Originally posted by exigentskystrong chess will soon be the domain of computers only, "advanced chess" will be no more than "computer" chess ...
No, it is theoretically impossible to solve chess. However, it will not be long before computers completely surpass humans in chess ability.
i wish rhp allowed this to be seen here as it happens.
(maybe it does a little, but we do not know who until they are 3(b))
I guess with the exponential growth in computing power, a solution to chess does become more of a possibility. But we are still an extremely long way from that now.
The power of chess engines comes from the efficiency of its alpha-beta pruning algorithm. The ability of a good engine (and computer) to get to a depth of 20-30 ply depends on this algorithm, the basic purpose of which is to eliminate unpromising branches of the "tree". If an angine had to fully process and evaluate every possible position, the depth of analysis would be severely limited, and even the most powerful computer will grind to a virtual halt several ply deep.
Yet, while this pruning occurs, there is always the possibility that a computer will abandon a promising line, or overlook a subtle threat. This is particularly the case in complex, positional play, where the differentials in an engine's evaluation accross many candidate moves is very small. It is in these situations where strong human players, applying sound theory, can still have the upper hand.
This may sound crazy, but i think chesscan and will be solved. The game of chess has been played well over 1500 years, and there has been many masters, many good games, many excellent moves. The game has gradually been being solved as we speak. Ever since they began keeping history of the moves, thats as far back as we can go for data. I think after you plug in that data, along with all the data you need that didnt get from the history of EVERY SINGLE GAME ever played you will have a system that can beat anyone. Because to say that a man can beat a computer with every single move ever played, and all the best players ever combinded into one, to say it could be beat is just plain talking silly.
-trallphaz
Originally posted by trallphazEvery recoerded game in chess history between masters yields fewer than the number of theoretically possible games after five moves.
This may sound crazy, but i think chesscan and will be solved. The game of chess has been played well over 1500 years, and there has been many masters, many good games, many excellent moves. The game has gradually been being solved as we speak. Ever since they began keeping history of the moves, thats as far back as we can go for data. I think after you pl ...[text shortened]... plain talking silly.
-trallphaz
Originally posted by bassoAfter Hydra v Adams I find it difficult to agree.
Kasparov weighed in on this a few months ago on the Charlie Rose show. He said that humans would always be able to beat computers because computers will always lack an essential quality of good chess-playing that humans have -- intuition.
Nobody has mentioned this yet and I am curious where the thread will go...
I have been playing with end game tablebases for a while. All possible 5 piece combinations are solved and work is well underway on the 6 piece combinations. The file sizes are getting rather large for some of the 6 piece combinations, but eventually all of them will be solved and work will begin on the 7 piece combinations.
End game tablebases do not play perfect chess, but they do play blunder free chess. They will always win a won position, even if maybe not via the most direct means. Regardless, it will never draw or lose a won position, and it won't lose when there is a theoretical draw.
Given all of this, and the advances in computing and data storage, does anyone honestly not believe there will be a day when 32 piece tablebases are solved?
Originally posted by WulebgrI dont know that much about the history of chess, mainly just the way to play and stuff. I dont worry about what has been as much as what can and/or will be to come. In other words, i dont know chess history only chess theory if you will?
Every recoerded game in chess history between masters yields fewer than the number of theoretically possible games after five moves.
Originally posted by TippedKingThat day is so far off that it cannot be anticipated.
Nobody has mentioned this yet and I am curious where the thread will go...
I have been playing with end game tablebases for a while. All possible 5 piece combinations are solved and work is well underway on the 6 piece combinations. The file sizes are getting rather large for some of the 6 piece combinations, but eventually all of them will be solved ...[text shortened]... orage, does anyone honestly not believe there will be a day when 32 piece tablebases are solved?
All of the 3- and 4-piece tablebases require a mere 30 MB of storage, but adding the 5-piece jumps the storage requirements to a bit over 7 GB. On my computer, I have 26 of the 6-piece tablebases (a mere fraction) and these occupy 17.1 GB on my hard drive. Most of us do not have a hard drive large enough to handle the 7-piece, if they existed.
Of course, when 64-bit processors are standard, and RAM is measured in terabytes, our hard drives will be much larger. These may well accomodate 7-piece and even 8-piece tablebases, but I doubt we'll be able to process 9-piece tablebases in a reasonable amount of time before our processors are running at 128-bit.
Originally posted by exigentskyAgreed, looking at the man vs machine results and Hydra-Admams match, Its not far that machine will have superiority over humans in chess
No, it is theoretically impossible to solve chess. However, it will not be long before computers completely surpass humans in chess ability.