The End of Chess?

I'm posting this again because the other thread commited suicide. I received a PM to repost it, so here it is...


You have probably noticed the chess battle going on, 'Brains in Bahrain'? This is a man vs machine challenge, German chess computer Deep Fritz taking on the world champion Vladimir Kramnik for Russia. It currently a draw between the two, which is quite surprising as Kramnik was a big favourite to win the game. About five years ago former world champion Gary Kasparov fought a similar battle agains IBM's Big Blue, which he lost.

There's a big difference between the two chess computers. Deep Fritz can be run on a regular pc, while Big Blue is a super computer which can do billions of computations per second. Deep Fritz on the other hand relies heavily on using clever algorithms to win a game. In a battle between the two computers Deep Fritz won easily.

There's a big difference between the two man vs machine games also. When Big Blue defeated Kasparov, Kasparov had no way to prepare for the game, because he did not have the opportunity to study how Big Blue. Big Blue's programmers on the other hand had taped Kasparov, and knew his stengths and weaknesses. After the game Kasparov claimed the whole game was a big joke, and that Big Blue was not a good chess player, and wouldn't be able to defeat any other of the top ten chess players of the world other than Kasparov himself. He said that Big Blue was programmed specifically to defeat him, an none other.

The Deep Fritz game is different. Vladimir Kramnik has had lots of time to study Deep Fritz. He knows how the machine plays and he has even created a strategy specifically to defeat Fritz. He has found out that in ceartain positions the computer has extra advantage of it's computation power. Kramnik knows he has to avoid getting the pieces in these positions. But Kramink has had little luck so far, and many now experts think he will not defeat Fritz.

So, the title of the post, what's that to do with any of this? Well, former world champion Bobby Fisher is now afraid that we're close to finding the ultimate mathematical solution to chess. The mathematics that proves the perfect play. That would mean that by using these mathematical rules you would always win when starting a game. And would also mean the end of chess (it's no fun playing videogames when cheating, right).

Although this might seem a little strange it has happend before. An easy example is tic-tac-toe. If both players play a perfect game it will always be a tie, right?

Another, much cooler example is a game called Nim, a somewhat complex game played for hundreds of years all around the world. There is an excellent flash version of the game here: http://www.transience.com.au/pearl2.html The mathematical solution for the game is located here: http://world.std.com/~reinhold/math/nim.html
After Nim was 'broken' of course it became pointless to play. Playing correct, the player who starts will always win. So a fun game was ruined by maths!

And now maybe the same will happen to chess. We'll better develop a more complex game for our genious kids to play with...

When you look at the solution to that Nim game... it's quite a simple game in reality. A solution to chess would probably be much longer and much more complicated. And even if there was a solution - people still play noughts and crosses (tic-tac-toe), don't they? If the solution to chess is some completely amazing but remarkably simple three-line proof, then sure, people may stop playing it in favour of something else. But if it's a piece of mathematics that only a supercomputer can calculate... then chess will remain as popular as it is (and has been) for a long time to come. JMHO

btw, Kramnik tied with the computer in the end, didn't he?

I think people will still play chess,

in the cybercafe where I spend some of my free time, we play chess alot.
sometimes even going over moves and stepping back some 3 steps back (or even more) to find out how that one stupd move would effect the outcome.
We play all this on a normal wooden board, no fancy compute aided ches game (would help with te stepbacks ;) ).

I think chess is a good game to test your skills of logic and mathematics..
But most of all your strategic skills.

I'd keep playing it, even if there was a way to lways win. But don't think that "Perfect sollution" will ever come.

But perhaps it is just the time, young ppl finding chess boring and geeky??
I personaly know a couple of hardcore-punks (yeah the kind with the safetypins and the mohawks) that play chess, but hey punk isn't hip anymore so that doesn't matter.

Perhaps it's just that pop culture replaceing culture thing going on here, and the DeepFritz puter helping it.

* the_JinX isn't making any sence now, is he ??

I don't think development of an algorithm for the perfect chess game will kill the playing of chess. It will probably put an end to the human vs. computer challenges, but I think that chess between two humans will still survive.

A mathematical formula for a perfect chess game will be complex, and probably beyond anything that the human mind will be able to calculate on its own. If two people, without the benefit of computers sit down to play a game of chess against each other, the fact that their is a mathematical formula to predict the perfect move exists, does not mean that either one of the competitors can calculate that formula in their heads. Chess will remain as a battle of strategy and intellect between two people.

Perhaps the game will lose its aura of mystery and complexity if there is a proof that if two people play correctly neither will win, but that does not mean the game will die. Humans will have to come up with a different game to play, but that doesn't mean people will stop playing chess. Tic-Tac-Toe, which cannot be won if both players play correctly, is still played by children until they are able to figure out how to play so as to never lose. I think it will be likewise with chess. People will continue to play it until they figure out a way to adapt the mathematical formula for the perfect game into something their minds can comprehend.

Technology brings us forward. That is what it is supposed to do. I can't imagine that even if a mathematical algorithm was found for chess, how many regular people would know about it, or be bothered to learn it? Most people play chess for fun.

Quote:

---

An easy example is tic-tac-toe. If both players play a perfect game it will always be a tie, right?

---

I know people still play tic tac toe, and not always ending in a tie, so I doubt the end of chess is near.

Quote:

---

I can give my thoughts by saying 4 words: \/\/ho gives a §hit this is a security site what the phuck does it have to do with this BS. BS in Amercia means BullShit liekt his thread. Your are getting a negative points cause I feel like lame ass. The power of the antipoints system how ****ing retarded and thats my hacker thoughts.

---

This was entirely uncalled for. If you want to remain at this site, you will have to learn respect as well as how to contribute productively. Do yourself a favor and delete your post, and start over without the immature attitude!

reposted as requested :

All the algorithms in the world will never create a flawless chess game as it is a game of war based on strategy, as with any war and strategic scenario you cannot predict 100 % the outcome if that was the case you would be able to input the stats to the computer and predict the outcome of any war for real and that will never happen as you cannot predict or dictate innovation and you can never prepare for the human spirit which can and will always surpass any technology !

All computers can do is process the data and also now they can also turn that data into information but hasnt got the emotions to be able to use that in every possible way !

We humans are a amazing breed and will never be able to surpass our own being !

The whole Vladimir Kramnik vs. Deep Fritz match now ended as a draw. I wonder what would have happened if Deep Fritz was given the raw calculation power of Deep Blue, besides its ultimate algorithms.

A couple of weeks ago there was a very similar event, I don't know if you noticed it. It was a match between Chessmaster 9000 and Grandmaster Larry Christiansen, the reigning US Chess Champion. Chessmaster 900 defeated Christiansen 2.5 - 1.5 in this four game match.

I highly doubt there can be an exact mathematical solution to win a game of chess. Firstly there are simply way too many possibilities for moves. After both white and black have moved once, the position can be any of 400 different ones and after that the numbers go up exponentially. Sure, Deep Fritz won't calculate the most stupid moves but still there are too many variations. And usually when playing against computers the masters do some unexpected move right in the beginning of the game just to drop the computer from reading its opening variation libraries. Secondly, you can't know what your opponent is going to move next. No way. You can guess that he moves either, say, knight to e6 or rook to b3 and you can calculate the best moves against either one of them and a move thats an average option but that's always a compromize. Doing compromizes a lot just doesn't guarantee that you will win for 100% sure, I'm sure you see it too.

Btw, you can buy and view some information about Fritz here.

(World chess program olympics (neither Fritz nor Chessmaster participated as I believe commercial products are lined out of this competition): http://www.cs.unimaas.nl/olympiad2002/ )

EDIT/ADD: The number "400" was gotten by first counting the possibilities of white to move and then multiplying it by itself as black has exactly the same moves. At first you have 8 pawns and any of them can move either 1 or 2 squares, giving us 16 moves. Add two possible squares for both knights and you get 16+4=20. Then 20^2=400.

I must disagree with you Geepod on the matter of Chess. I believe it is possible to eventually write a series of algorithms, tests, etc. that would perfect a game of chess so that every move was responded to by the best counter move. This does not mean that war can be predicted. Chess is a highly variable game that allows for tremendous variation and innovation but it is still a bounded system. The rules are there and they do not change, nor can they be changed by the players.

I also agree, however, that this is immaterial to chess itself. Unlike the pearl game where the algorithm is fairly easy to understand and follow, or like Tic Tac Toe which is not only easy to follow but easy to formulate the chess algorithm would require a computer to do it. Therefore a human, while aware it existed, could not put the algorithm into effect against another opponent. So, as stated earlier, this might end the whole senseless fiasco of man vs. machine chess it would have no affect on man vs. man chess.

To say that if a computer beats a Grandmaster spells the end of chess is like saying the pole vault is dead because a plane can go higher. The track meet is dead because every event could be won by a machine rather than a man. Football is dead, just give me the ball in my Jeep Cherokee and I can score everytime.

An easy example would be my own chess playing. I am not a chess enthusiast (I prefer other more Philistine games) however occassionally I will play. When I do I normally play people of my calibre. A good chess player will defeat me easily. He may enjoy it, I may learn but overall the game is not very satisfactory. This doesn't "end" chess for me it means I seek an opponent who challenges me but against whom I have a chance of victory. The better player would need to teach me not play me. Then, if motivated, I can improve and move up the ranks and levels of players. The beauty of the game is that no matter what your skill level there will be players to challenge you and, with the exception of the Grand Masters, mentors to teach you.

The sadness of this whole man vs. machine debate is that Great Chess Players like Kasparov have to defend themselves when they are not playing "Chess" but are playing a specialized form of the game. Then to hear people talk about their failure is ludicrous. Someone who can't beat Chessmaster 3000 (like me ;) ) denigrating a Grand master because he didn't beat the best a team of programmers and computer scientists could develop over the course of decades is ridiculous in the extreme. If anything it points out the failure of our programming capability that these great men were able to even challenge a machine with the raw computing power contained in any normal desktop computer today.

Chess is not at risk, it is a game that has lasted the ages due to its complexity, depth, and breadth. The Grandmasters have not lost their edge they were not competing in a "Chess Match". It is good that we are improving our Mathematics and Algorithms to conquer this intriguing and complex problem with the help of machines. However, just because machines help us do things better than we can alone does not mean I will stop enjoying a walk with my wife even though the car could get us there faster.

Most chess games are similar to the General's perfect battle plan: The plan is perfect until the first shot is fired. Same with poker; don't play with someone who doesn't know the odds, they'll first get your dough while you're trying to figure out what they're doing (...absolutely nothing logical...) and then they'll beat you again when they don't even know they have a winning hand, just like they beat you when you know your hand is a looser. Come to think of it, i'd really like to have the money back it cost me to learn how to play pool against the kids in the Phillipines... Hmmm, ...or maybe they were just midget professional pool sharks dressed up like kids... too bad i didn't think of that while i was shelling out the lesson dough. But then that strawberry soda pop they kept serving up might have had something to do with it also. But back to a thoughtful note, just because AI can plot a certain player's preferred moves then beat him doesn't mean diddly to me, because i'm not in that league anyway. We have three chess sets, one is up all the time, and sometimes i can win and sometimes my wife can win. I used to try to keep 4 or 5 moves ahead suspended in memory, but that got to be worthless because she never moves where i think she should. i used to have an on-screen chess game and could beat the AI up to about the 3rd level but it was not the same as playing a carbon-based lifeform. Besides, have you ever tried to get a 'puter to pass the chips, or go fetch a cold drink from the fridge? Naw, even if AI could beat every carbon-based lifeform on the planet it still would not be the same, and most of us would not care anyway, and we'd still be looking at the next carved chess set and thinking how good it would look on our little table. AI profficient chess programs will never replace chess nor reduce the number of people who will play. JMHO! :D

Chess has always been a game that has been heavilly analysed (by humans!).

For years there have been books published on vast numbers of openings i.e. the first 10-30 moves in most games. Translating all of this to software is easy. One thing the software is very good at is to try to stick to one of these standard positions, and then use its raw power when there are fewer pieces left on the board (and hence fewer possibilities to calculate 10+ moves in the future). Which is exactly the way that most grandmasters play.

In the recent match, Vladimir Kramnik failed to make any headway in the last match, bceause he had prepared a special variation of an opening, but the software was astute enough to transpose it back into a standard position.

In many ways Kramnik was behaving exactly like the computer program, because he was saying if we play these 11 moves each, then we get to a position that neither of us has seen before. He had spent days doing detailed analysis of what to do if that should happen, which he was sure would give him an advantage. A technique that has been used by grandmasters for decades, but doesn't work if the software (or a human opponent) can see it coming.

To improve the software, you would do an exhaustive analysis of what would happen in this new position, and add it to your standard database ..... which again is exactly what human grandmasters do.