Chess computer weakness: Blocked positions and 50 move rule.

edit

It would seem appropriate to say something about block positions. All chess computers are susceptible to losing against humans in blocked positions as Hikaru Nakamura shows in this game against the top rated computer Rybka. http://www.chessgames.com/perl/chessgame?gid=1497429 SunCreator (talk) 20:30, 30 March 2009 (UTC)Reply

That was Rybka 2. What computer was it? Has Hikaru Nakamura played Rybka 3 on a Quad computer? Mschribr (talk) 22:21, 7 March 2010 (UTC)Reply
It is my understanding that Rybka was withdrawn from playing against humans that use such anti-computer tactics/strategies. Regards, SunCreator (talk) 00:23, 8 March 2010 (UTC)Reply
Why has Nakamura or anyone else not played Rybka 3 on a Quad computer in public in the last 2 years? Because they have no chance. Rybka was handicapped when playing GM. Even puny cells phone can beat a GM. Mschribr (talk) 03:20, 8 March 2010 (UTC)Reply
Firstly the above game was within the last two years and played on the public Internet Chess Club servers. And secndly as I said above, my understanding is that Rybka was withdrawn from such encounters. Imagine the negative publicity if the computer losses. Note the above game was played with a computer friendly 3 minute time control. It's not the only game either, here against Crafty a similiar bloackade technique was employed. Regards, SunCreator (talk) 04:24, 8 March 2010 (UTC)Reply
In the game between Nakamura and rybka we don’t know who was running the rybka? What settings were used on rybka? What computer was used? Was it a Pentium 1? All these things could make a difference of 100s of points. A public game means all this information is known. When I said 2 years, I meant a recent rybka running on a recent computer. We do not know this for the Nakamura rybka game. We know this was blitz and usual for a top computer to lose a blitz game in 2008. We know in 1994 Kasparov lost to Fritz 3 in Munich in a blitz tournament. Therefore, for a top program on a top computer to lose a blitz game to any human in 2008 is unusual. There must have been a problem with the computer or the program not set properly. Crafty is an important program for programmers to learn and definitely not 1 of the top programs. Crafty is 500 points less than rybka on the ssdf rating list. Show me a public game with 1 of the top 10 programs on quad computer in the last 2 years. Mschribr (talk) 18:18, 8 March 2010 (UTC)Reply

Title

edit

Technically, shouldn't this article be moved to Human–computer chess matches, the en dash indicating an "X vs. Y" sense, as opposed to, for example, an "X-Y hybrid" sense? - dcljr (talk) 20:55, 11 May 2011 (UTC)Reply

More info

edit

Interested parties should check out [1] (an archived version of [2], part of the larger collection [3]) for additional information on this topic. It's a personal webpage—no citations unfortunately—but it might give some ideas of what else could be said about computers playing chess against humans. - dcljr (talk) 21:03, 11 May 2011 (UTC)Reply

Oh, and of course there's always Computer chess. [g] - dcljr (talk) 21:06, 11 May 2011 (UTC)Reply

List of matches

edit

Moved here from the article:

  • 1989 Deep Thought. Vs G, Kasparov 2775 2-0 Human Wins
  • 1995 Genius Vs G, Kasparov 2008 2.5-1.5 Human Wins
  • 1995 Fritz Vs G, Kasparov 2795 1.5 0.5 Human Wins
  • 1996 Deep Blue Vs G, Kasparov 2775 4 2 Human Wins
  • 1996 FritzVs G, Kasparov 2770 1.5 0.5 Human Wins
  • 1997 Deep Blue Vs G, Kasparov 2795 2.5 3.5 Computer Wins
  • 1998 Rebel Vs Anand 2795 1.5 0.5 Human Wins
  • 2002 Rebel Vs Loek 2714 2 2 Draw
  • 2002 Fritz7 Vs Kramnik 2807 4 4 Draw
  • 2003 Hiarcs Vs Evgeny Bareev 2729 2 2 Draw
  • 2003 Deep Junior7 Vs G, Kasparov 3 3 Draw
  • 2003 X3D Vs G,Kasparov 2800 2.5 2.5 Draw
  • 2005 Hydra - Adams (2005)
  • 2006 Deep Fritz10 Vs Kramnik 2800 4 2 Computer Wins by 2 full points.
  • 2007 Rybka Vs M, Hayward 3150. 6 game match. Hayward draws 2.5 2.5=Unafisual Worlds Record.
  • 2007 Krempov Vs Deep Junior 10. Rated 2900. 1 Game Match. Score a draw.
  • 2013 Krempov vs stockfish3. Rated 3200. 6 Game Match Krempov gets a draw

Might be useful. GregorB (talk) 14:21, 30 November 2013 (UTC)Reply

Take back

edit

Can a person beat a program if the person has no time controls and can use other programs and can take back as many moves as they want? I came here hoping to find answer. — Preceding unsigned comment added by 47.72.78.79 (talk) 06:09, 22 October 2017 (UTC)Reply

edit

Hello fellow Wikipedians,

I have just modified 2 external links on Human–computer chess matches. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}} (last update: 5 June 2024).

  • If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
  • If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—InternetArchiveBot (Report bug) 17:35, 8 November 2017 (UTC)Reply

Some links that may be included in the main article

edit

human player rating requirement for engine lost game being allowed, in this article

edit

I am aware on repeated reversal of contribution about recent online chess games won by humans against recent version of Stockfish engine.

There is a mixture of disdain for the youtube references, I think twice made for the reversal, and of something else I would like to elucidate here with the help of many, not just the reverters. The last version of the contribution included the reproducible or factual online link to the uniqueID of the games played, on lichess, which is quite good at keeping clean data, and keeping it open, I think. The source code being open, offers for some garranties of being able to find the exact constraints of engine play.

Quote from last revert header: "that is youtube fancruft... nothing like the big kasparov matches..." This seconde revert is using condescending language, indicating some loaded intentions going further than the argument being used. That is a minor of this post here, but it pushed me to intervene, to figure out if the other separate argument about the players on the human side of the games not being champions of history, is really a blocking point. If the contribution were to keep only the lichess references, would it still fail on the reputation of human player argument?

So I bring this behind the scene apparently hot-head inducing question to the many possible onlookers here in the talk page. so that level headed discussion can happen, and at least something visible about the contribution remains.

The question is, where in the title of the article is it said that only the past pantheon of chess players could be worthy of being mentioned here. I would think that not only champions are able to find holes in engine play, as such engines are often optimized by engine competition only purely competitive measures. The question of engines going highly competitive in some corner of chess space beyond human ability to discern how biased that might be, is important to me, and I see any human game as possible data for the question of the title.

I do think that the engine constraint need to be made explicit about the games, but that can be said of many past examples. I wonder if the problem might be about the online or digital aspect of the game being referenced, beside some hidden requirement about player historical standing, or competition pyramidal level.

Here is the last contribution that has been dropped per quoted intervention mentioned up here. https://en.wikipedia.org/w/index.php?title=Human%E2%80%93computer_chess_matches&oldid=1078973247#Stockfish_defeated_in_quick_online_matches_(2020-2021) Dbague (talk) 15:59, 24 March 2022 (UTC)Reply

The addition violates WP:NOR. Links to Youtube matches from personal Youtube accounts and Lichess links do not qualify as WP:RS. OhNoitsJamie Talk 16:16, 24 March 2022 (UTC)Reply
"Lichess links do not qualify as WP:RS"
I could guess for youtube, but i am not that clear on Lichess, using my common sense and superficial (for now) understanding of the wikipedia policy. I will keep reading, but maybe a link to a lichess unique ID, that is meant to keep accessible and unique for as long as lichess organization is online, is not a source because it is not text? and would require further manipulation to get to the game conclusion? Then perhaps the link should become an API call from public browser URL to obtain the PGN text file (a chess standard) which does contain the automatic termination result in plain text (open source coded, lichess implemented chess rules of termination)j. I apologize if this is expected common sense. otherwise please explain more. thanks. Dbague (talk) 18:01, 24 March 2022 (UTC)Reply
I'm not explaining more; that's why we have policy pages, which I've already linked numerous examples of. You're welcome to post here at the original research noticeboard if you'd like more people to tell you the same thing. OhNoitsJamie Talk 18:56, 24 March 2022 (UTC)Reply
The problem is, that that are all primary sources. You cannot find a youtube video and a game on lichess and add it to the article. You need a secondary source talking about this. Like an article on chessbase etc at least some media websites. This 15 second game is a random game streamed online... it's not a big "human-compute chess match" fitting this article. -Koppapa (talk) 05:02, 25 March 2022 (UTC)Reply
There are 3 elements. to the objections. I was trying to separate them. If not using youtube reference but still the lichess game as a game. I see perhaps "match" is the problem, as in english it means a series of games. I forgot. I think both I and the contributor may have missed that.
I understand also that the article in wikipedia, per policy, needs to be about a intermediate first commentary of some sort, and not raw data, even if textual, and factual. This is not about journal-ling all the games. Matches do offer some sampling size which would indicate significance. I think this is more understandable than the policy being invoked about reliable source, which when read appears to be subjective (the line where it might stop is). What is a publisher. I don't think a policy can be self-explanatory always. But I understand. I think the contributor might too. I do reject the obvious no need to explain arguments used about using lichess game id as reference. That would need discussion, I believe. and here would be the place... not in some reverser space. I did ask if youtube was out, what objections would still be left. I also don't think the question is about champions of human chess, as the matches themselves againt engine by same human, should be already an interesting result to share under this title. But we don't have that. So I hope not to have wasted too much vertical space. thanks for the last replies.. There is enough arguments there for me, not all of them, but enough. Dbague (talk) 05:20, 25 March 2022 (UTC)Reply