Recommend
6 
 Thumb up
 Hide
39 Posts
1 , 2  Next »   | 

Chess» Forums » General

Subject: The impact of computerized chess players on the game rss

Your Tags: Add tags
Popular Tags: [View All]
Matthew M
United States
New Haven
Connecticut
flag msg tools
admin
8/8 FREE, PROTECTED
badge
513ers Assemble!
Avatar
mbmbmbmbmb
This was orignally going to be a thread asking what the longest forced-mate sequence was - but a little googling revealed this:

Quote:
The ability of computers to do extensive search has had ramifications on endgame play, where only a few pieces are left on the board, such as a king, bishop and knight versus a king and a pawn. Certain endgame arrangements were always thought to represent a draw -- no human had ever seen a way to win. Nevertheless, through deep -- very deep -- computer searches, some of these positions were proven to be a forced win. In one setup, a Connection Machine supercomputer found a forced checkmate in an astounding 249 moves! There is a scientific paper on the subject in the original German, which said playing against a computer programmed with these endgame sequences was "Schachspielen wie gegen Gott"-- chess playing as if against God. Garry Kasparov himself said it best: "Sometimes quantity becomes quality."

Once some of these endgames were shown to be wins, a few tournament players tried to memorize the necessary sequence of moves. This leads to a fascinating point: Whereas researchers began by trying to make computer chess systems imitate the style of humans, paradoxically it turns out that some humans play their endgames by imitating computers! Thus, at least in some aspects of endgame play, machines are clearly superior to humans.

Source: http://www.research.ibm.com/deepblue/learn/html/e.8.1.html



Which now leads me to wondering what the chess fans on BGG thinks of the impact that the current era of computerized grand masters is having on the game - is it good for chess? Will computers eventually remove the soul of chess, or worse - solve the game to the point where it is nothing more than complicated version of tic-tac-toe?

I have no real investment on any side of the issue - just curious about what discussion might come from this thread so I'm throwing this out there to start things off.

-MMM

PS - here's a link where you can see a forced mate in 270 using a legal (if unlikely) setup! http://www.chessbase.com/puzzle/games/puz7.htm
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Valdir Jorge
Canada
Montreal
Quebec
flag msg tools
badge
Avatar
mbmbmbmbmb
Hi Matthew!

As long as we can insure that humans are not "connected" to a computerized GM in face to face tournaments, I think Chess will live forever. I mean, there are many people (and I mean, many people) who play the game at tournament level and don't know a basic ending like KBN x K or cannot tell if a KP x K ending is winnable or not without making the moves on the board. I'm only a class C level player (my rating is around 1500) but I've seen players one or two levels up that don't grasp some of the concepts that are clear to me.

Oh, and I looked at that 270-move mate. That was mighty boring!
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Philip Thomas
United Kingdom
London
London
flag msg tools
badge
Avatar
mbmbmbmbmb
The 270 move mate is also a draw, you know. I am assuming no piece is taken or pawn moved in the sequence: the game ends after 50 moves. According to the rules I was taught anyway.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Severus Snape
Canada
flag msg tools
Pascal said, "The eternal silence of these infinite spaces terrifies me."
badge
"The heart has its reasons which reason knows nothing of."--Pascal
Avatar
mbmbmbmbmb
In face-to-face competition between human beings, all will be well on the surface. Behind the scenes, however, computers have been used for sometime to help players at the highest--and richest--levels prepare, especially in their study of openings.

I guess this is just technology at the service of chess, but it is a long way from the 1960's and 1970's where players like Bobby Fischer and Bent Larsen (where have I heard that name before?) went it alone, unless you are going to be nit-picky and talk about how Larry Evans, and even Bent Larsen(!) helped Fischer "prepare" before a few big events. But how often did this happen, and who do you think ultimately called the shots? goo
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Matthew M
United States
New Haven
Connecticut
flag msg tools
admin
8/8 FREE, PROTECTED
badge
513ers Assemble!
Avatar
mbmbmbmbmb
Philip Thomas wrote:
The 270 move mate is also a draw, you know. I am assuming no piece is taken or pawn moved in the sequence: the game ends after 50 moves. According to the rules I was taught anyway.


There is a pawn move every 20 moves or so for most of it and the end is a big ole slaughter.

-MMM
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Michael Kandrac
United States
Grand Prairie
TX
flag msg tools
badge
Avatar
mbmbmbmbmb
The "50 Move Rule" states that a game is a draw when 50 consecutive moves are made without there being a capture or a pawn move. I looked at the game for the same thing, and of course there are pawn moves every 20 moves or so, therefore no draw.

Gg
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Paul DeStefano
United States
Long Island
New York
flag msg tools
designer
badge
It's a Zendrum. www.zendrum.com
Avatar
mbmbmbmbmb
My first job was working at the largest independent retailer of chess computers. This was mid 80s-90s. They were just attaining master level.

The same discussions came up repeatedly then.

The same asnwer comes up now:
People still run the mile, even though a car can do it faster. It has always been assumed the game can be solved. Humans won't be the ones to do it. Does the game change for all of the players if somewhere a computer does what a human cannot?
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Ava Jarvis
United States
Bainbridge Island
Washington
flag msg tools
badge
Avatar
mbmbmbmbmb
The famous Kasparov dethroned by IBM's Deep Blue incident was a little bit rigged. For example, it was agreed upon that both Kasparov and Deep Blue could study all the public games of the opponent. Many of Kasparov's games are, of course, public. But IBM claimed all Deep Blue games to be private, and thus Kasparov was playing cold, while Deep Blue was not.

Nevertheless, Kasparov also had a bit of a temper during the tournament (he'd never drawn/lost before), which very much did not help. He was unnerved. He probably could have drawn with Deep Blue if he had kept his temper.

Kasparov of course wanted a rematch with Deep Blue, but it isn't in IBM's best interest to do so, so there aren't any (nor are there ever likely to be) plans for a rematch. Kasparov is one of the few chess players who got sharper with age, but there's probably a limit on that.

The incident probably did plenty of damage to chess's image. But even if chess is solved, to many people it won't make much of a difference. Solving chess would involve making the optimum move no matter what your opponent did, or else to have optimum moves mapped to each of your opponent's possible moves. Deep Blue and AIs similar it that tackle hard games (even games like checkers, which we tend to think of as not that complicated) are neural nets, a sort of "thinking" program that can learn. The best you could say for a solved chess is that you come up with a very, very large simple program listing that covers every case down the branches that would result in it winning. And that is something very few (if any) people can memorize.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nelson Lamoureux
Canada
Natashquan
Québec
flag msg tools
badge
Avatar
mbmbmbmbmb
Solving chess is currently not possible through brute force algorithm. The number of nodes on the analysis tree is larger than the number of atoms in the galaxy. So as of now, solving chess is impossible. Regarding this subject, there is an excellent article on wikipedia about computer chess:

http://en.wikipedia.org/wiki/Computer_chess

And no, it would be impossible to memorize the solution to chess.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Paul DeStefano
United States
Long Island
New York
flag msg tools
designer
badge
It's a Zendrum. www.zendrum.com
Avatar
mbmbmbmbmb
haver wrote:
The number of nodes on the analysis tree is larger than the number of atoms in the galaxy.


That would be fascinating, considering chess has a relatively confined tree. Do you have a source for that.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Valdir Jorge
Canada
Montreal
Quebec
flag msg tools
badge
Avatar
mbmbmbmbmb
Hi Paul!

If you consider an average of ten possible moves at any one point (a low estimate), a 40-move game would have 10**80 possible positions. Even discounting a lot of duplicates (because there are two of each rook, bishop and knight), we're still on the 10**40 range. Check this page for more details: http://mathworld.wolfram.com/Chess.html
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Jeff Dawson
United States
Missouri
flag msg tools
Avatar
mbmbmbmbmb
Two Points:

First, if any euro board game were put to the scrutiny and analysis that Chess has been over the past few decades, it would certainly be "solved". But Chess has not been solved, and will never be solved.

Secondly, theoretically, there is no such thing as a "forced win" in chess. There is no sequence of moves by white (or black) that can "force" a win. Theoretically, if white and black played perfectly, the BEST result white (or black) could ever hope for is a draw.

Chess games are won and lost by material and positional blunders. That is what makes computers so good at chess, they NEVER make tactical material blunders that you and I make and then ALWAYS find the tactical blunders we make (The Computers that look 40 or more plies deep that is). The only advantage the top grandmasters have over a computer is a better understanding of positional considerations. That's the intuition and experience that cannot be as easily calculated as cold, hard tactics. (Incidentily that is why computers are not good at a game like Go.) Deep Blue made huge strides in this area. An interesting read on this subject is "Behind Deep Blue", the book written by Feng-Hsiung Hsuthe, the creator of Deep Blue.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Jorge Montero
United States
St Louis
Missouri
flag msg tools
badge
I'll take Manhattan in a garbage bag. With Latin written on it that says "It's hard to give a shit these days"
Avatar
mbmbmbmbmb
checkmate123 wrote:
Two Points:
Secondly, theoretically, there is no such thing as a "forced win" in chess. There is no sequence of moves by white (or black) that can "force" a win. Theoretically, if white and black played perfectly, the BEST result white (or black) could ever hope for is a draw.


Care to post a link with a proof? Based on my undergrad math, I would think that there is no way of proving what you just claimed without solving chess.

How can you claim that if both players play perfectly, the natural result is a draw without knowing what 'playing perfectly' is? Since there is no perfect position evaluation function in chess, how in the world are you going to identify which move was the best, without going through the entire position graph?
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Jeff Dawson
United States
Missouri
flag msg tools
Avatar
mbmbmbmbmb
hibikir wrote:
checkmate123 wrote:
Two Points:
Secondly, theoretically, there is no such thing as a "forced win" in chess. There is no sequence of moves by white (or black) that can "force" a win. Theoretically, if white and black played perfectly, the BEST result white (or black) could ever hope for is a draw.


Care to post a link with a proof? Based on my undergrad math, I would think that there is no way of proving what you just claimed without solving chess.


I wasn't intending to form a proof, however Martin J Osborne's "An Introduction to Game Theory" would be a good place to start.

http://books.google.com/books?vid=ISBN0195128958&id=Ep7bPXVT...

My statement is based on experience, intuition, and logic. If it were possible for white to always make the perfect move, black will be able to counter it (at least for a draw) with perfect play. The fact is, the initiative of white is not strong enough to force a win, there needs to be a blunder somewhere by the other person to win any game of chess, even if it is infinitesimal. And even with an infinitesimal "error", it is often possible for the losing side to draw with good technique.

Look at a "solved" game like checkers. Every game is a draw with correct play. A game can only be won or lost with a blunder. Or take a much simpler game, tic-tac-toe. A game can only be won or lost with a blunder. This, in my mind, is the fundamental principle of what makes a game a "fair" game. If one side has an advantage that cannot be overcome with "perfect play", I don't want to play it. Do you? I don't need a computer to crunch out every possible variations for every possible move to believe this to be true. Experience and game theory tells me it is true.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Jeff Dawson
United States
Missouri
flag msg tools
Avatar
mbmbmbmbmb
Here is an article on solved games though.

http://en.wikipedia.org/wiki/Solved_board_games

And I need to make an amendment.

Checkers endgames up to 9 pieces (and some 10 piece endgames) have been solved. Not all early-game positions have been solved, but almost all midgame positions are solved. In January, 2005, the opening called White Doctor was proven to be a draw. Contrary to popular belief, Checkers is not completely solved, but this may happen in several years, as computer power increases.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Jeff Dawson
United States
Missouri
flag msg tools
Avatar
mbmbmbmbmb
"I think it's almost definite that the game is a draw theoretically."
- Bobby Fischer speaking of Chess

He knows more about Chess then all of us put together. That's good enough for me.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Matthew Barratt
United Kingdom
Royal Leamington Spa
Warwickshire
flag msg tools
Avatar
mbmbmbmbmb
Quote:
"I think it's almost definite that the game is a draw theoretically."
- Bobby Fischer speaking of Chess

He knows more about Chess then all of us put together. That's good enough for me


But if he isn't completely definite, then I don't see how any of us could be sure that chess is drawn with perfect play.

Quote:
My statement is based on experience, intuition, and logic. If it were possible for white to always make the perfect move, black will be able to counter it (at least for a draw) with perfect play.


There do exist symetrical positions where the first player to move can force a win ( A trivial example being: WK on h1, BK on h8, WR on b7, BR on b2, WR on a6, BR on a3, WP on a4, BP on a5. Whoever moves first can mate by moving the a file rook onto the back rank. Even simpler is if both players have pawns on the same, otherwise empty, file that are one square away from an unprotected back rank square).

Although it is 'almost definite' that the starting position for chess is drawn I don't see how we can be sure that it isn't a win for white nor indeed to absolutely rule out the vastly more remote possibility that the starting position is a win for black.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Steve Hope
United States
Woodside
California
flag msg tools
badge
Likes: Mountains, Tundra Turn-offs: Serpents, Marsh
Avatar
mbmbmbmbmb
I agree w/Bobby Fischer. And with the posters who say that there remains a chance that if Chess is ever "solved", it will be a White win (and I guess theoretically a Black win is possible, though it would be one of the most unlikely/counterintuitive discoveries in the history of the world). And with Jeff when he says that any Eurogame with perfect information and only 2 players (don't know how many of these there are, though I'm sure there are some) would be "solved" if the same amount of effort was applied to analyzing them that is applied to Chess.

I'd actually be interested in a list (Geeklist?) of relatively simple games (played with no more pieces or spaces/hexes/whatever than Chess and with no random or hidden elements) which are HARDER to solve than Chess.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Jeff Dawson
United States
Missouri
flag msg tools
Avatar
mbmbmbmbmb
MBarratt wrote:
Quote:
"I think it's almost definite that the game is a draw theoretically."
- Bobby Fischer speaking of Chess

He knows more about Chess then all of us put together. That's good enough for me


But if he isn't completely definite, then I don't see how any of us could be sure that chess is drawn with perfect play.


OK fine, it's almost definite that chess is a draw theoretically yuk semantics. Plus we have some computer analysis that supports this that Fischer did not have.

Quote:
There do exist symetrical positions where the first player to move can force a win ( A trivial example being: WK on h1, BK on h8, WR on b7, BR on b2, WR on a6, BR on a3, WP on a4, BP on a5. Whoever moves first can mate by moving the a file rook onto the back rank. Even simpler is if both players have pawns on the same, otherwise empty, file that are one square away from an unprotected back rank square).


Sorry I fail to see you point. Could you clarify? I'm talking about from start with all 32 pieces.

Quote:
Although it is 'almost definite' that the starting position for chess is drawn I don't see how we can be sure that it isn't a win for white nor indeed to absolutely rule out the vastly more remote possibility that the starting position is a win for black.


Well, if there is a forced win from the start of the game, hypothetically devil, I am sure (or should I say almost definite) it is for white and not black.

Interesting conversation. What else?
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Olivier Clementin
France
Paris
Paris
flag msg tools
badge
Avatar
mbmbmbmbmb
I'm willing to bet $100 that the solution of chess is a white win. (I will make a provision in my will if there's any taker)
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Jorge Montero
United States
St Louis
Missouri
flag msg tools
badge
I'll take Manhattan in a garbage bag. With Latin written on it that says "It's hard to give a shit these days"
Avatar
mbmbmbmbmb
checkmate123 wrote:

I wasn't intending to form a proof, however Martin J Osborne's "An Introduction to Game Theory" would be a good place to start.

http://books.google.com/books?vid=ISBN0195128958&id=Ep7bPXVT...

My statement is based on experience, intuition, and logic. If it were possible for white to always make the perfect move, black will be able to counter it (at least for a draw) with perfect play. The fact is, the initiative of white is not strong enough to force a win, there needs to be a blunder somewhere by the other person to win any game of chess, even if it is infinitesimal. And even with an infinitesimal "error", it is often possible for the losing side to draw with good technique.

Look at a "solved" game like checkers. Every game is a draw with correct play. A game can only be won or lost with a blunder. Or take a much simpler game, tic-tac-toe. A game can only be won or lost with a blunder. This, in my mind, is the fundamental principle of what makes a game a "fair" game. If one side has an advantage that cannot be overcome with "perfect play", I don't want to play it. Do you? I don't need a computer to crunch out every possible variations for every possible move to believe this to be true. Experience and game theory tells me it is true.


I have a degree in computer science. I've studied game theory as part of the curriculum years ago, and you know it. Pointing me towards an itroductory book is nothing but demeaning.

Your statement is based on experience an intuition... i don't see much logic in there. There are plenty of perfect information games that are won by the second player. Given that you send me to a book in game theory, I'd expect that you've at least read about them.

Why do you say that the initiative in white is not enough to win? Is it because a good amount of grandmaster games ends in a draw? As we learn more about a game, the advantage of initiative tends to get bigger, not smaller. Look at Go, for example: Komi, a score handicap to the starting player, is a relatively modern invention, to make the game more even. With new players, you don't need komi at all: the starting player's advantage becomes more and more noticeable as you improve your knowledge of the game. I might be wrong on this one, but I seem to remember reading that the value of komi was increased lately, because the starting player was still winning.

Game theory says NOTHING about the value of initiative, so it's not really helping your case. If you had read my previous post carefully, you'd see that I talk about position evaluation... if we had perfect position evaluation, game theory would help your case... not in this case. As far as experience goes, I had a 18XX ELO. you don't get past 1800 by playing 20, 50, or even 100 games. I'd like to think that I know something about chess. I'm not saying that perfect play doesn't lead to a draw, just that it's debatable: mathematics do not say that the perfect solution is a draw.

The most interesting part of your post is the last paragraph, where you claim that you would not like to play a game that would end in anything other than a draw with perfect play. I find that very telling: Since you enjoy chess, you just WANT chess to be drawn on perfect play, anr belive this is probably true: I call that faith, not reason. There's nothing wrong in belief without proof, you're in your right to do that: just don't try to sell it to me as an argument.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Steve Hope
United States
Woodside
California
flag msg tools
badge
Likes: Mountains, Tundra Turn-offs: Serpents, Marsh
Avatar
mbmbmbmbmb
olivier6 wrote:
I'm willing to bet $100 that the solution of chess is a white win. (I will make a provision in my will if there's any taker)


Out of curiosity, what makes you say this? The ratio of White wins to draws at top level play?
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Matthew M
United States
New Haven
Connecticut
flag msg tools
admin
8/8 FREE, PROTECTED
badge
513ers Assemble!
Avatar
mbmbmbmbmb
I'm guessing he meant a white win as opposed to a black win.

-MMM
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Jonathan Tang
Malaysia
KL
WP
flag msg tools
badge
Avatar
mbmbmbmbmb
haver wrote:
The number of nodes on the analysis tree is larger than the number of atoms in the galaxy. So as of now, solving chess is impossible.

I'm not quite sure I understand this. How exactly do we know the number of atoms in the galaxy? And what does that have to do with solving Chess?

An 8-digit calculator doesn't need 99,999,999 atoms to make its calculations.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Jonathan Tang
Malaysia
KL
WP
flag msg tools
badge
Avatar
mbmbmbmbmb
JeremiahClayton wrote:
There are more permutations in Chess than an 8-digit calculator could calculate.
Haha, I think my point is that you don't need 1 atom per calculation.
Even if we knew the number of atoms in the galaxy.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
1 , 2  Next »   | 
Front Page | Welcome | Contact | Privacy Policy | Terms of Service | Advertise | Support BGG | Feeds RSS
Geekdo, BoardGameGeek, the Geekdo logo, and the BoardGameGeek logo are trademarks of BoardGameGeek, LLC.