besides that, all of the database is from prior human games, so there was significant human-derived chess knowledge built into its system. As far as I know, no poker pro was needed to assist the Libratus team.
But this match seemed very biased in favor of Libratus. Using Doyle's game rules (stacks reset after each hand) is not like playing a cash game and not like playing in a tournament either. So the poker pros are not accustomed to it. The strategy is obviously very different, and much more complex, to play tournament style vs. Doyle's rules.
Pots were split when both sides were all-in, depending on the probability of each side winning. This is not a standard poker rule anywhere. So basically they have "invented" a very specific form of poker that their computer is designed to play, and are subjecting the humans to it. It should have been the other way around. Knowing more about this match, I don't consider it as significant an achievement as was touted. It's a bit surprising to me that the players not only agreed to these rules, but also the number of hands and hours that they are playing. The computer doesn't get tired or go on tilt, but it's a huge human factor.
Are there any computers that have demonstrated playing high-level poker with standard rules?