On Fri, Jul 16, 2010 at 4:17 PM, Mark Boon <[email protected]> wrote:
> On Fri, Jul 16, 2010 at 7:28 AM, Don Dailey <[email protected]> wrote: > > I kind of like to think of games (of perfect information) in terms of > what > > chance does a top human (or future human) player have a beating or > drawing a > > player who is omniscient in the game. If that chance is very close to > > zero, it's a good game and it doesn't make it a "better game" to make > the > > chances even lower. > > I completely disagree with this point of view (to which you are > entitled of course). > If this was the only measure, then we could all just as well play the > same game. No need for the differential between chess, checkers, > draughts, go, or any other game of perfect information. They'd all be > equally good. We don't disagree, you just misunderstand what I'm saying. And what I'm saying is that I don't buy the idea that 19x19 is a "better game" than 21x21 or some other size. But I certainly agree it is DIFFERENT to the point that small boards are almost like a completely different game. Whether one game is "better" or not is purely subjective in every sense of the word. I used Checkers as an example of a game that is "practically played out", and when the world champion of checkers died in 1995 (Marion Tinsley) he was clearly the dominant player in the world - nobody came close. So if you or I took up the game right now we could be at it for a lifetime and still be challenged by it. Sure, it may well be subjective to prefer one game over > the other, but that doesn't mean the differential doesn't exist. If an > omniscient player can win with 100% certainty against the best player > in say checkers, does it make the game equivalent to Go on 19x19? > Checkers is a poor example, because I think the top checkers players would still draw most of the games against a perfect player. So the answer is NO, checkers is not as good a game as go, even on a 9x9 board - because I think go on a 9x9 board is more challenging than checkers. You have to pick a game where the chance of playing perfectly by human master is really small. Then for all practical purposes the game is challenging and it's a matter purely of preference. Marion Tinsley often drew games against players not even in his league because that is the nature of checkers, it just isn't a good game in the sense I'm talking about. > Personally I don't think so. It's the diversity of intellectual > challenges that 19x19 Go poses that make it attractive to me. A > diversity I don't find in checkers. But you can probably narrow Go > down by reducing the size. Maybe 7x7 or 9x9 would approach checkers in > complexity and diversity. Even though we might not be able to beat an > omniscient player in either, to me checkers is a lesser game. And so > is 9x9 Go. > I think checkers is more interesting and complex that 7x7 go, but not 9x9. If the top humans played 7x7 almost every game would end up as draw assuming they used 9.0 komi (if that's what it is.) It would be even more so than in checkers which usually ends in a draw. I'm not a go expert, but my sense of it is that 9x9 at the very top levels would still return very frequent decisive results. But perhaps I'm wrong - in which case I would think a bigger board would clearly make for a "better" game. > > I don't know if at 17x17 the game is much different than on 19x19. > Maybe not to me. But it may be to a 9-dan pro. But even if 17x17 isn't > all that different to a pro, 9x9 surely is. Even to me it is, and I > don't enjoy playing 9x9 nearly as much as 19x19. Even though I'd have > zero chance against an omniscient player on 9x9. > All you are saying is that it's subjective - a point I agree with. I'm sure there are people who like checkers much more Go, but does that prove it's a better game? I don't think so! So the fact that you find 19x19 more interesting doesn't mean anything. > > Do you know the (kids) game that is played with a matrix of random > numbers? One player picks (and removes) a number from a row, the other > picks a number from any column in the row of the previous player. The > player with the largest total in the end wins. It's a stupid game, > played badly by humans and very well by computers because it's very > easy to make a program do the search way beyond a human can. Make the > matrix big enough and no player will be perfect, whereas a computer > easily can be made to beat any human 100% of the time. > > Is this game just as 'good' as Chess or Go? > I don't know enough to judge it. If you are the type of person who can really get into it it might be profound. There are people who have phenomenal calculation ability and it's quite likely that with practice and a lifetime of devotion to such a game it would turn out to have deep strategy. Of course it might also turn out to be trivial once a few basic principles and lot's of practice is concerned. > > Mark > _______________________________________________ > Computer-go mailing list > [email protected] > http://dvandva.org/cgi-bin/mailman/listinfo/computer-go >
_______________________________________________ Computer-go mailing list [email protected] http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
