GameKnot: Chess related, Top Engine(s) vs Top Corr Player(s) Match . . .

Chess related: Top Engine(s) vs Top Corr Player(s) Match . . .

From	Message
baronderkilt 01-Nov-10, 23:33	Top Engine(s) vs Top Corr Player(s) Match . . . In another thread navig18 brought up the Question of how we think the top Chess engine(s) would fare against top Corr Chess player(s) ... playing under Time Control and conditions of corr play ? This interests me as well, and I no longer am certain of the result of such play, tho I have long favored the human chances. I would like to get opinions as to what would constitute appropriate conditions for such a game or match. Or if you wish to take it more personally, what conditions would You like for playing such a challenge yourself, or view it played by another human player ... in order to provide the optimum experience, such that it would not favor the computer in any way for its simply being a computer ? For egs, it could not be allowed to fatigue the human or simply wear them down by its willingness to play-on forever in nonsensical nor hopeless positions. Do we agree on that? Or what else? Of course any comments on who would win, or draw etc, are certainly welcome .
baronderkilt 01-Nov-10, 23:50	Some of my thoughts . . . I would like to see the corr player be able to: _______________________________________________ * Choose their color, if only playing one game . Have choice of the corr Time Control used. Be permitted Nalimov Endgame TableBases if the computer has them built-in to its programming. * Have ample time to Postpone for life aor health matters * Be assured that their Engine Opponent cannot receive Human advice ~! * The engine should be considered Resigned if it falls beneath some particular assessment value in its own calculation. Not sure what, but perhaps something in the range of 2.25 to 3.5 'points' !? ... in consideration of the fact these would be Master + quality games, and such players Will convert such an advantage. With that in mind, perhaps my range is too high? Or should it perhaps have to be sustained some number of moves? Or would you do away with such as this altogether, and make the human Play Till Mate ... to be fair to the E-Box ?! * The human should have some manner of acceptible incentive for playing, winning or scoring on the engine? Such as? (Honor of all Mankind? Trophy? Script added to Profile? $$ CASH $$ ? Selfsatisfaction? Or my personal favorite, GK T-Shirt~! or ballcap. Maybe say something appropriate, like what? "I Beat The Box~!" or "Fritz is a Patzer" "I Ate The E-Fish" * * * * * * * * HELP~! I have run out of ideas ... }8-D
ganstaman 02-Nov-10, 17:14	Well, I guess we'll try to make this thread more popular to discuss this since the other thread already has a number of posts. In general, the extra time of CC helps the human much more than it helps the computer. Because the game tree of chess grows exponentially, the computer takes more and more time to get to each ply deeper. The human generally starts of more shallow, so in the extra time that the computer gets another ply or 2, the human can nearly catch up by a good number of ply. However, there are so few mistakes made by top computers these days. It would be hard for me to imagine such a match being anything but a draw. Just my semi-random guess.
baronderkilt 04-Nov-10, 01:02	That's the thing ... It seems like a human must work awfully hard to more than Draw ... or know some specific weakness in the play or opening book of the engine, perhaps. (Big-Secret: Fritz 8 or ChessTiger dont seem to play particularly well against a Scotch/Goring Gambit, at least at 10 min play ... shhh. And some even past 2600 blitz ratings still may be had with a KI, WT Lopez, or similar pawn positions of slow building but prolonged attack ... tho I've never seen Rybka play. But surely if even I have some anti-computer Secrets; the Top players must know some real whoppers to try! Yes!? )

« Back to forum

chess online