|
|
#41
07-25-2007, 03:19 PM
|
|||
|
|||
Re: What games are bots unable to beat?
[ QUOTE ]
[ QUOTE ] I challenge any bot to a $50,000 heads up 1,000 BB NLHE freezeout [/ QUOTE ] The Unabomber is doing this already against the bots that the University of Alberta team created. [/ QUOTE ] No. The recent UofA event is not the same type of challenge stated by the OP. For it to be the same, one of the sides invovled would need to lose money. There was zero downside for the pro players in Vancouver other than time spent. It is NOT real poker until all players risk losing money. 
|
|
#42
07-25-2007, 04:35 PM
|
|||
|
|||
|
Re: What games are bots unable to beat?
[ QUOTE ]
[ QUOTE ] Quote: Quote: I challenge any bot to a $50,000 heads up 1,000 BB NLHE freezeout [ QUOTE ] The Unabomber is doing this already against the bots that the University of Alberta team created. [/ QUOTE ] [/ QUOTE ] No. The recent UofA event is not the same type of challenge stated by the OP. For it to be the same, one of the sides invovled would need to lose money. There was zero downside for the pro players in Vancouver other than time spent. It is NOT real poker until all players risk losing money. [/ QUOTE ] Would this actually be the case against a computer opponent? I'm not sure. Money changes things for the human players, but for the computer, it's just numbers anyway. In that sense, playing for significant real money puts the human players at something of a disadvantage in most cases (some players play better under those conditions, but not many). I agree to some extent that without some sense of risk, poker loses some essential elements. So I guess the real question is, will a computer ever be able to play 'real' poker, in the sense that you describe?
|
|
#43
07-25-2007, 04:55 PM
|
|||
|
|||
|
Re: What games are bots unable to beat?
Bot schmooot!
|
|
#44
07-25-2007, 05:55 PM
|
|||
|
|||
|
Re: What games are bots unable to beat?
[ QUOTE ]
[ QUOTE ] [ QUOTE ] Quote: Quote: I challenge any bot to a $50,000 heads up 1,000 BB NLHE freezeout [ QUOTE ] The Unabomber is doing this already against the bots that the University of Alberta team created. [/ QUOTE ] [/ QUOTE ] No. The recent UofA event is not the same type of challenge stated by the OP. For it to be the same, one of the sides invovled would need to lose money. There was zero downside for the pro players in Vancouver other than time spent. It is NOT real poker until all players risk losing money. [/ QUOTE ] Would this actually be the case against a computer opponent? I'm not sure. Money changes things for the human players, but for the computer, it's just numbers anyway. In that sense, playing for significant real money puts the human players at something of a disadvantage in most cases (some players play better under those conditions, but not many). I agree to some extent that without some sense of risk, poker loses some essential elements. So I guess the real question is, will a computer ever be able to play 'real' poker, in the sense that you describe? [/ QUOTE ] I'm not exactly sure what you mean by "case against a computer opponent"; maybe you can clarify. I definitely agree that the money factor is the great equalizer when all things are considered. I don't see the human player at a disadvantage any more than the human that developed the robot. In each case you have a human risking money (I don't think robots can do this on their own [img]/images/graemlins/laugh.gif[/img]). If anything I'd say the human programmer has voluntarily chosen to hamstring himself by limiting intelligence to only that which can be programmed; the previous statement probably reflects my personal preference if I were involved in such an event; however, I'm willing to admit that I could be wrong and that something might be gained from strong poker ai. If I were forced in to wagering my life on the proposition of whether or not bots are already doing this everyday in the online game then I would take the "yes-they-are" side of the prop bet with little or no fear for my life. It's interesting to note that this proposition (even if for small money wagers) is not resolvable in the online game because there is no way to prove that a robot was not assisted by it's human during play (in which case the prop. bet is invalid). The flip side of this is that you cannot prove that a human was not computer assisted at some point during play. The only way to resolve this type of 'real poker' thing is to get everybody into a live game so everybody can see that the robots were not aided by humans and the humans were not aided by software tools. This is one of the major factors being emphasized by the PBWC guys. They are placing a very high priority on real world tournament conditions so much so that their guiding design principle (according to some of their threads) has been one simple question: "What if the WSOP allowed bot players?" The PBWC event is apparently their answer to this question.
|
|
#45
07-26-2007, 03:19 PM
|
|||
|
|||
|
Re: What games are bots unable to beat?
[ QUOTE ]
I'm not exactly sure what you mean by "case against a computer opponent"; maybe you can clarify. [/ QUOTE ] The statement was that 'It is NOT real poker until all players risk losing money.' In responding 'Would this actually be the case against a computer opponent?', I was asking if it matters to the computer whether it was real money or not. You make an interesting point when you say: [ QUOTE ] I definitely agree that the money factor is the great equalizer when all things are considered. I don't see the human player at a disadvantage any more than the human that developed the robot. In each case you have a human risking money (I don't think robots can do this on their own ). [/ QUOTE ] In this case, you consider the computer to be a tool for the programmer to win money. As the technology stands, this is probably the right way of seeing things; I'm not sure it addresses my point, however. The programmer isn't exactly 'playing poker', any more than the guys that stake poker pros in major tournaments are. Like them, the programmer is placing one bet, sort of a prop bet, that the computer will win more money than its opponents. I think the point that RIIT was trying to make was that the prospect of loss is an essential element of poker (please correct me if I am wrong), and that since the UofA challenge had no prospect of loss on the part of any of the players, it wasn't 'real poker'. I think I agree with him on that, but if this is the case, can a computer ever play 'real poker' if the prospect of loss has no meaning for it?
|
|
#46
07-26-2007, 03:47 PM
|
|||
|
|||
|
What games are bots unable to beat?
[ QUOTE ]
What games are bots unable to beat? [/ QUOTE ] Live cash games (limit or NL). I have yet to see C-3P0 at the casino, although im sure this is a thing of the near future. Also, parcheesi. I will play any bot at parcheesi for any amound of money.
|
|
#47
07-27-2007, 08:44 AM
|
|||
|
|||
|
Re: What games are bots unable to beat?
[ QUOTE ]
[ QUOTE ] I'm not exactly sure what you mean by "case against a computer opponent"; maybe you can clarify. [/ QUOTE ] The statement was that 'It is NOT real poker until all players risk losing money.' In responding 'Would this actually be the case against a computer opponent?', I was asking if it matters to the computer whether it was real money or not. You make an interesting point when you say: [ QUOTE ] I definitely agree that the money factor is the great equalizer when all things are considered. I don't see the human player at a disadvantage any more than the human that developed the robot. In each case you have a human risking money (I don't think robots can do this on their own ). [/ QUOTE ] In this case, you consider the computer to be a tool for the programmer to win money. As the technology stands, this is probably the right way of seeing things; I'm not sure it addresses my point, however. The programmer isn't exactly 'playing poker', any more than the guys that stake poker pros in major tournaments are. Like them, the programmer is placing one bet, sort of a prop bet, that the computer will win more money than its opponents. I think the point that RIIT was trying to make was that the prospect of loss is an essential element of poker (please correct me if I am wrong), and that since the UofA challenge had no prospect of loss on the part of any of the players, it wasn't 'real poker'. I think I agree with him on that, but if this is the case, can a computer ever play 'real poker' if the prospect of loss has no meaning for it? [/ QUOTE ] WJL: The key factor is not that a player has "feelings" about the risk; the key factor is simply that they "are" at risk to begin with. The only difference between a human player and a botting player is that the botting player has voluntarily decided to commit all poker decisions to a playing algorithm; technically speaking it's still 2 humans risking money albeit one of them is using a pinch hitter. I'm quite sure the botting player cares very much about the money at risk. Yes, the UofA was at risk based upon their declared structure - they had to pay the opponent if the bot lost a session. The pro players risked little if nothing; I do recognize that personal egos were at risk but that's not the same as risking money. The risk of losing money is the great equalizer all things considered. So 'real poker' means that somewhere some human is connected to the risk of losing money for the given player (human or robot). The bot itself is connected to the risk algorithmically but not emotionally; the bot developer is connected to the money in every possible way.
|
 |
|
|
Linear Mode
