Talk:Chess

You're right. Then, we've got Ches v/s šâh (shah), let's discuss this and interchange ideas! --Neigel von Teighen 18:17, 30 Dec 2004 (UTC)

• • Heh - I don't think you realize how many theories there are, but every major current theory certainly.

Why? There is already a specific page explaining the origins of chess. Dan100 21:59, Jan 6, 2005 (UTC)

Computer Chess

The point or fairness of comparing a human player to a computer program becomes ridiculous and merely a marketing tool for the computer manufacturer, and a source of publicity and income for the volunteer chess player. Unless the games become timed on a speed equivalent basis, they remain absurd, as the computer should either be given 1 microsecond to make its move, and the human player be given 1 day, because that is how long it takes both to calculate 100,000 positions.

I'm removing the entire above paragraph as POV. Whether it's fair or ridiculous is purely POV. Also, time to calculate 100,000 positions is irrelevant, because human chess algorithms and human hardware are totally different from computer chess algorithms and computer hardware. --Prosfilaes 03:20, 3 Jan 2005 (UTC)

Good work. What I first took to be an innocent editorial mistake by a newcomer is now appearing to be a tasteless obsession. The responsible editor is countermanding our work in a neverending manner and spiting against our every word of constructive advice. Vigilance will be required. --BadSanta

--Gunter 16:44, 3 Jan 2005 (UTC) The text has been modified as such: The computer's speed advantage and the timing of the games to never give the human equal processing time make the outcomes moot should the computer win. The highly publicised computer-human games are publicity stunts by the hardware manufacturers.

BadSanta, since you are unable to point out what you think is not factual but opinion in my two sentences, let me try for you: 1. You are unaware that computers are faster than humans? 2. You are unaware that companies like IBM are using these games to promote their corporate image? Spare me your terse useless comments and inane warnings, instead tell me clearly how you cannot comprehend these two facts and why you keep removing them? They are a perfectly valid addition to the article, they are factual. Wikipedia is a community, this Chess article is not your baby.

Stop fighting! Is there any reason to delete that text? I don't see any POV, but only facts (maybe I'm wrong). Stay cool and don't let this go into an edit war. --Neigel von Teighen 17:23, 3 Jan 2005 (UTC)

The proper scope of this article is not be to engage in speculation about the motives of corporate hardware manufacturers to host computer-human contests. Furthermore, an accusation against them for purely engaging in "publicity stunts" instead of "advancing AI" is inappropriate.

Your assessment of the advantages computers have against human opponents is not altogether factless yet your conclusions are imbalanced to the extreme and incomplete as you totally ignore the compensating advantages top, well-trained human players have against computers. [In my studied opinion, this article is definitely not the proper place for a thorough explanation of such difficult-to-define, psychological and human learning details, however.]

How do you reconcile the facts that Kramnik fought Deep Fritz to a draw in 2003 and Kasparov fought Deep Junior and X3D Fritz to draws in 2004 with your assessment? Your original write-up was so salacious, one must have drawn the conclusion that it is impossible for ANY human to compete against top chess supercomputers and programs yet the best human players at their best ... still do.

Please read, think and learn a lot more about what you seek to write with authority. --BadSanta

You're right: Kramnik (I saw that live through Internet. What a match!) and Kasparov did it, but not anyone can. These are super-trained and professional chess players and I can assure you that I could never beat such a machine. Obviously, this fact shows that there's no place for the word 'impossible', maybe 'very difficult'. About publicity, we agree, that can't be in the article. --Neigel von Teighen 18:14, 3 Jan 2005 (UTC)

--Gunter 18:53, 3 Jan 2005 (UTC)It is not speculation as to why e.g. IBM made Deep Blue, to promote IBM hardware, but ignorance on your part to think they did it for some greater glory of furthering AI (which they did not). IBM saw a marketing opportunity and took it, that is what they do.

The matches are speed biased. The computer program failing to win only shows the inadequacies of the programmers ability. If both computer and human used the exact same logic for playing, the computer would always win, due to it's speed. So as i pointed out, unless the computer is limited to the same number of moves it can calculate as the human, what are you proving? The obvious, that computers are faster than humans?

Since the current section on Computer Chess implies some amazing feat in computers beating humans at chess, i would like this balanced to give a more realistic view of this.

Now I understand your point about publicity. But, i don't think it should be placed here. Perhaps the IBM article is a better place. This is about chess and not about publicity in chess. The point to discuss here is if we should include the speed fact (I would). --Neigel von Teighen 19:01, 3 Jan 2005 (UTC)

If computers and humans used the exact same logic for playing, it would depend on the logic. If humans tried to play chess the way a computer does, the human would get squashed. But you seriously underestimate the power of the wetware the human has; just because it doesn't multiply fast, doesn't mean it's not powerful. Humans do pattern-matching stunts that computers can't even come close to, and given the amount of neural matter devoted to those problems, it's like humans are much, much faster than computers at those problems. A human can look at a board and see whether it's a good situation or not without really looking at any future moves, and defintely without doing an exhaustive multiply search, where a computer would be forced to check out the decision tree to make the same judgement. If a computer tried to play chess the way a human does, it couldn't beat the elementary school chess champ. Human wetware and computer hardware are in no ways isomorphic, and comparisons that claim that they are are unrealistic.

IBM is a publically owned corporation; its sole ultimate goal is money. While the researchers who worked on the machine understood that it was a publicity stunt, it was also an important study as to where computers stood against the best in the world. --Prosfilaes 22:19, 3 Jan 2005 (UTC)

My thoughts exactly. Dan100 22:05, Jan 6, 2005 (UTC)

Vote

I see that we have a revert war going on over whether the following paragraph should be included in the Computer Chess section of this article:

The computer's speed advantage and the timing of the games to never give the human equal processing time make the outcomes moot should the computer win. The highly publicised computer-human games are publicity stunts by the hardware manufacturers.

The vote is this: Should the above paragraph be included in the article. The four possible answers are: Yes, No, Abstain, and Other (it should be included, but in a different form). So, let the voting begin.

Yes

No

1. Far too much POV Samboy 23:47, 4 Jan 2005 (UTC)
2. Not encyclopaedic, though I don't see POV. --Neigel von Teighen 00:19, 5 Jan 2005 (UTC)
3. Poorly-balanced facts admixed with emotional, extreme opinions. -- BadSanta
   • --Gunter 12:35, 5 Jan 2005 (UTC)poorly balanced? i'm trying to balance the article that is already there with it's biased opinion. You keep raving about opinion, are you sure you know the difference between fact and opinion?
4. Some reference to processing speed is valid, but this is not well presented -- and the use of "moot" is just silly. Perhaps, after the current first paragraph of the "Computer chess" section, and as a lead-in to the recounting of recent matches, we could add something like: "In the 1940s [or was it the 1950s?], Mikhail Botvinnik suggested that a computer capable of playing at grandmaster level would have to be as large as the University of Moscow. The performance of the best chess-playing computers has improved as the programming has become more sophisticated (with the aid of some human grandmasters, such as ____ [I think Joel Benjamin has been a consultant]), and as more powerful mainframes have increased the number of computations that can be performed in a given time." As for the question of corporate motives, I think the price of IBM stock went up after Kasparov's loss; if so, that could mentioned as an interesting sidelight to the reference to that match. JamesMLane 02:56, 5 Jan 2005 (UTC)
   • --Gunter 12:35, 5 Jan 2005 (UTC)That should be a yes or other vote since you agree with both my points.
     • No, I don't agree with your wholesale disparagement of the meaningfulness of the human-versus-computer matches, and I don't agree with including unattributed speculation about corporate motives. It would be better to have nothing than to have the paragraph we were asked to vote on. My suggested language is an attempt to bridge the gap, by including facts (grandmaster involvement in programming, IBM stock price movement) and properly attributed opinions (Botvinnik's). I was trying to address the subjects you raised but to do so in an encylopedic manner. Does my suggested language seem OK to you (subject to confirming facts and filling in gaps)? It would be better if we could work out consensus language than to just decide something by a vote in the poll. JamesMLane 19:11, 5 Jan 2005 (UTC)
5. No, because it's entirely possible that the human has more processing power. To the extent that we don't know the results of any given contest, it's not moot. Is there any point in just saying the contests are for publicity purposes? Pretty much everything of the sort is, to some extent or another. --Prosfilaes 06:51, 6 Jan 2005 (UTC)
6. For all of the reasons above. Dan100 22:19, Jan 6, 2005 (UTC)
7. Human "processing power" is such a vague quantity; I expect we'll have to make significant advances in cognitive science before we can even begin to make a meaningful comparison. Until then, any estimate of human brain-time spent attacking a problem is a Meaningless Indicator of Processor Speed. As long as we're trying to reference other people's opinions, I should note that Raymond Kurzweil's The Age of Spiritual Machines advocates the position that, empirically speaking, chess tournaments demonstrate that machines are already doing things which we once thought only humans could do. In other words, he accepts the validity of the demonstration, regardless of IBM's motives. Furthermore, as added trivia, in Bruce Pandolfini's "blow-by-blow" account of the tournament, he quips at the last move, "The Black King looks pale. It's time to buy IBM stock." —Anville 00:46, 8 Jan 2005 (UTC)

Abstain

1. What exactly would one put in here? If I intend to abstain, surely I don't have an opinion to add.

Other

1. I would like it added (obviously), but reworded so as to conform to Wikipedia's standards.--Gunter 00:24, 5 Jan 2005 (UTC)

Note- Now our dear boy is arguing with the votes he doesn't like.

• I have no problem with that, as long as it doesn't mess up the tally, so I've reformatted. Also, I suggest that everyone refrain from comments like "our dear boy". Let's stick to the issues. JamesMLane 19:11, 5 Jan 2005 (UTC)

Arimaa

The article as it now stands implies that there are several chess variants which computers play poorly relative to humans. I know of none other than Arimaa. Can anyone name a few, and give evidence for their intractability to comptuers? The $10,000 prize money for Arimaa insures that real programmers are actually trying and failing, which probably isn't the case for other variants, but I'm curious about even minimal evidence to the effect that new variants are being spawned in which human intelligence dominates. Thanks. --Fritzlein 01:55, 10 Mar 2005 (UTC)

There are at least a few AI experts here as editors who may have furthergoing valuable, detailed input than I. However, I can tell you that the unrivalled intractability and theoretical depth of Arimaa is attributable mainly to its turn-order based upon each player having 4 moves per turn.

The vast majority of chess variants, likewise to standard FIDE chess, allow only 1 move per turn per player. Nonetheless, a small minority of chess variants (such as the well-known Marseillais Chess as well as the games in the Symmetrical Chess Collection) allow 2 moves per turn per player. There are probably a few others somewhere in the large literature of over 1000 chess variants. Perhaps, there is even another game(s) where 3 or 4 moves per turn per player are allowed. As a general rule, the theoretical depth at 1 move per turn is exponential time complete, at 2 moves per turn is hyper-exponential time complete, etc.

--BadSanta

I've played with writing a DragonChess game on my own time. It's a 3d variant of chess with 84 pieces and 288 squares. I may not count as a "real" programmer, and I would not be entirely surprised if there were some fatal weaknesses that showed up after greater processing, but it seems a lot more complex for the computer. Moving the number of moves at the start from 10 to over a hundred seems to make it a lot harder for the computer to process.

All of which is unsubstantiated, original research. YMMV. --Prosfilaes 07:41, 10 Mar 2005 (UTC)

I think in Xiangqi and Shogi the best human player is still better then the best computer program, due to higher game-tree complexity. However it is predicted that by year 2010 this will change, see Computer Shogi (http://portal.acm.org/citation.cfm?id=512155). Other chess variants, which can be difficult for computers are Dark Chess or Kriegspiel, as they are games with incomplete information and require more startegical planning and some psychological reasoning then deep move searching. Andreas Kaufmann 07:47, 10 Mar 2005 (UTC)

Thanks for the other examples. It's hard for me to judge how unique Arimaa is. Is a sufficiently large game-tree complexity all that is necessary for humans to excel relative to computers? Intuitively not. As a gamer I expect that a game must also contain strategic themes that the human mind can grasp. For example, if pawns in chess could move backwards, I expect that it would increase game-tree complexity, but also make computers even more dominant over humans, since it would remove a strategic point from the game. And isn't the endgame, with its reduced game-tree complexity, exactly the area of chess where humans are still unambiguously better that computers?

My hunch is that Arimaa simply a brilliantly-designed game. I could probably make a new game tomorrow with ten times the game-tree complexity of Arimaa, yet have it be hideously intractible to humans relative to computers. It isn't clear to me that human intelligence is such that a large playing field is all that is necessary to excel. --Fritzlein 14:48, 11 Mar 2005 (UTC)

The endgame of Chess is one of the few completely analyzed parts of Chess. At one time, it was thought that 50 moves without capture would never change the fate of a well-played game; computers proved that there are certain endgames (K+R versus K, IIRC) that can won if there was no such limit. If there are five or fewer pieces on the board, then a computer can play perfect Chess, using Ken Thompson's tables. King, rook and bishop beats a king and two knights in 223 moves by computer analysis. (First article in the Risks digest (http://catless.ncl.ac.uk/Risks/12.58.html))

You reduce the game-tree complexity enough, and computers can quickly become unbeatable. There used to be a paper on the web by the solver of Connect-4 listing the major games and how computers were doing on them. If you sorted them by game-tree complexity, nothing with a higher game-tree complexity than Chess had a contender for the world title, and everything with a lower game-tree complexity had computers the permeanant holder of the World Master.

It's true that upping the game-tree complexity generally makes it more intractable to humans. But in general, humans work by pattern-matching, which isn't as vulnerable to those increases. An increase in game-tree complexity directly ups the difficulty for computers. I don't think Arimaa is all that brilliantly designed or unique in this manner. --Prosfilaes 01:24, 13 Mar 2005 (UTC)

I am willing to assert more thoroughgoing that Arimaa is, in fact, a badly-designed game- a game designed with the single-minded purpose of being intractable for computer players at the sacrifice of many other known, quality game-design criteria (with the notable exception of human playability).

The inventor of this game is unknown to the chess variant community except for this one entry and as such, has demonstrated no general mastery of quality gameworks.

The $10,000 prize (if the money really exists) will never have to be awarded, of course, because it is generally recognized by AI experts that orders of magnitude of increase in CPU technology are required before a computer can possibly play this game competently, much less incisively enough to beat a human expert. As such, the $10,000 reward is little more than just a "publicity stunt" for the purpose of drawing attention to and popularizing this game (which has been remarkably successful).

The first-move-of-the-game advantage (for white) is generally the greatest enemy to stability in chess variants as a class of board games. For instance, in standard FIDE chess, appr. 55%-60% of the games which do not end in draws end in victories for white. This is with a turn-order based upon each player having 1 move per turn. Now, as an existential theorem, imagine how much more unstable a game must be in which each player has 4 moves per turn. Perhaps, catastrophically. This is true regardless of the fact that a game winning opening book for white has not yet been discovered due to the intractable nature of the game to present-day computers. If such a thing CAN exist in theory, then the game is fatally flawed.

--BadSanta

Request for references

Hi, I am working to encourage implementation of the goals of the Wikipedia:Verifiability policy. Part of that is to make sure articles cite their sources. This is particularly important for featured articles, since they are a prominent part of Wikipedia. The Fact and Reference Check Project has more information. Thank you, and please leave me a message (http://en.wikipedia.org/w/wiki.phtml?title=User_talk:Taxman&action=edit&section=new) when a few references have been added to the article. - Taxman 18:54, Apr 22, 2005 (UTC)

What do we mean by CHESS?

Chess in this article seems to be defined uniquely in the header as the modern, european chess (64 squares, 8 pawns, 2 rooks etc...). Wouldn't it be more appropriate to define first chess in a general way since many variants of chess have different rules and configurations? I would define chess as a strategic bord game where two armies are confronted and the goal is to take the King of the enemy. Then, it can be explained that the word chess taken alone is mostly used to denote the european, modern game, and if one wants to denote a variant, one says chenese chess, nordic chess, thai chess etc. --Philipum 11:52, 26 May 2005 (UTC)

OK, now I understood when I clicked the link Chess (disambiguation). But why not explain that directly in the article itself instead of forcing people to find this link? --Philipum 12:00, 26 May 2005 (UTC)

Suggestion

I think it would be nice to have a weekly(or longer) chess collaboration to expand bio's, openings, books and other chess related things. Thoughts?Falphin 19:26, 9 Jun 2005 (UTC)

Rukh

From what I understand, rukh is actually Persian for "chariot". It only happens to sound like the Arabic for roc.[4] (http://www.chessvariants.org/piececlopedia.dir/rook.html) — Gwalla | Talk 00:35, 11 Jun 2005 (UTC)

Age range?

Where 2 that age range come from ? Didn't Capablanca learn the game when he was four ?

Also the playing time > 1 hour ? There are very fast games, and also games that last months.

I Think thoses two parameters really make no sense in chess.