Wikipedia talk:WikiProject Chess/Archive 22

Schachgesellschaft Zürich – the world's oldest chess club
Is anyone who can read German interested in writing an article on Schachgesellschaft Zürich, the world's oldest chess club (founded in 1809)? Their website doesn't have much information, but the de.wikpedia page is pretty good: Quale (talk) 01:51, 18 August 2012 (UTC)
 * http://www.sgzurich.ch/index.php – official web page (in German)
 * Schachgesellschaft Zürich
 * http://www.chessbase.com/newsdetail.asp?newsid=5582 - chessbase.com article on the 200th jubilee
 * http://www.chesshistory.com/sgz/ - brief history by Richard Forster (in German)
 * Richard Forster (2009), Schachgesellschaft Zürich. 1809 bis 2009


 * Prior to the 200th Jubilee, there was also a 175th, held there in September 1984 and won by John Nunn - some basic info - Nunn 6/7, Bellon, Gheorghiu, Korchnoi, Seirawan, Sosonko, Spassky 5.5, Hort, Kindermann 5, Ekstrom, Forintos, Hess, Keller, Tatai, Zuger 4.5. Herzog, Hug, Wirthensohn 4 … (7 round Swiss, 22 players took part) Brittle heaven (talk) 02:06, 19 August 2012 (UTC)

Pawn duel
Should Pawn duel be in the chess project? It is really a variation of nim. Bubba73 You talkin' to me? 01:04, 4 August 2012 (UTC)
 * Yep, but it's low-priority. I would say it's similar to the knight's tour in importance.--Jasper Deng (talk) 01:09, 4 August 2012 (UTC)


 * It is in the category of chess variants, but I don't think that is appropriate. The pieces don't move like chess pieces and could be anything. At least the knight moves like a chess piece.  Bubba73 You talkin' to me? 01:11, 4 August 2012 (UTC)
 * I concur that this is not a chess variant.--Jasper Deng (talk) 01:13, 4 August 2012 (UTC)


 * Hexapawn is about the same, except that the pieces do move like chess pawns. Bubba73 You talkin' to me? 03:01, 4 August 2012 (UTC)
 * I wouldn't consider it a full variant - again, it's closer to nim than chess. But then what is it?--Jasper Deng (talk) 03:16, 4 August 2012 (UTC)


 * Seems like a math puzzle to me. That uses a partial chessboard and pawns equipment as convenience. (Could just as well use coin, go stone, Halma pawn, meeple, or generalized token). What even slight meaningful connection to chess here? None that I see. Ihardlythinkso (talk) 07:39, 4 August 2012 (UTC)


 * To me both Pawn Duel and Hexapawn are more mathematical than chess. Pawn Duel has nothing to do with chess, but in Hexapawn at least it moves like a chess pawn (except for the initial 2-square move, en passant, and promotion).  Since I first listed Hexapawn above, though, I found that it is listed in Pritchard's Encyclopedia of Chess Variants.  Bubba73 You talkin' to me? 16:09, 4 August 2012 (UTC)


 * OK I have removed the "WikiProject Chess" banner for now. Please feel free to revert if this is not appropriate. SyG (talk) 08:05, 19 August 2012 (UTC)


 * The Chess variants section has plenty of variants where pieces do not move like chess ones: circe chess, checkers chess, chessence etc. So following this argument all those variants need to be removed, which does not sound right. The game can be set up using standard chess set and has the same nature. Okkay (talk) 00:34, 26 August 2012 (UTC)

Glossary of computer chess terms
I created a new page, Glossary of computer chess terms. I started by thumbing through just a single source so it needs a lot of expansion. If you are interested in computer chess and would like to expand or otherwise improve the page, please have at it. Anyone else is also welcome to improve my turgid prose or improve the wikilinks even if you are not interested in the topic. Quale (talk) 04:25, 20 August 2012 (UTC)

Conversation from talk:chess re history of chess
See Talk:Chess The chess article's talk page was previously trancluded here. I converted the transclusion into a link to avoid all those categories from showing up on this page. jonkerz ♠talk 08:38, 28 August 2012 (UTC)

Arturo Liebstein
Is he a notable enough player to have a biography? While he is a three-time Uruguay champion, (not exactly a chess hotbed) the only tournament records that are available have him scoring incredibly poorly at some smaller tournaments, where only a few names are instantly recognizable. (With scores like 4/16 and 3.5/14, second to last or tied for last) I can't even find (nor does the bio provide) any clue on when and where he was born, or whether he is even dead or alive. On a tangential note, however, I think it's safe to say that like Miguel Najdorf, he was a Jew who escaped from Europe to South America during WW2, especially since his real name is Isaac Liebstein. ChessPlayerLev (talk) 16:53, 28 August 2012 (UTC)
 * Normally I'm in favor of keeping bios of national chess champions, even when the country is not a chess world power. If there are good references, then I believe in WP:BTW, and building the web requires that links have targets.  I take a broader view than most editors here about bios of minor players because I think even a small scrap of information about a Uruguayan champion such as date or place of birth might be of interest or use to someone.  However in this particular instance I don't see any good references, so I can't say that it should be kept. Quale (talk) 01:36, 31 August 2012 (UTC)

George Henry Mackenzie Article
I added quite a lot to this article, but would like the help of other interested editors. Firstly, I am wondering if anyone is interested in doing a quality assessment of the article at present (I think it is up to C-class now), and also whether "Mid-Importance" should be changed to "High-Importance" considering that Mackenzie was a world-class player in his heyday? In addition to this, there are some disputed parts of Mackenzie's biography with different sources claiming different facts. In particular,


 * 1) Did Mackenzie die of pneumonia or tuberculosis?
 * 2) Did Mackenzie immigrate to the United States in 1861 or 1863?
 * 3) Did he win the handicap tournament against Adolf Andersson in 1861 or 1862?

I'm also going through the various chess notes that Edward Winter has written about Mackenzie and adding anything that I find relevant. Thanks a bunch to any editors that want to help. ChessPlayerLev (talk) 03:50, 19 August 2012 (UTC)


 * Many thanks to Bubba73 for doing a quality assessment. As for the much-disputed question of Mackenzie's death, Edward Winter wrote a fantastic piece about it.  Maybe it's only interesting for me (after all, it's just a biographical detail for a strong, but not immensely famous master) but I definitely encourage people to check it out.  If nothing else, it shows how difficult it can be to determine precise historical facts, even from newspaper accounts of that time period.  ChessPlayerLev (talk) 00:53, 3 September 2012 (UTC)

Chessmetrics
We have User:BrashBrannigan's opinion here, and (on the other hand) we have User:ChessPlayerLev nearly systematically and single-handedly removing Chessmetric rating evaluations from articles Isidor Gunsberg, George Henry Mackenzie, James Mason, and Charles Jaffe recently.

Question: Isn't this (Chessmetrics includability/value/reliability) a matter for ProjChess to discuss and form consensus on (like it has on, say, includability of Fritz and other computer-generated analyses in articles), before one user unilaterally removes all traces of Chessmetric evaluations from articles? (I don't have opinion about Chessmetrics one way or the other. Am inquiring about process here, for those who might.) Ok, Ihardlythinkso (talk) 18:52, 24 August 2012 (UTC)


 * Yes. Chessmetrics is the best objective measure we have for ratings before ELO.  Also, it uses all available data for players.  Like any rating system it has problems comparing players of different eras but I think it is good for comparing to players of the same time (e.g. Chessmatrics pput him at #9 in the world from ...).  Bubba73 You talkin' to me? 19:43, 24 August 2012 (UTC)


 * Chessmetrics ratings have been discussed here a few times before. I think the first discussion relevant to your points was in June 2008:  WT:WikiProject Chess/Archive_12.  In May 2010 I raised some issues I saw with the way that we were using chessmetrics, in particular articles that suggested that Chessmetrics and FIDE (Elo) ratings are directly comparable. The thread Wikipedia Talk:WikiProject Chess/Archive_18 attracted some discussion.  Since we were (and still are) using chessmetrics in some ways that I consider not valid, I was a scold again in April 2011: Wikipedia talk:WikiProject Chess/Archive_20.
 * I think Sonas (the creator of Chessmetrics) is an expert in chess rating, and he has been consulted by FIDE concerning ratings including the recent change in k-factor. I think Chessmetrics are useful for players whose peak occurred before the introduction of FIDE ratings and can be used in the way that Bubba73 describes.  The best brief recap of the facts as I see them was by Peter Ballard in 2008:
 * "Chessmetrics is very good, but (1) it is unofficial, (2) it is the work of one person, and (3) it could change if Sonas thinks of improvements in the future. For these reasons, I think Wikipedia should freely use Chessmetrics, but always make it clear that is a chessmetrics result, not an absolute result or ranking. So for instance, 'Geza Maroczy was the the world #1 according to Chessmetrics between 1904 and 1907' is OK; 'Maroczy was the world #1 between 1904 and 1907' is not. Peter Ballard (talk) 04:21, 27 June 2008 (UTC)"
 * I don't know if Sonas will take up historical ratings again in the future; he seems to have quit in 2005. He may consider his work complete. Briefly, I don't think chessmetrics ratings should be excised from Wikipedia.  I do think claims that chessmetrics ratings show that a player was of GM strength must be removed, and I agree with Peter Ballard that any chessmetrics ranking must be clearly labeled as a chessmetrics ranking, not a world ranking.  Also chessmetrics ratings and rankings must not be used for players in the modern rated era.  Quale (talk) 22:14, 24 August 2012 (UTC)


 * Thanks for all that context, Quale. Is there enough data here for Proj Chess to form a consensus on policy convention vis-a-vis use of Chessmetrics in articles? Can someone summarize or propose a clear summary? Ihardlythinkso (talk) 05:51, 25 August 2012 (UTC)


 * Regardless of consensus on its use, Chessmetrics is not the same as FIDE Elo ratings, and has nothing to do with the GM title (which is won through norms, not any form or rating) or "world rankings". These are the only things I have edited out in the articles mentioned above, and which I note below.  However, on top of this, I would definitely like to see the use of Chessmetrics significantly reduced in articles to no more than a mere external reference or short note.  This is an encyclopedia, and should include specific accomplishments and sourced statements, not some crazy rating system a guy cooked up in the early 2000s and then completely abandoned, which has worse predictive power than Elo and ranks Lasker 12th in the world in 1914. ChessPlayerLev (talk) 06:09, 25 August 2012 (UTC)

Look, I don't have a problem with keeping Chessmetrics as a reference or short note in the biography of pre-WW1 chess players. (Although I will mention my problems with their algorithm at the end) If you will notice, many of those articles still retain some mention or reference to the site. What I DO have a problem with are the following;

Now, my problems with the algorithm. Chessmetrics is a form of rating that is uniquely unsuited to judging old players. Anyone who has studied chess history will immediately see this. Go ahead and read their methods first.
 * Using Chessmetrics interchangeably with Elo ratings. This is simply wrong.  Chessmetrics is completely different than Elo, and should never be confused with it.  Chessmetrics gives different ratings based on whether someone was White or Black, and strongly penalizes players for lack of playing activity.  It was instituted much later than Elo for players who never had Elos.  (And Dr. Arpad Elo's guesses for the ratings of certain historical players are completely different than those of Chessmetrics in almost every instance)
 * Using Chessmetrics to assign world rankings. This is foolish.  Saying someone was "#6 in the world" at one time because of Chessmetrics is pretty silly, especially when this changes to "he was #14 in the world a few months later" (during which he might not have even played any tournaments!).  Really, guys?  Even in modern chess, this doesn't happen so fast. As a lovely example mentioned on the Chessmetrics Wikipedia entry itself, Emmanuel Lasker, the reigning, dominant world champion who finished 1st or 2nd in every major international tournament except Hastings 1895 (when he was recovering from pneumonia) for 45 years, was ranked 12th in the world in early 1914, right before his win at St. Petersburgh 1914.  And people actually take Chessmetrics seriously as a form of rating?! We're talking about a guy who was still considered top 10 by actual reputable chess sources in 1934, at the mere age of 65.


 * Players have their older games factored in far less in the algorithm, both in terms of how many games you played since then as well as how many MONTHS ago they were! (50% for 2 years ago, 2% for 4 years ago) Before World War I, there were very few major international chess tournaments.  Guys like Lasker, Tarrasch, and even a slightly later player like Capablanca would play maybe 1 major event a year.  In some cases, they would play 1 major event every 2-3 years!  All the older players have very few available games compared to modern players.  By implementing an algorithm that places overwhelming emphasis on recent results when everyone's playing activity was exceptionally low by modern standards, you punish the actual better players, and reward the inferior players!  Steinitz doesn't play for 24 months?  Huge drop in the rankings!  But oh, his rival Mackenzie played twice during that span, so he is the new number one, ranked ahead of Steinitz!  Never mind that Steinitz obliterated Mackenzie in their individual games  (+6 −1 =3) and finished ahead of him in every single tournament they ever played in together.
 * They way they "flatten" performance ratings makes players appear far closer in rating range than they really are! Imagine a hypothetical tournament where there are 10 entrants.  They are all unrated.  8 of them has a "real" rating strength 1400.  Another one has a real rating strength of 1700.  Another has a real rating strength of 2000.  If they played against one another, not only would the 2000 and 1700 players likely have very similar results (making it hard to tell them apart), but one of the real 1400s would do well and have a close rating to them as well!  Now, this isn't a problem in major international chess tournaments today, but it was a huge problem back in older tournaments.  There were a LOT of really weak players who were just free points for the established masters.  Even Hastings 1895, one of the strongest tournaments ever, especially for that period, featured a "Beniamino Vergani" who probably plays at the strength of a modern 1300 Elo player.  In fact, that might even be generous, when you look at his games.
 * The predictive power of Chessmetrics is astonishingly bad, for all these reasons and more. For instance, Tarrasch played Carl Walbrodt in a match in 1894.  Anyone who has seen these games knows what a one-sided beating they were.  Tarrasch destroyed Walbrodt 7.5-0.5, but even that doesn't do the actual games justice.  This was worse than a 2000 rated player beating down on a 1400 player.  I mean, these guys were HUNDREDS of points apart in actual chess strength at the time.  What does Chessmetrics predict as their relative ratings and what the result should have been?  Why, it assigns Tarrasch a rating of 2753 and Walbrodt 2689, predicting a razor-close match which Tarrasch would have won by the skin of his teeth; 4.6-3.4.  Most insultingly, Walbrodt is ranked #5 in the world on little more than playing activity.  Again, you will laugh if you have ever looked at the guy's games and are a decently strong player yourself.

Anyways, Chessmetrics is not a reliable source, and while I am okay with putting in it at present, I would eventually like to see it fazed out. I have recommended this to Quale already, but if you want a much better way of ranking old players check out GM John Nunn's section "The Test of Time" in "John Nunn's Chess Puzzle Book". Not only was Dr. Nunn a top-15 chess player in his prime, but he is also a doctor of mathematics at Oxford who used to be a lecturer there. ChessPlayerLev (talk) 23:34, 24 August 2012 (UTC)


 * I'm not making a blanket defense of chessmetrics, but I have some disagreement with a couple things you've written. On your claim that Chessmetrics "has worse predictive power"—I'm not familiar with the paper, but someone published something that said that according to their study chessmetrics ratings were more accurate at predicting outcomes than FIDE/Elo ratings. I think it's in the talk archives.  As I recall the study was concerned with determining whether Soviet players had colluded.  I'm not vouching for the value of the study, just saying it's out there, and I'm not certain that the contrary claim that chessmetrics has worse predictive power has been demonstrated.  (I actually argued in that old talk that a single study perhaps not peer reviewed shouldn't be accorded too much weight, but I never thought it should be completely ignored.)  Second, Jeff Sonas is not "some guy".  He's a recognized expert in chess rating, and FIDE consulted him when they considered adjusting the K-factor.  Ultimately FIDE did make an adjustment, and I think Sonas' suggestions played a part, along with the other people they consulted including John Nunn.  User:Philcha was the strongest defender of Chessmetrics in the chess project.  I don't think he shared my concerns over what I consider to be invalid uses of chessmetrics ratings and rankings in articles, but I wish he could add his two cents here. Sadly, he won't be able to participate any longer.  Quale (talk) 07:27, 25 August 2012 (UTC)


 * If there is a paper about Chessmetrics' predictive power, please link me to it. I didn't find anything in the Talk archives, and can't respond to something only vaguely ("someone published something") alluded to.  I have clearly given my reasons for why I think their predictive power is awful.  Feel free to respond to my arguments.  Also, are we really calling people "experts" now because FIDE consulted with them?  This is the same organization that | didn't "accept" the resignation of a corrupt vice president (FIDE presently has 5 of them!) who banned multiple countries from the 2012 Olympiad because they were suing FIDE...while HE WAS SUING FIDE HIMSELF.  The organization is run by a man with ties to the criminal underworld (very nearly convicted of murder of a journalist), is endlessly corrupt, and routinely makes horrible decisions about rules, tournaments, time controls, and championships.  Now someone they consulted once is anointed "expert" status?  ChessPlayerLev (talk) 09:54, 25 August 2012 (UTC)


 * Look in Wikipedia talk:WikiProject Chess/Archive 12 for mention of the paper that discusses the quality of chessmetrics in predicting outcomes. (I linked that section early in this thread.) While I don't disagree with your characterization of FIDE, you've definitely left the reservation now.  Talk is fairly wide open as long as the discussion relates to improving wikipedia, but inside articles WP:V and WP:RS determine article content.  Your (or our) personal feelings about FIDE don't have any place inside articles, which must be based on what is published by reliable sources and also avoid undue weight.  I think you have made a strong and mostly valid critique of chessmetrics ratings, and I don't have any need or desire to try to refute all of it.  In fact I agree with most of it.  My two main points of disagreement are that I don't think you have proven that chessmetrics has "worse predictive power"—just because you claim it doesn't make it so,  It would require a fair amount of research to determine that one way or the other.  Your characterization of Jeff Sonas as "some guy" is just intemperate and needless.  When you have a good argument I think you just make it weaker by making reckless claims that aren't necessary to demonstrate your point.  Unfortunately this is just seems like a flashback to my first encounter with you on Talk:Judit Polgár ("blatantly dishonest", etc.)  Quale (talk) 21:01, 26 August 2012 (UTC)


 * My point was simply that being consulted by FIDE does not confer expertdom on its subject. That's all.  It truly is a joke of an organization on every level, a statement you seem to agree with.  (And for that matter, so have dozens of GMs from the 60s onwards)


 * As for the vaguely alluded-to paper ("someone published something"), the only thing I found was | this. Unfortunately, I cannot download the paper unless I pay for it.  Also, the abstract tells me nothing related to your claims.  For one, it makes absolutely no mention of Chessmetrics.  Secondly, it examines chess events from 1940-1964, which is a completely different animal than the pre-WW1 era I was describing above.  Have you read the article yourself, Quale?  If you did, perhaps you can be more specific on what this paper has to do with Chessmetrics and its predictive power.


 * And of course I haven't "proved" that Chessmetrics has awful predictive power. I simply wrote my personal impression.  However, neither has anyone "proved" that Chessmetrics does NOT have awful predictive power and that it's not just another random rating system someone cooked up in a few hours on their computer and which subsequently received a little bit of attention.  I know reliable sources are hard to come by when it comes to historical chess research, and ranking many old masters is exceptionally hard work.  Thus, I totally understand the appeal of Chessmetrics.  It's an easy way out.  Rather than scrutinizing games, tournament results, matches, and trying to find accurate sourced statements, you can just consult Chessmetrics instead and report their findings!  Unfortunately, what you get back is hardly accurate. ChessPlayerLev (talk) 21:51, 26 August 2012 (UTC)


 * You've got it exactly right with "trying to find accurate sourced statements". The important question then is whether chessmetrics can be used as a reliable source.  I think parts of your critique rest largely on your personal opinion.  Your opinion could be correct, but I don't think we have any valid reason to suppose that we should value your opinion over that of Jeff Sonas.  Suppose that Jeff Sonas were to write a rant, roughly equivalent to yours, calling ChessPlayerLev "just some guy"—how should we decide who was right?  Now it is not necessarily necessary to impeach Chessmetrics in order to conclude that it doesn't meet Wikpipedia's requirements for a reliable source.  Aside from the single paper claimed, Chessmetrics may lack independent verification.  To the paper in question, I have not read it nor have I ever claimed to have read it.  (I thought that my original statement was fairly clear on that point: "I'm not familiar with the paper, but someone published something that said that according to their study chessmetrics ratings were more accurate at predicting outcomes than FIDE/Elo ratings.")  If the paper doesn't say anything about chessmetrics then clearly the paper is irrelevant.  If the paper makes the claim reported by another Wikipedia editor, then it is relevant, even if you or I believe it may be incorrect.  In that case we would need a reliable source to impeach it. Quale (talk) 03:55, 27 August 2012 (UTC)

Like most things written on the Talk Page, (as opposed to the articles themselves) my thoughts on Chessmetrics are my own opinions. However, I believe they're quite logical, and you admit to mostly agreeing with them. If anyone disputes them, they are free to elucidate any errors in reasoning I made above. Personally, I'm still waiting to hear on any reasons (opinions) on why Chessmetrics is any good besides "Jeff Sonas was once invited to a FIDE conference" or "it's easier and more convenient than doing real research".

As for the paper, you originally claimed that it showed something about Chessmetrics's predictive power, and demonstrated one of my statements wrong. Now you admit that you haven't read it, and in fact, have no clue if they even mention Chessmetrics, or in what capacity. This is also completely ignoring it that has zero relevance to what I wrote above, since they cover the period from 1940-1964, not the pre-WW1 era. We also can't go by heresay on what one editor claims the paper said if we can't actually look at the source itself, read the whole thing, and make up our own minds. ChessPlayerLev (talk) 04:23, 27 August 2012 (UTC)


 * I too, mostly agree with CPL. Never used Chessmetrics, never will. But having not been officially discredited, it exists whether we like it or not, for the inexperienced/lazy editor's convenience, or in some cases to ram home a point of nationalistic pride. We can take a strong stand against the latter; if an article already contains sufficient sourced material to prove a point, then a chessmetrics entry to further reinforce it could be seen as excessive and contrary to NPOV considerations. Similarly, if a brief biography contains an excessively long passage on Chessmetrics’ findings then that might also be regarded as excessive. I'd have no problem with removal or part removal in either of these cases (of course, using Chessmetrics to contradict an expert opinion would also be incorrect unless in support of a rival expert claim). I am reminded of other wiki policies which for example take a dim view of stating and restating a person's country, their jewish background, of overly displaying their flag (or the wrong flag). It is important that we remain vigilant against inappropriately motivated inclusions and undue weight. I can however see how statistics may be a useful substitute where there is otherwise a void, or to lightly back up an expert opinion. As a second point, I'd personally recommend that articles reflecting substantially post-1970 contexts should stick with Elo and exclude Chessmetrics - I can't see the sense in having two entirely different entities, with confusingly similar numbers running side by side – that must be wrong. As for ‘predictive’ uses, I’m not sure if any articles say stupid things like Fischer would have beaten Capablanca in a match because Chessmetrics says so, but it should be easy enough to dismiss such uses as ‘not encyclopedic’. Probably the only exception to these recommendations would be the article Comparison of top chess players throughout history where necessarily everything is thrown into the mix, but of course without any conclusion or opinion on validity/accuracy. Has anyone any issues with these suggested principles? Would they help reduce excessive use of Chessmetrics? Could they go further? Do they need to? Brittle heaven (talk) 15:17, 27 August 2012 (UTC)


 * Yeah, Brittle heaven brings up a good point; most times when I see Chessmetrics used, it's as an argument for how amazing a particular player was, out of all proportion with their actual accolades and abilities. And Chessmetrics makes this exceedingly easy; as a type of fluid performance rating that discounts anything older than 2 years, if a player has a run of one or two good tournaments (as most masters, of any level, will tend to do now and then), their Chessmetrics rating will rank them higher than they ever really were.  ChessPlayerLev (talk) 00:36, 28 August 2012 (UTC)


 * Am completely neutral as mentioned (no opinion since never researched Chessmetrics in order to have one [an opinion]), tho CPL somehow misconstrued and accused of carrying a negative opinion on his recent edits of something I know nothing about (already admitted). My single suggestion was re the benefit of having a ProjChess convention re application of Chessmetrics in articles. (Then, changes, e.g. like what CPL has initiated, could be evaluated for consistecy with group goals, and any editor can have confident base to know if consistent w/ ProjChess standard or not.) Quale, when all data is in could you summarize a simple bulleted list what are the convention criterion? Thx for consider. Ihardlythinkso (talk) 19:35, 27 August 2012 (UTC)

Ihardlythinkso, as I mentioned above, my edits to the articles above are completely independent of what one thinks of Chessmetrics. I didn't completely get rid of Chessmetrics in those articles, but only in those instances where it is used in a factually inaccurate manner;


 * I got rid of statements that used Chessmetrics ratings interchangeably with Elo ratings (factually wrong).
 * I got rid of statements using Chessmetrics ratings to make arguments about someone being "GM strength" (factually wrong again; the GM title is awarded based on norms, and not any any type of rating, even Elo).
 * Finally, I got rid of statements using Chessmetrics to make statements about "world rankings", as if they were some official tracker. (factually wrong a third time)

Even if the discussion here concludes that Chessmetrics is the greatest rating system ever devised by man, it won't affect the particular edits I made. So while I welcome this conversation to state my views on the rating system and limit its implementation beyond the edits I made, the nature of my edits so far is unrelated to this. ChessPlayerLev (talk) 00:36, 28 August 2012 (UTC)


 * Just one last note by me here, and I should be done in this thread. I agree with #1 and 2 in ChessPlayerLev's list—both of those are poor practice and just wrong.  For #3 (rankings) I'm more ambivalent, as I've argued in the past that this is a better use than the others.  Really the Chessmetrics numbers aren't very useful alone.  If you say a player's historical chessmetrics rating peaked at 2678, what does that number mean?  Nothing, really.  You need something to compare it too.  But Lev's point that the Chessmetrics world rankings are too volatile is important.  I'm also remiss in that I didn't really pay attention to the way that Chessmetrics is used in Emanuel Lasker.  It appears in the references about fifteen (!) times, which is at least ten times too many in my opinion.  In many places on that page it's used to try to indicate how strong Lasker's opponents were, which I think is not very helpful.  Fortunately I haven't seen any other pages rely as heavily on Chessmetrics; the bad things on other pages tend to be the issues that Lev has been correcting. Quale (talk) 05:35, 28 August 2012 (UTC)


 * "Chessmetrics numbers aren't very useful alone. If you say a player's historical chessmetrics rating peaked at 2678, what does that number mean? Nothing, really. You need something to compare it too." I agree; is this an example what you mean, changing this:"According to Chessmetrics, at his peak in October 1876 Mason's play was equivalent to an Elo rating of 2715, and he was ranked number 2 in the world, behind only Wilhelm Steinitz."to this:"According to Chessmetrics, at his peak in October 1876 Mason's play was equivalent to a Chessmetrics rating of 2715, and he was ranked number 2 in the world, behind only Wilhelm Steinitz." as was done by this edit? Because the latter doesn't make any sense to me. Ihardlythinkso (talk) 19:56, 1 September 2012 (UTC)


 * What part doesn't make sense to you? There is a type of rating called the Elo rating system, named after its founder, Dr. Elo, which FIDE uses.  Then, there is the rating system used by Chessmetrics which has virtually nothing in common with it, and no national or international chess organization uses.  Speaking of which, I need to go back and edit that entire section anyways; using Chessmetrics to assign "world rankings" is lazy and inaccurate.  ChessPlayerLev (talk) 05:03, 2 September 2012 (UTC)


 * The part that doesn't make sense to me is: "equivalent to a Chessmetrics rating of 2715". (If the rating is a Chessmetrics rathing, then how could it be "equivalent" to anything? The word "equivalent" was previously in the text as a compare between Chessmetrics and Elo. You took the Elo mention out, so "equivalent" refers to nothing.) Please read Quale's comments about having a base point for compare. (I think it's necessary, too.) Okay, I'm getting confused ... let me try again ... Ihardlythinkso (talk) 06:05, 2 September 2012 (UTC)


 * I was reading it wrong. (Sorry.) You are saying in the text Mason's play was "equivalent to a Chessmetrics rating of 2715". Okay, two things: That is "round-about" (and confusing) language, and should be (can be) simplified to simply say: "According to Chessmetrics, at his peak in October 1876 Mason's play was equal to a Chessmetrics rating of 2715", or better and simpler: "According to Chessmetrics, at his peak in October 1876 Mason's play was equal to a rating of 2715." But in any case, this has no meaning to the reader (does it?), unless it is an implicit compare (equivalency) with Elo. (Else, you are putting on the reader to understand that Chessmetrics is something special and different rating system from Elo. But, isn't Chessmetrics intention to estimate the Elo rating? So to suggest it is different but not an intentional estimator of Elo, ... that's confusing!) Ihardlythinkso (talk) 06:20, 2 September 2012 (UTC)


 * (What was wrong with the first way before you changed it: "According to Chessmetrics, at his peak in October 1876 Mason's play was equivalent to an Elo rating of 2715"? Because that seems clear Chessmetrics is nothing more than an estimator of equivalent Elo.) Ihardlythinkso (talk) 06:30, 2 September 2012 (UTC)

Once again, you are not reading what is written. Like I have written no fewer than 3 times above, Elo is NOT the rating system that Chessmetrics uses. The statement "it seems clear Chessmetrics is an estimator of equivalent Elo rating" is simply wrong. Chessmetrics is its own completely unique rating system, not the one Dr. Elo pioneered and a version of which is used by FIDE, the USCF, and other national/international chess organizations to this day. I don't know how many times I can repeat this. The reason I specified "Chessmetrics" a second time in "Chessmetrics rating" is because "rating" is understood by most readers to mean "Elo rating". In fact, it's practically synonymous when it comes to chess. As for the use of "equivalent to", that's not even my edit. That was there before I ever got to the page. You can change it to "equal to" if you think it makes a difference. ChessPlayerLev (talk) 06:43, 2 September 2012 (UTC)


 * I think you are saying that Chessmetrics does not endeavor to estimate Elo ratings. You mention "rating" by itself means "Elo" to readers. Right, agreed. The original sentence read: "According to Chessmetrics, at his peak in October 1867 Mason's play was equivalent to an Elo rating of 2715 [...]". (That was User:Krakatoa's contribution. Was he wrong?) If you write: ": "According to Chessmetrics, at his peak in October 1876 Mason's play was equivalent to a Chessmetrics rating of 2715", that is circular and redundant, and doesn't make sense. (Like saying, "According to John, John said he was hungry for a taco." Which should be simplified to: "According to John, he's hungry for a taco." Or in context: "According to Chessmetrics, at his peak in October 1876 Mason's play was equivalent to a rating of 2715." If we both agree "2715" means "Elo", and User:Krakatoa added that specific language including "equivalent to", then it satisfies your requirement that no one is saying they are the same exact thing, and the sentence then makes sense. ("Equal to" or "equlivalent to" is just causing confusion in this discussion, so please ignore that I brought up word "equal".)


 * If you still don't understand me, I think at this point User:Quale needs to bridge the communication confusion, since my understanding is I'm echoing his same message above about a base reference point to give meaning to any Chessmetrics rating quoted in an article. (It has to be relative to something, and can't be relative to Chessmetrics, since that's circular.) Anyway, who knows maybe I'm getting confused too ... Ihardlythinkso (talk) 09:19, 2 September 2012 (UTC)


 * Isn't part of the confusion here semantics? From Chessmetrics: "It is intended as an improvement over the Elo rating system." In other words, "more accurate than". (Not a different animal, just a more precise one.) So, in context: "According to Chessmetrics, at his peak in October 1867 Mason's play was equivalent to a refined/improved/reworked/better-rendered Elo rating of 2715". (It's still a basis/reference point to Elo, not some other animal, just better defined, more accurate, etc.) But we need others' input now ... Ihardlythinkso (talk) 09:39, 2 September 2012 (UTC)


 * Yes, Krakatoa was wrong with that statement. No, it's not a question of "semantics".  It's nice that you read the brief blurb about Chessmetrics, but considering that you started this entire discussion, it's amazing that you haven't even read up on how either Chessmetrics or Elo ratings work!  I can't emphasize this enough; you started this topic asking a question about rating systems, particularly Chessmetrics, and after making a number of replies about it, you STILL haven't read how they work!  What was the point then, Ihardlythinkso, if you have zero knowledge and zero interest in learning about the topic you're so passionately arguing about?  Saying that Elo ratings and Chessmetrics ratings are merely different versions of one another is like saying that Elo ratings and number of first place tournament finishes are different versions of one another since hey, they're both numbers that can be used to rank chess players!


 * Summarized very, VERY briefly, Chessmetrics uses a linear regression to calculate "performance ratings" for entire events (not games) and average those with performance ratings within only the last 2 years, giving more weight to more recent results/those with more games. Everything outside of 2 years is ignored.  Ratings change wildly with inactivity, and a single tournament is enough (even if they have never played in another in their life) to make someone #1 in the world, since it's just performance rating.  This isn't a hypothetical example, either; there are lots of instances of something exactly like or similar to this occurring in the Chessmetrics database.  Elo, meanwhile, is not a performance rating, and takes into account the results over a player's entire career.  It uses a table of expected win percentages against players of a certain rating to calculate results from each individual game, adding them up over the course of a tournament.  Ratings do not change with inactivity, either.


 * They're two completely, wildly different approaches to rating players. One is not a version of the other, and using "Elo" to refer to Chessmetrics ratings is like using "dog" to refer to a cat because they're both animals with four legs.  ChessPlayerLev (talk) 10:12, 2 September 2012 (UTC)


 * Okay, they are two wildly different approaches. Got it. I started the topic to understand what consensus is and is not, for use of Chessmetrics in articles. I'm neither "passionate" nor "arguing" on the topic. I'm just wondering what sense this will make to readers: "According to Chessmetrics, at his peak in October 1876 Mason's play was equivalent to a Chessmetrics rating of 2715". First, it seems redundant or circular expression. Second, I believe this is example what User:Quale was referring when he expressed view: "Chessmetrics numbers aren't very useful alone. If you say a player's historical chessmetrics rating peaked at 2678, what does that number mean? Nothing, really. You need something to compare it too." As already mentioned, I think I'm just echoing Quale on his point, with a specific example. (Do you think it *isn't* example what his view is about? Or if you think it *is* example what his view is about, then your answer needs to apply to that user too, not just to me. What is your reply to Quale, then?)


 * Last, I have a question for you CPL, and then a request... Question: Do you think the lead in the Chessmetrics article, "It is intended as an improvement over the Elo rating system", is okay? Or do you think the statement needs an adjustment? Request: Assuming you'll continue correcting use of Chessmetrics in articles, could I ask that your next article for doing so be this article: George H. Gossip. (I think it would be helpful possibly, beneficial, in sorting out and clarifying what are the right or best rectifications there, since, there are relatively lots of eyes on that article, and something either good or bad, clear or unclear, will then likely come out. Thanks for considering.) Ok, Ihardlythinkso (talk) 14:37, 2 September 2012 (UTC)


 * I intend to re-write that entire section dealing with Mason's playing strength in a week anyways. If you dislike the word repetition (which in general irks me too, although I have mentioned my reasons for using it in this case), rest assured it will be changed very soon.  As for your questions, I think the lead-in for the Chessmetrics article is fine.  The key word is "intended".  Yes, the intention was a brand-new, completely different rating system that would be an improvement over the previous Elo.  Now, I think they failed miserably at this, but that was certainly their hope.  Thus, the lead-in is completely accurate.  As for the George H Gossip article, it's quite wonderful, one of the best chess entries I have come across.  (Which we largely have Krakatoa to thank for)  It certainly deserves its FA status and was complimented by no less an authority than Edward Winter.  That being said, my only minor quibble was with the use of Chessmetrics to argue he was better than his dismal reputation.  Out of every single player that regularly participated in international tournaments around the time, Gossip had the lowest rating for most of his career.  Also, the way that Chessmetrics calculates performance ratings with linear regression, the difference between 2700 and 2470 (cited by the article) is like the difference between 2800 and 2200 under the Elo system. (That's a guess off the top of my head, but it's definitely way bigger) The way Chessmetrics is used is far from terrible; they mention Dr. Elo's historical calculation first, some chess magazine ranking in 1904, and give Chessmetrics proper context.  Still, I would say it's improvable.  It's definitely misleading to suggest that Gossip was ever anywhere close to #17 in the world.  I mean, he finished tied for 17th-18th out of 20 at a second-tier international event missing a ton of the world's best players.  And THAT was his best performance ever.


 * Which actually brings me to my last point on Chessmetrics and why it became so popular. What Chessmetrics does is calculate performance ratings for an event and based on playing activity, this can be all or most of a player's rating during a period.  Now, EVERY chess player, including masters, has a string of good tournaments every now and then.  But whereas with Elo, they can only gain so much from it, with Chessmetrics it might be their entire rating!  Thus, every single player gets wildly overrated through Chessmetrics' system.  I have seen players barely cracking the top 100 that had rankings of 40th in the world through Chessmetrics.  If you were actually a fringe top 50 player, you would probably make it to the top 15-20 at one point through Chessmetrics.  And if you were ever a top 10 player, Chessmetrics would probably rank you #1 at one point.  In fact, that's why Chessmetrics, especially in an era of lower playing activity, has so many different #1 players during eras where there was a dominant world champion who would win every major international tournament in sight.  ChessPlayerLev (talk) 16:00, 2 September 2012 (UTC)


 * Fair enough. Quale expressed view, and I agree if I properly understand it, that there's lack of meaning when a reference to Chessmetrics is left "hanging" in an article, without a comparative context to help it make sense to readers. The George H. Gossip article, however, seems to have lots of context, with three different ratings evaluations mentioned. But other articles aren't as extensive in scope or detail. But I still think your change(s) to Gossip would be useful, earlier than later, for distilling out issue(s), if there are any. Ok, Ihardlythinkso (talk) 16:56, 2 September 2012 (UTC)

Category:Chess theoreticians
Is there any real reason for Category:Chess theoreticians to exist? I never thought it was a good idea, but I didn't say anything about it either. Peter Ballard broached that issue four years ago on Category talk:Chess theoreticians and I agree with his comments there. The category inclusion criteria aren't clear enough to know who goes in, with the result that the category is pretty indiscriminate. Nearly anyone who has written about chess could go in (unless they confine their writing to history, which is quite rare), and some seem to think that anyone who gets an opening named after himself belongs in. How else to explain why Preston Ware was put in the category? The Ware Opening and the Corn Stalk Defense don't indicate that Ware was a chess theoretician; rather they suggest instead that he wasn't a very good chess player. Nearly everyone in this category will already be under Category:Chess players, and it's so subjective I don't see any point in it. The odd cases of people like Mark Dvoretsky, Graham Burgess, Jeremy Silman, and the like who are best known for chess-related work other than as players have ample appropriate categories already such as Category:Chess writers and Category:Chess coaches. These categories have the advantage of more objective and clearly defined inclusion criteria. What say you? Quale (talk) 08:31, 14 August 2012 (UTC)


 * I completely agree with both you and Peter Ballard on the matter. (On a side note, it's a shame that Peter Ballard, an outstanding chess editor whose old comments and edits I have found enlightening and helpful, got burnt out and left)  For all the reasons mentioned, this list seems to be a complete waste, and should be deleted.  Virtually every strong player ever can be considered a "theoretician" as their games and analysis alter our chess understanding.  Similarly, every player in history who has played an opening novelty or done an endgame study can also be considered a "theoretician".  It's an extremely vague and often irrelevant label when, as Quale mentioned, far more concrete and relevant ones exist.  I propose deletion.  ChessPlayerLev (talk) 23:02, 18 August 2012 (UTC)


 * Some people are better known as theoreticians than players. Bubba73 You talkin' to me? 23:44, 18 August 2012 (UTC)


 * See, that's my problem with "theoreticians". It's such a vague, unclear, general term.  You are probably thinking of people like IM Mark Dvoretsky or IM John Watson, but we can include them in much better lists like "List of Major Chess Authors" or "List of Important Chess Coaches", which are far more specific and give the reader actual information about what they did.  ChessPlayerLev (talk) 23:55, 18 August 2012 (UTC)


 * Actually I was thinking about people like Alexey Troitsky and maybe Grigory Levenfish. But I'm not against deleting the theoretician category (too vague and subjective). Bubba73 You talkin' to me? 00:03, 19 August 2012 (UTC)


 * Thanks, I think Troitsky is a very good counterexample to my thesis, as he does qualify as a theoretician. Category:Chess composers is more obvious and we could probably stand to have a Category:Chess endgame study composers subcategory, but his contribution goes a little farther.  (If there were a endgame study composers subcategory it shouldn't be mutually exclusive of a general composers category.  I'm sure many composers have done both, but some may have done only one or the other.  I don't know if Richard Réti did any middlegame compositions, but his K+P study is pretty famous even to those of us who know nothing about compositions.)  Quale (talk) 00:10, 19 August 2012 (UTC)


 * André Chéron is probably another, but I agree that there are problems with the category. Bubba73 You talkin' to me? 02:21, 19 August 2012 (UTC)


 * Interesting; I consider Réti and Levenfish primarily as players and secondly as very influential writers. Troitsky is a good example of neither, but as noted, he is a legendary composer and rightly included in that list.  I also think an endgame composer subcategory makes a lot of sense.  ChessPlayerLev (talk) 03:31, 19 August 2012 (UTC)


 * Reti is clearly much more prominent as a player than theoretician, but if it wasn't for the Levenfish & Smyslov book, I might not have heard of Levenfish for many years. Bubba73 You talkin' to me? 03:48, 19 August 2012 (UTC)


 * It's a truly wonderful and amazing book, isn't it? In fact, I want to expand that part of Levenfish's biography, since "Rook Endings" is probably one of the 10 greatest, most influential chess books ever.  Generations and generations of players have studied this book; it's right up there with "Zurich International Tournament 1953", "My System", and (more controversially!) "The Road to Chess Improvement".  As for Levenfish as a player, maybe it's because I was born in the USSR, but he was widely known as a solid GM.  ChessPlayerLev (talk) 07:00, 19 August 2012 (UTC)


 * I agree about Levenfish & Smyslov Rook Endings, but I think that Averbach's Comprehensive Chess Endings volume on rook endings is about as good. I wish it would be reprinted in English and algebraic notation.  My first copy is in algebraic, but it is a paperback from the 80s and the paper is very brittle.  I have some old hardback copies in pretty good shape, but they are in descriptive notation. Bubba73 You talkin' to me? 14:57, 19 August 2012 (UTC)


 * Agree the category as it is, with 165 entries, is arbitrary and meaningless. But isn't that because the interpretation what qualifies as "theoretician" allows each editor's subjective understanding what the word means? I'm thinking the chess literature identifies certain unmistakable theoreticians (like Steinitz, Tarrasch, Nimzovitsch). Fischer and Petrosian probably didn't see themselves as theoreticians, and neither does the literature. I suggest to solve the problem of the current arbitrariness and meaninglessness, not to abolish the category, but to restrict it to what the chess literature unmistakably tells us. (The number of entries in the category, under that definition, would greatly reduce, and have meaning by reflecting what chess literature tells us. Of course there would be argument whether a player still qualified or not according to the literature recognition criteria, but, does anyone really argue that Nimzovitsch wasn't a theoretician?) I suggest the category be changed to be limited to unmistakable entries, like Nimzovitsch, rather than to abolish it in toto. Ihardlythinkso (talk) 06:06, 19 August 2012 (UTC)


 * I'm OK with that. While we're taking a much needed hatchet to some of the Chess people categories, someone should take Sam Sloan out of Category:Chess historians.  That one really stands out too.  I should do it myself, but some days I don't have the will to fight with people who might argue that Sloan's crappy book claiming that chess originated in China demonstrates that he is a chess historian.  Quale (talk) 06:40, 19 August 2012 (UTC)


 * What exactly would your definition for a chess "theoretician" be, Ihardlythinkso? This is the same objection that Peter Ballard brought up 4 years earlier, and which no one has answered yet.  Steinitz, Tarrasch, and Nimzo were all world-class players who were also very dogmatic in the way they wrote their chess articles and books.  (Although not nearly as dogmatic in the way they actually played, especially Tarrasch)  However, note how old all of these names are.  That's because chess has not had a truly dogmatic world-class player for the last 70 years!  What it does have are strong players, chess writers, chess coaches, and chess composers, ALL of whom fundamentally alter our theory and perception of the game.  Furthermore, all of those are better and more precise terms than the vague, undefined "theoretician".  I propose deleting the list and replacing it with better, more specific lists.  ChessPlayerLev (talk) 07:00, 19 August 2012 (UTC)


 * ChessPlayerLev, it's not *my* def, but what the chess literature unmistakably recognizes as theoretician, that s/b the new criteria. (We all know Nimzovitsch was theoretician. If we scrap the category, then readers wouldn't be able to profit from the knowledge, that there are unmistakable theoreticians in chess, and some figures are clearly noted as such -- other than via individual mentions in individual bio articles.) There is no "exact" def to give. But judgement & consensus can weigh in, like everything else in the Wiki (e.g. notability). I'm suggesting a stricter criteria, where overwhelming consensus will exist about the figure according to the literature (e.g. Steinitz, Nimovitsch). It would result in a much-reduced list. (I think this is consistent with Ballard's remarks, because he is referring to the absence of criteria which has allowed the list to grow arbitrarily, producing the problem of meaninglessness. With a much-restricted replaced criteria, based on consensus, it solves the problem he's describing.) Nuking the category, rather than restricting it, I think takes value away from readers unnecessarily. (We chessplayers *know* there are notable theoreticians from the literature; let's share that knowledge with readers via cat, rather than nuking cat because it can't be precisely defined.) Ihardlythinkso (talk) 07:50, 19 August 2012 (UTC)


 * The way I think should proceed, is that the cat is retained, but its current list is emptied. Then we could start to add names, via consensus. (There would be discussion if disagreement, certainly all contributors would be challenged to know their literature and would learn something, too. It would be challenging, fun, and in the end of value to readers. A win-win scenario.) Ihardlythinkso (talk)


 * Def: "Widely regarded as a [chess-topic] theoretician" (in the body of chess literatures). Where there was no consensus that the figure was widely regarded as theoretician in the chess literature, the figure wouldn't get in the list. (If it did, the list would start to deprecate to meaningless again.) Ok, Ihardlythinkso (talk) 08:08, 19 August 2012 (UTC)


 * Okay, please tell me what "the chess literature unmistakably recognizes as a theoretician". Tell me what specific "chess literature" you are talking about, and tell me what their definition is.  A word without a definition is completely useless, a fact I hope you can recognize.  Amusingly, this is the same problem Peter Ballard ran into 4 years ago.  He would ask people what the hell a theoretician was and to provide reliable sources that had a definition of the term.  And they were unable to.  Since no one has answered this basic question, I believe the list should be scrapped.  We know what a "chess writer" is.  We know what a "top player" is.  We know what a "chess coach" is.  We know what a "chess problemist" is.  But after 4 years, we still have no idea what a "chess theoretician".  Or at least no definition has been provided for us.


 * By the way, don't confuse my thoughts on the matter. I consider Nimzo and Tarrasch as being no more or less the "theoreticians" than Fischer, Petrosian, or endgame composer Chekhover were.  All have contributed massively to our understanding and theory of the game of chess, albeit through different ways.  ChessPlayerLev (talk) 08:32, 19 August 2012 (UTC)


 * You're not getting it, revealed by your statement, "I consider Nimzo and Tarrasch as being no more or less the "theoreticians" than Fischer, Petrosian, or endgame composer Chekhover were", shows why you are not understanding. (The point isn't who *you* or any other editor considers to be a theoretician, but what consensus is on who the body of chess literature widely considers to be theoreticians. If you don't feel that Nimzovitsch, for example, is widely recognized as theoretician in the body of notable chess literatures, but other ProjChess members and other interested editors disagree, well then, Nimzovitch will be added to the list and consensus did not go your way. It'd be the same as other WP consensus-building.) But now I'm repeating myself. Ihardlythinkso (talk) 10:25, 19 August 2012 (UTC)


 * I already explained "chess literature" isn't subject to a precise definition; I would say it consists of all notable books printed by notable chess writers. So you want me to specify an exact title. You're not hearing me. Ihardlythinkso (talk) 10:30, 19 August 2012 (UTC)


 * "Amusingly, this is the same problem Peter Ballard ran into 4 years ago." ChessPlayerLev, I really don't see anything humorous going on here. If this is amusing to you, it's not helpful to tell that. "He would ask people what the hell a theoretician was and to provide reliable sources that had a definition of the term. And they were unable to." I read the Talk item Quale linked to, and I don't see that kind of exhaustive query that you imply. (And I don't see any bolsterous "hell" language, either.) Can you stop with the hyperbole here; it doesn't contribute to anything worthwhile. Ihardlythinkso (talk) 10:34, 19 August 2012 (UTC)


 * ChessPlayerLev, I think another mistake you're making, is raising the level of categorization to that of article content. I don't think the same standards of reliable refs, apply. (If a figure ends up in category "Chess theoretician", it is because the consensus of ProjChess members and other interested editors put it there. And that is a pretty good measure. It is not vastly serious if a "mistake" is made. There are no WP:COI, WP:OR, WP:SYSNTHESIS, WP:RS, WP:VERIFY issues involved. I'm guessing that you are raising the bar of verification re categorization to that of challengable matierial in a mainspace article. And that would be an inappropriate comparison. Additionally, if a figure were categorized "theoretician", but the mainspace article was not mentioning it or not able to support it with reliable refs, then it probably wouldn't have gotten categorized that way in the first place. So what would be the concern here?) Ihardlythinkso (talk) 10:45, 19 August 2012 (UTC)


 * I say we start right now. (Step 1: strip the current category list empty. Step 2: User:ihardlythinkso proposes to add Aaron Nimzovitch as first entry in the category list. Step 3: Do others agree? User:ChessPlayerLev, apparently does not feel that Nimzovitch is widely recognized in the body of chess literature as an [openings] theorist. How do others feel?) Simple. If consensus does not go my way to include Nimzovitsch, I can live with that. If there isn't overwhelming consensus to include him, he doesn't get into the list. If he gets into the list, there is some value to be had for encyclopedia readers from the consensus-built list. A win for everyone, and a lot of fun & learning to boot. Ihardlythinkso (talk) 11:10, 19 August 2012 (UTC)


 * Actually I think Lev is right. Nimzovitsch, could go in, and so could Philidor, Steinitz, and Tarrasch.  Add Troitzky as well.  Those are easy, but now, the rubber hits the road: Who else do you put in and why?  Does Alekhine get in?  Lasker?  Nunn?  Von der Lassa, Bilguier, and Schlechter for the Handbuch?  I suppose Suetin and Pannov are in, but what about Boleslavsky and Bronstein (theory of the KID)?  Quale (talk) 15:11, 19 August 2012 (UTC)
 * What's wrong with just keeping it to the obvious ones? (And for the record, ChessPlayerLev has not agreed on Nimzovitsch or any other figure as theoretician, so that is disagreement, not agreement.) Why does the list have to be a certain minimum size? (For my 2 cents, what I know about Lasker and Alekhine, how they're thought of in the literature, I would not !vote to include them. What I know about Suetin, I would !vote to include, plus Nimzovitsch, Steinitz, Tarrasch, and Troitzky. I don't know enough about how the literature treats the others, to !vote. I'm sure others do, though.) Is there something wrong with this process? It seems easy to start it "rolling", besides fun and educational. And the product would offer some value to readers. Ihardlythinkso (talk) 16:28, 19 August 2012 (UTC)
 * (Regarding your question: I think a general suggestion is that a category should have at least seven members, but that is not a hard and fast rule.) Well, Lev's concern over classifying Nimzowitsch as a theoretician is stronger than mine.  I think that's a very appropriate characterization, and I also think we'd be able to come to quick consensus there even if not universal agreement.  And I understand that many categories have pages that are borderline on the inclusion.  The issue I have with this category is that aside from maybe 7 or so fairly obvious cases, every single other player is borderline.  Anyone who might even slightly plausibly be considered a theoretician will require extensive discussion as to whether they belong in the category.  Or else we can just do what we do now which is put them all in, diluting the category to uselessness.  I actually appreciate discussion as I think wikipedia gets better when people with different views hash it out.  I just don't think this category gives enough value to have 150 debates, and that's what would be required—There are 165 members currently and I think we'd be hard pressed to find 10 that everyone would recognize as obviously belonging.
 * Instead let me go back to something that I think Lev suggested: let's listify and delete this category. That is something that is commonly done for categories that shouldn't be categories.  The beauty of a list here is that each entry should be annotated with a description of the person's theoretical contribution to chess and cited with sources.  This gives a lot more info than a category.  If desired, the list could be broken into sections, maybe Opening theorists, Middlegame theorists, and Endgame theorists.  A list would also centralize the discussion to the list page, rather than scattering it across 165 chess bios.  If the list was successful and we thought we really wanted a category we could always reconstitute it later with the requirement that everyone in the category had to be sourced and on the list first.  Quale (talk) 17:01, 19 August 2012 (UTC)


 * If we only need 10, I think that would be achievable (everyone doesn't have to agree, just a consensus; how do we know without trying?). The category thing is really equivalent to the list, and list subdivisions are like subcategories, yes? (Don't see much difference. The category list isn't scattered when the category contents is brought up for view.) I really don't see much difference here if we have list and sublists "xxx theoretician", etc. (I see, however, that a list (single page) would be easier to change/update. Perhaps that is what you meant. Yes in that case.) I'm all for Lev's idea of a list, just don't understand how that changes the problem much, other than ease of updating. (Aren't the same issues re evaluations involved -- who qualifies into which list or sub-list?) Ihardlythinkso (talk) 17:17, 19 August 2012 (UTC)


 * We already have mini-consensus (two editors) on five in the (new) list. That was easy. (Imagine more time; more contributing participants.) Make it conservative. A short list. Vast consensus. (I see no harm; only advantage. If the category is eliminated, that suggests, as ChessPlayerLev likes to do, that identifying figures who are widely thought of as theoreticians is impossible, because the semantics rips apart the notion of identification. I don't think that's consistent w/ what the literature tells us. Whether and what theoretician subgroups to umbrella, could be choice of ProjChess members. Why not? The cutoff what is notable to include as player article is a similar decision by ProjChess, no?) Ihardlythinkso (talk) 16:45, 19 August 2012 (UTC)


 * On second thought, I can see the category being removed, out of expediency. (My suggestion to start the list over with restrictive criteria, will necessarily make new discussion points for members. Ala AfDs. And I think perhaps that implies too much work to "look forward to", for ProjChess. Too much discussion. [Although, come again, an editor could just open a Talk section with a name that he/she wants to include, and see how much consensus develops. And only include the name if an obvious consensus musters. Not too hard.] So I can understand *DELETE* on that basis [added work, added discussion]. Even I think besides intrinsic value it could be fun, educational, others might likely not.) Ihardlythinkso (talk) 17:00, 19 August 2012 (UTC)

This is the problem with arguing about a list which has NO DEFINITION. (In the midst of his multiple replies, Ihardlythinkso even admitted that no such definition exists, and that he can't provide one) Who is to say that Kasparov can't be included in such a list? Or Fischer? Or any strong player of the last 150 years whose games have shaped the THEORY of chess far more than those of any non-elite writer?

If you say "let's have a list of chess players that are zorban; we already have a few books that refer to Petrosian and Smyslov in this manner! That's two guys on our list already!" and I ask you "what the hell is a zorban?" and you reply "well, there is no stated definition, but since a few books use this word to describe a few specific players, let's use it anyways!". This absurd discussion is exactly we're having right now.

Again, there is nothing involved in the list "List of Chess theoreticians" that isn't far better explained by "List of Chess Authors" or "List of Chess Composers". Scrap this meaningless list with no stated definition. ChessPlayerLev (talk) 21:13, 19 August 2012 (UTC)


 * A composer or author is not the same as a theoretician. A theoretician comes up with some knew general understanding.  An author generally just writes about something and compositions are generally a dead end with no contribution to theory.  Bubba73 You talkin' to me? 21:19, 19 August 2012 (UTC)


 * Ah! We finally have our first definition!  "A theoretician comes up with some new general understanding".  You are wrong about composers, by the way; endgame compositions frequently contain valuable contribution to theory.  That being said, there are two problems with this definition;


 * 1) It's vague and unclear.  Are opening innovations a type of "new general understanding"?  I would argue that they are.  Many would disagree with me.  What about endgames?  What about tactics in the middlegame?
 * 2) By this definition, every elite player of the last 150 years was a theoretician.  Are you going to tell me that Fischer didn't usher in a new era of general chess understanding?  That Bronstein's innovate and creative play didn't?  (Oh, and he wrote one of the 5-10 greatest, most influential books ever, too; "Zurich International Tournament of 1953", that shaped the way generations of top players understood the game, AND made major opening contributions)  Adolf Andersson wrote extensively and his games were certainly a source of study for countless masters of his time.  Morphy didn't write as much, but his games represented a huge leap forward in chess understanding, and are studied to the present day.  His contribution to the understanding of chess was probably no less than that of Nimzovich.  By the way, Peter Ballard said he would also consider Vishy Anand a theoretician for his countless contributions to opening theory and his innovative style of play, all of which vastly pushed forward our chess understanding.  I would agree with him.


 * You see now the problems with this list? Under your own definition (not mine), it quickly turns into a list of mostly elite past players. ChessPlayerLev (talk) 21:52, 19 August 2012 (UTC)
 * [This isn't directed at any individual, but rather a hope that the tone of our discussion remains collegial. I am glad that people are being firm advocates for their views, but I hope that emotions don't run too high on this or any similar issue.  I'm as guilty as anyone of getting too wrapped up in a disagreement].  I agree with ChessPlayerLev about this. "Contributing to the general understanding of the game of chess" is a good definition of a theoretician, but it is broad enough that any player who made a contribution to any opening that might have been of use to another player could qualify.  I think our intent is to restrict the category to people notable for "contributing to the general understanding of the game", but to me this still seems like an endless argument over whether each individual in Category:Chess players qualifies.
 * As an alternate idea, we could keep the status quo. I can do what I've done for the past four years which is pretend that Category:Chess theoreticians doesn't exist.  I never add any pages to it, but I never remove any either.  The existence of the category doesn't harm me and it certainly doesn't offend me, but I really don't see what purpose it is intended to serve.  Even if we reform it, I don't think it is workable.  I still think an annotated list with references could be very nice. Quale (talk) 23:01, 19 August 2012 (UTC)


 * I propose that we simply get rid of "Chess Theoreticians" and instead do a much better job of updating and maintaining "List of Major Chess Authors" and "List of Chess Composers". When a new list adds nothing important or substantive to an encyclopedia, there's no reason to add it.  Also, if the same energy that went into this discussion went into improving those two lists, there would be less general confusion.  For instance, I WOULD include Nimzovich in a list of "Major Chess Authors" but NOT Vishy Anand.  There is a much clearer definition there.


 * On a completely unrelated note, how do I get rid of Sam Sloan from the list of chess historians? I completely agree with Quale and his reasoning on this matter.  (By the way, Sam Sloan is an editor here who has threatened legal action against Wikipedia if articles related to him were altered in any manner...I wouldn't be surprised if he added himself to that list, as he has done to several other articles in the past).  ChessPlayerLev (talk) 00:07, 20 August 2012 (UTC)


 * Legal threats can get an editor banned (permanently). If you are talking about the category, just edit the article and delete the category (it will be near the end, under the last section).  Bubba73 You talkin' to me? 00:27, 20 August 2012 (UTC)


 * ChessPlayerLev, you misquoted and misinterpreted what I wrote. You said "Ihardlythinkso even admitted that no such definition exists". No I didn't. What I said was that I think there is no "exact" definition. (Big difference.) You said I "couldn't provide a definition". But I did provide one: "Def: 'Widely regarded as a [chess-topic] theoretician' (in the body of chess literatures)." The fundamental idea I have been proposing here, that I'm sure you still don't understand me, is that I think all the meaninglessness that has accumulated in the category content, is due to the fact that editors who add a name to the category are using their own judgement about what "theoretician" means. (And, every editor probably might have a different idea what it means.) My suggestion was a way to avoid the problem. By not using our own judgement who is or is not a theoretician, but instead just let the body of chess literature tell us. (When you say "Who is to say that Kasparov can't be included in such a list? Or Fischer?", my answer would be that Fischer would not be included in list of theoreticians, since the body of chess literature really does not identify him as such. If you personally think Fischer was a theoretician, that's your opinion, but that opinion isn't echoed in the body of chess literature.)


 * Quale wanted to dispense with the category in favor of documented lists, which was your idea ChessPlayerLev, and, I'm fine with that (however, I don't see how the problem of identifying who is or isn't theoretician goes away if the list or sublists include "theoretician", "openings theoretician", "endings theoretician", etc. that Quale made reference to). I think Quale is right -- specialist lists would be easier to maintain than categories. However I think his idea of lists differs from yours, as far as retaining theoretician identification, or sublists like "openings theoretician". (I think you ChessPlayerLev like to scrap any mention of word "theoretician", yes?) Perhaps it's good to clarify before this thread goes dead.


 * p.s. I'm curious about two of your other statements too, ChessPlayerLev: 1) "endgame compositions frequently contain valuable contribution to theory." So it should be easy to identify one of those, right? Can you? 2) "Are you going to tell me that Fischer didn't usher in a new era of general chess understanding?" You addressed this to Bubba, not me; but for me, yes, I would tell you that. Fischer was a great player and had a universal style. But I'm curious what "new general understanding of chess" Fischer ushered in to constitute a "new era"? (I have no idea what your answer might be. As much as I admire Bobby, I can't think of one. [He brought to light idea that Black can fight for a win from move one; perhaps that is an example and what you mean!?]) Ihardlythinkso (talk) 12:54, 20 August 2012 (UTC)


 * The idea of breaking the list into sections by phase of the game was just a thought, not an essential aspect demonstrating superiority over a category. (If it were that simple, then obviously you could just use three categories instead of one. But there are other reasons that lists are easier to maintain than cats.)  Later it occurred to me that a list of chess theoreticians might be more informative if it were organized chronologically rather than alphabetically.  The chrono order would have to be by birth date rather than date of chess contribution as that would be too hard to determine.  I think chrono might be preferable because it's easy to find a particular name in a list in chronological order if you have either a vague idea when the person lived or by simply using your browser's find functionality (usually bound to Control-F), but it's nearly impossible to get a chronological view out of an alpha list.  (Sortable wikitables are cool and can give you multiple orders, but they don't work as well when the tables get very large.)  Anyway it's vastly easier to maintain a list than a category because everything is in one place and you can watchlist the edits.  Unless watched very closely, over time categories can become Category:Chess theoreticians and spawn discussions like this one.  We had some problems with Category:Chess players so in 2005 I wrote the explanatory text on the page (expanded a bit in the years since then), the key bit being "people notable as chess players".  Actually people don't read categories before they dump articles in them so I'm sure there are a few pages in there that don't belong.  I've had to clean up the American chess players cat a few times to take out celebrities like Kubrick and Bogart and random unknown people who have a Wikipedia page and happen to like chess as a hobby.  (We have Category:Amateur chess players for them.)  I haven't take the time to comb through the other chess player subcategories.  Quale (talk) 21:24, 20 August 2012 (UTC)


 * So Ihardlythinkso's definition of a "theoretician" is "someone who is called a theoretician in certain chess books"? That's not a real definition, and again demonstrates how vague and useless the term really is.  Replace "theoretician" with "zorban" above and here is a good approximation of our discussion thus far;


 * Quale- "I don't see the point of a List of Chess Zorbans; the term is vague and unclear. No one has provided a definition that isn't better suited by a list like Chess Authors or Chess Composers.


 * Me- Completely agree.


 * Ihardlythinkso- We should include Chess Zorbans like Nimzovich, Tarrasch, and Steinitz in the list.  Clearly, they're Zorbans!


 * Me- But what's a Zorban?


 * Ihardlytinkso- Well, it's whoever is called a Zorban by the books!


 * Me- But what about the people who are NOT called a Zorban by the books? Without any specific definition for the term, it's impossible to tell!


 * And yes, I think the idea of changing it from a category to a list doesn't get rid of the problem at all. Look, "chess theoretician" is an old, outdated, vague term from the time when chess books simplistically characterized leading players as "tactical attackers" or "strategic, positional players", although leading players regularly displayed prowess and implementation of both qualities since at least the 1930s, if not earlier.  (And feel free to note that Steinitz, Tarrasch, and Nimzovich all disseminated their ideas well before that date, and did so for mostly beginners...)


 * As for your two chess questions Ihardlythinkso, this really isn't the place for me to give you a full chess lesson. (I usually charge for them!)  However, to avoid you accusing me of ignoring them, I will give each a brief reply;


 * 1) Even the celebrated Reti pawn study has major effects on chess play and understanding. There are countless practical examples where mixing the threats of getting inside the range of a pawn AND promoting one's own turns a loss into a draw.  If you want more practical examples beyond those already mentioned in the article, read GM Neil McDonald's "Practical Endgame Play".  He has a nice section on it.
 * 2) You expect me, a mere USCF-rated expert, to give you an impromptu lesson on Fischer's contributions to the theory and understanding of top-flight chess?  Seriously?  Here is Grandmaster Gregory Serper, a former top player and outstanding writer/teacher, spending 4 lengthy, detailed parts merely to explain Botvinnik's contributions to chess understanding, which occurred decades earlier, and which are necessarily simpler.  While I can't go into the nitty-gritty of his exact contributions, note than the overwhelming majority of top-20 players during the last 30 years, when asked who they modeled their games after, and who was their main inspiration with his play, reply "Fischer" the a great deal of the time.  Not Steinitz, not Tarrach, and not Nimzovich (by the way, many of them have not even read anything by Steinitz or Tarrasch or barely glanced at their games, and many haven't even read "My System"), but "Fischer".


 * Once again, this is a useless list. Keeping it would just be more unnecessary clutter that wouldn't illuminate anyone about chess.  ChessPlayerLev (talk) 22:32, 20 August 2012 (UTC)


 * As wonderful as Reti's famous endgame study is, that alone doesn't make him a theoretician. Surely people knew about that idea before then.  In fact, there are games that used it.  Bubba73 You talkin' to me? 22:45, 20 August 2012 (UTC)


 * Bubba73, I never said that it alone made him a theoretician. I was merely responding to your claim that most endgame compositions are meaningless in terms of chess theory and play.  Which is wrong.  You might be correct that people knew about this idea before Reti's study (although what's your source for this?  What is the game you had in in mind?), but the same is true of the principles written about by Steinitz, Tarrasch, and yes, even Nimzovich.  The top players were highly aware of them already.  ChessPlayerLev (talk) 23:08, 20 August 2012 (UTC)


 * CPL, not sure you answered my Q about your assertion of Fischer ushering in new era of chess understanding. (How so? I wasn't interested to hear about Botvinnik.) My position is simple: "theoretician" is in the chess literature; for it to be absent from WP would be odd. Letting the body of sources determine who is theoretician is consistent with the Wiki; doing it by our own individual definitions and opinions not grounded strictly in the literature, is WP:OR. Ihardlythinkso (talk) 22:53, 20 August 2012 (UTC)


 * My point is that we have simple, clear definitions for what constitutes a "Chess Writer", "Chess Teacher", and "Chess Composer", and meanwhile, no definition for "Chess Theoretician" beyond the fact that a minority of the chess literature (make no mistake; most books don't use this term to describe players) refers to a few very old players this way. We also have no way of determining whether a specific player is a "theoretician" or not.  That's not a good list.  ChessPlayerLev (talk) 23:08, 20 August 2012 (UTC)


 * CPL, you seem to be assuming it is somehow mandatory, that specific players are classified one way or the other. Why? I don't see any requirement or obligation to do that. It seems also, you're contending there are no modern theorists, only those of bygone era. (Are you? ... I'm not so sure about that.) Books can describe a player's contribution to chess theory, without ever using the word. As intelligent readers, we know when the authors are describing contributions to changes in chess understanding, evolutions of playing styles, etc.; all without using the word itself. On WP, we're supposed to take from sources and then put in our own words. (That would be a fair example of doing so, without misreading, misconstruing, or inventing.) Ihardlythinkso (talk) 23:39, 20 August 2012 (UTC)


 * Agreed, "theoretician" shouldn't be a cat. The Q seems to be whether it's represented in Quale's lists plan, or not. (Just to keep clear the issue.) But you agree w/ me that moving the problem from cat to list doesn't solve problem. But, I think it moves forward from Quale's housekeeping perspective -- the category can go byebye. Ihardlythinkso (talk) 23:55, 20 August 2012 (UTC)

Actually, this is a question you should answer. Are there any modern theorists, Ihardlythinkso? So far, you have named Steinitz, Tarrasch, and Nimzovich. And since you have given no definition of "theorist", I can't tell whether someone like Botvinnik, Fischer, or Anand qualifies or doesn't. Also, what part of the work of the first 3 was different than that of the latter 3? All six were top players. All six radically changed opening theory. All six wrote books and had in-depth analysis of their own games going into their own unique views on chess. All six greatly changed the way other players approached and viewed the game. Funnily enough, I might have agreed with you 12 years ago. Back then, when I was first really studying chess, concepts like "tactical, attacking player" and "solid, positional player" made sense. As did "chess theoretician". Now I realize that they're just vague terms meant to make a complicated game like chess easier for a beginner to group in his mind. Botvinnik was a pretty amazing and brilliant "tactical, attacking player" and Tal was an outstanding "solid, positional player" who was one of the best endgame players of his day. Similarly, Fischer was a great theoretician and Nimzo was a hell of a great, creative player. ChessPlayerLev (talk) 00:08, 21 August 2012 (UTC)


 * I don't agree "theoretician" is a word or concept cooked up for the benefit of beginners. CPL you keep repeating, saying I didn't offer a definition. I did. It's whatever the literature says. (If I gave my *own* definition of theoretician, what value would there be in that? I've been contending that is the type of thing leading to the meaningless category content in the first place.) You still haven't explained how Fischer, great as he was, "ushered in a era of new chess understanding", and what is that "new understanding of chess". As far as your Q who are modern theorists, I really don't know. I would suppose there are some. (But my not knowing, doesn't prove anything.) I think Quale should start his list(s) and the category thing can be nuked. (I need a drink.) ;) Ihardlythinkso (talk) 00:43, 21 August 2012 (UTC)


 * Am I the only one who thinks that this entrenched debate is a little pointless? "Theoretician" is surely covered as well as it needs to be in the Chess theory article. Okay, just a thought. Brittle heaven (talk) 18:26, 21 August 2012 (UTC)


 * So anyone who ever contributed to opening, middlegame, or endgame theory is a theoretician? GREAT!  Not only does every single elite player of the last 150-200 years make it in, but so does every master that ever played a novelty, everyone who has ever made an endgame composition, people like Trompowsky, Michael Basman, etc.  The list should run well over a thousand.  As for the article, it's fine, because it talks about the "theory" of chess, instead of meaninglessly labeling individuals "theoreticians".  It's a word that makes sense when talking about the game, but not its players.  ChessPlayerLev (talk) 18:50, 21 August 2012 (UTC)


 * The article labels Nimzovitsch a theoretician. Also Reuben Fine. Fischer, who you said "ushered in a new era of general chess understanding", isn't mentioned. Ihardlythinkso (talk) 19:00, 21 August 2012 (UTC)


 * I also think Fischer did user in a new era of general chess understanding. First, in the way that he dedicated his entire life to chess raising the expected level of preparation for chess professionals in a way that all future top level players would emulate, but also more specifically by adopting variations such as the Poisoned Pawn.  Based on general principles it was thought by most at the time to be inferior, but Fischer relied completely on concrete analysis ignoring common principles as elucidated by Steinitz and Tarrasch.  Today I think it is even considered drawish.  Watson writes about this, and it is a classic theme of modern chess that concrete analysis (generally provided by home preparation) trumps general principles.
 * But anyway, Brittle heaven is right—I think everyone has had their say and we aren't covering any new ground. I propose that we leave things as is.  The category isn't a WP:BLP concern as would be something like Category:Arsonists so we can just leave it be.  Keep in mind that actually removing the category would require that it be nominated at WP:CFD, and even if we were unanimous in deciding it should be deleted, discussion there could overrule (or even just ignore) us.  If someone wants to work on a List of chess theorists I think that would be good, Ihardlythinkso thinks it would be OK, and ChessPlayerLev thinks it would be bad, but we could worry about that on Talk:List of chess theorists.  Quale (talk) 23:47, 21 August 2012 (UTC)


 * I agree with Bubba's simple definition, "A theoretician comes up with some knew general understanding." Philidor and Steinitz, about the value and use of a pawn. Tarrasch, development over material. Nimzovitsch, control the center from the flanks; over-protection of squares, etc. But the things you name about Fischer (devotion; home-prepared opening lines), don't translate to sub-objectives of checkmate or ways to play on the board, but rather how he prepared for tournament and match play.


 * But I think CPL's view is just playing with semantics, which can be seen by analogy to topic of philosophy... CPL might say: "We all know what a philosophy teacher is. (Professor at a college who teaches it.) We all know what a philosophy writer is. (A person who writes books about it.) So what is a philosopher? What's the definition? Anyone who's had a life, and a certain minimum I.Q., has probably thought about the meaning of life, ethics, metaphysics, and has made a contribution to philosophy. Anyone who wrote a novel or book (fiction or non-), has probably made a contribution to philosophic thought. So are all novelists philsophers? Is anyone who ever wrote a book a philosopher too?" And if I were to tell CPL that my definition of word "philosopher" is someone who is identified as philosopher in the body of philosophic literature, like Socrates (who wrote no philosophic text), then CPL would likely say that my def is no def and avoiding giving a def. And he would likely challenge the idea that Socrates should be considered a philosopher any more than CPL's own grandpa, who often mused about the meaning of life and inquried into the human endeavor. "Why isn't my grandpa, who offered his philosophic thougths and reasoning and wisdom, any less a philosopher than Socrates? Why isn't everyone's grandpa a philosopher, who had a life and a minimum I.Q. and offered insights about existence? See? Everyone's a philosopher! The word is meaningless. Get rid of it. (Don't have it as WP category or list.)" (I don't mean to artificially quote CPL or mock him, I'm trying to prove my point re semantics, and how that line of reasoning isn't valid. I say back up to a definition like Bubba's.) Ok, Ihardlythinkso (talk) 11:15, 22 August 2012 (UTC)

The difference is that I'm not using my grandfather as an example here. Rather, the example is a world champion and one of the greatest players ever, Fischer. He was the main inspiration for 2 entire generations of top players and fundamentally altered the way the game was played at the highest levels. I would say that's a "new general understanding". Also, philosophy is its own self-contained subject. "Chess theoretician" really isn't. It's a vague subcategory in chess that was largely instituted in chess writings 100+ years ago as an easier way for leading players of that time to try to explain things to rank beginners. ChessPlayerLev (talk) 00:39, 25 August 2012 (UTC)


 * So then leave out the example of your grandpa in the given analogy, my argument wasn't dependent on it: "Any novelist or writer or anyone who has had a life and a minimum IQ has contributed to philosophic thought and is a 'philosopher', too." The topic of individual player Fischer was a side-issue (whether he was theoretician or not), and "altered the way the game was played at the highest levels" is still unclear what that means vis-a-vis contributing to chess theory the way Steinitz or Tarrasch or Nimzowitsch did, so I still have no idea what you mean. (Fischer did highlight playing for a win as Black from move one, but he did that out of a ferocious will to win games, I don't recall he proposed any theoretic basis for the practice.) Philosophy is a "self-contained subject" -- I have no idea what that means, or how or why it breaks the analogy. And Nimzowitsch's My System wasn't written for the kiddies, either. Ihardlythinkso (talk) 05:33, 25 August 2012 (UTC)


 * Regardless of what you have been told, My System was in fact very clearly written for much weaker players. Most players, myself included, studied it when we were 10-12 years old and of roughly 1300-1600 Elo strength.  The top players of Nimzovich's era were most assuredly not studying from it.  Also, the fact that Fischer's contribution to theory is way more advanced, complicated, and can't be as summarized as easily doesn't make his contributions any less valuable.  Again, your assumption that one can summarize his theoretical contributions in a quick Wikipedia message is laughable.  You can't even do that for the basic ideas of Nimzovich!  I gave you an example of how it takes a serious, involved, 4-part article by GM Gregory Serper to even scratch the surface of Botvinnik's theoretical contributions (which were still simpler than Fischer's).  However, as usual, you ignored it to keep asking me a rhetorical question.  If you like having a completely meaningless list around on Wikipedia, I guess you have your wish, since Bubba73 agrees with you, Quale agrees with me, but we don't care enough to force it through.  ChessPlayerLev (talk) 06:29, 25 August 2012 (UTC)


 * Actually I don't fully agree with you. I think "theoretician" is a label that has value to describe someone like Lipnitsky and Pannov among many others, I just don't think it works well as a category.  I also don't agree that it is impossible to briefly summarize a person's theoretical contributions to chess or explain why they are important.  That's the equivalent of claiming that it's impossible to briefly summarize something complex like quantum mechanics so it shouldn't be attempted.  Quantum mechanics is far more complicated than the most involved chess theory invented but it is certainly possible to summarize quantum mechanics and give an idea of its significance to physics.  I don't intend to participate in this particular discussion any longer because I think it's already run its course and we won't achieve anything today.  Maybe we can take it up again in a few months if that seems wise.  I know that you and Ihardlythinkso don't get along, but I think that's a shame.  You both are very dedicated and serious about maintaining rigorous standards for chess articles, but apparently there is a personality clash.  Maybe if you two hadn't gotten off on the wrong foot at your first encounter you would be able to work together collaboratively, but that doesn't seem possible now. Quale (talk) 07:08, 25 August 2012 (UTC)


 * Quantum theory is certainly more complicated than any chess theory, but at the same time, chess is a highly contextual game, while science is, by its very nature, generalized. Thus, you can make broad statements in science that are true "F=ma" while this simply doesn't work in chess; "bishops are bad in closed positions" will have a million counterexamples.  Thus, explaining Fischer's contributions requires me to present a board, a bunch of his games compared to those of his contemporaries, and a nitty-gritty discussion of the handling of certain pawn structures arising from popular openings.  This is a serious research job for a player of solid grandmaster strength that should take dozens of hours.  By the way Quale, did as good a job as anyone can of giving a "layperson's" explanation of Fischer's contribution above, even if it doesn't really say much about what he actually did.  I agree with some of the other things you write as well.  As for Ihardlythinkso, I have tried to set any personality differences we might have had on numerous occasions, the most recent one occurring  several hours ago.  I really only care about improving Chess Articles, not personal squabbles.  ChessPlayerLev (talk) 07:27, 25 August 2012 (UTC)


 * No, I don't agree. You seem to be saying that if it isn't possible to completely explain everything about a subject in the space available (in this case maybe a short paragraph), then we shouldn't or aren't allowed to say anything at all. I don't understand this.
 * I do agree that chess isn't particularly amenable to general rules. In fact, this is a characteristic of modern chess and one of Fischer's influences which might be considered a contribution to chess theory: concrete analysis trumps general principles.  However, you can still say "Fischer showed that bishops are more valuable than knights in particular positions" and as long as you provide a reliable source for the claim this is can be a concise and accurate summary.  Naturally you might want to say more, so if you want to explore Fischer's contributions in depth, you can create Bobby Fischer's contributions to chess theory where you will have a lot more room to present specific examples and go into more detail.  That article could be linked from several places such as Bobby Fischer.  You may in fact want to say a lot more, so if you run out of space in that article, you can spinout specific topics to subpages such as Bobby Fischer's theory of the two bishops or something similar.  But it is still always possible to briefly and accurately summarize these topics, and if you care to explore in depth Wikipedia has the space as long as you have the sources.  (If you want to create Bobby Fischer's contributions to chess theory I encourage it.  It would be a lot of work to do it well, but there have been several books on this topic published in English [and possibly more in other languages], so there should be reliable sources.) Quale (talk) 21:27, 26 August 2012 (UTC)


 * You're greatly underestimating how difficult something like this truly is. For instance, your statement that "Fischer's influences which might be considered a contribution to chess theory: concrete analysis trumps general principles." is simply wrong.  This is a standard element of chess in the post-WW2 era; it was so ingrained by all the top players by 1950 (a good decade before Fischer) that I don't know whether it should be ascribed more to Botvinnik, Keres, Boleslavsky, or a half-dozen other players.  You say there are books that go into Fischer's contributions.  Okay.  See if you can summarize them in a Wikipedia comment in a manner that is both informative and manages to avoid inaccurate statements.  (Like the example about concrete analysis above)  I gave you an example above of the type of extraordinary work GM Serper did to accurately cover Botvinnik's theoretical contributions.  ChessPlayerLev (talk) 21:37, 26 August 2012 (UTC)


 * I stand corrected. (Forgot re Fischer's contribution to theory, bishops' superiority over knights endings.) Ihardlythinkso (talk) 23:18, 26 August 2012 (UTC)


 * ChessPlayerLev, if you're saying that it's difficult to do, then I agree. If you are also saying it's too difficult for you to do, then I can't disagree with you although I think you could make a good start if you didn't make the task too large.  As to whether the notion that concrete analysis trumps general principles was so deeply ingrained in top players by 1950, well I'm certainly not an expert but on the face of it I'll have to say I disagree.  Certainly Boleslavsky and Bronstein contributed to the modern view, especially with their work on the KID but probably in other ways as well, but looking at the openings played at Zürich 1953 shows nothing like what is employed in modern praxis.  Some of this is just fashion, of course, but modern opening play tends to be much sharper and the initiative is valued much more highly today.  Kasparov was probably even more influential than Fischer in this, especially with regard to the value of the initiative.  Perhaps it could be argued that the value of wins is enhanced today by the trend to shorter Swiss tournaments over the lengthy double round-robins that were more common two generations ago, but I don't think the full realization of how playable truly risky lines could be came until computer-aided analysis became common in the 1990s.  Quale (talk) 02:45, 27 August 2012 (UTC)

I definitely think it's too difficult for me to do, heh. My last published rating was about 2050 (at the age of 19), and even if I have improved a little in the past six years, it's certainly not past the 2100-2200. That might be enough to write semi-logical assessments of pre-WW2 play, but once you get to Zurich 1953, let alone Fischer, it's certainly beyond my abilities. I trust leading players when they talk about the vast influence of Fischer and how it changed their styles, but that's about it! ChessPlayerLev (talk) 04:25, 27 August 2012 (UTC)

Great discussion ! I really like it to see wise guys coming with cunning arguments and discussing thoroughly some stuff. Please allow me to note hereunder a few points, to sum up my understanding. Please note I am not trying to sum up the whole discussion, but just a few points I found interesting, for each side of the argument.

Points "for"
 * 1) A definition can be found: Bubba proposed one, Ihardlythinkso proposed another. I personnally prefer the one from Ihardly, because it is really practical and "wikicompatible", as you just have to give a new reliable sources to back the statement that X is a chess theoretician. I understand also the counter-example with Zorban, but even here I see nothing wrong with it and if several reliable sources invented the word Zorban I think Wikipedia should have an article on Zorban.
 * 2) chess theoretician is a commonly used term in a lot of chess books. So Wikipedia should have something on that.

Points "against"
 * 1) Whatever the definition used, any list on "chess theoretician" will most likely end up having most strong chess players, including most world champions (probably Tal and a few others would be discussed). So what exactly is the value of having a list that is mostly a copy of another list ? (like, list of strong chess players).
 * 2) I am not sure I have a complete understanding of what we mean by "chess theory" there. Is it only openings ? or systems (Nimzovitch, Philidor, Tarrasch, Steinitz, ...) ? or also endings ?
 * 3) It seems we are getting to the situation in which we are inventing a definition ourselves. As Wikipedia editors, this is not something we should do because it is dangerously close to WP:OR. Instead, we should be able to say something like "the definition of a chess theoretician is XXXX because this is the definition used in the reliable and notable books YYY, ZZZ and WWW". If we are not able to say something like that because there is no commonly-agreed definition, then maybe we should forget about a "list of chess theoreticians".

Happy to hear from all of you ! SyG (talk) 21:01, 8 September 2012 (UTC)