Talk:Comparing top chess players throughout history

From Wikipedia, the free encyclopedia

This is the talk page for discussing improvements to the Comparing top chess players throughout history article.
Article policies
This article is within the scope of WikiProject Chess, a collaborative effort to improve Wikipedia's coverage of chess. For more information, visit the project page, where you can join the project and/or contribute to the discussion.
Chess Portal
B This article has been rated as b-Class on the quality scale.
High This article has been rated as High-Importance on the importance scale.
This article is in the list of Selected articles that are shown on the Portal:Chess.
Articles for deletion This article was nominated for deletion on 2007-01-02. The result of the discussion was keep and cleanup, source, and remove OR.

Contents

[edit] Article content

Much of this article (at least the stuff not in the "statistical methods" section, which is new) has been moved from World Chess Championship (see that article's talk page for my thoughts on why this material is better in its own article). I have, however, deleted the below section:

Hence, it is extremely unclear which single player truly is "the best", but nearly all lists of the best players include at least the following ten (in chronological order):

The list is completely subjective, and since we now have other lists which, while not really any more objective are at least attributed to particular sources, this one is probably unnecessary.

A note on Chessmetrics: it seems Sonas has recently overhauled the website, and a lot of the ratings seem to have changed (formerly, Capablanca had the highest 5-year peak average). The quote I've put in the article from him about it being impossible to compare ratings of players from different eras is from the old version of the website, but the ratings I've quoted are from the new version. I know this isn't ideal--it's a quick fix while I try to find a similar quote on the new site (if it turns out there isn't one... well, I'm not sure what the best thing to do will be then). --Camembert 18:39, 8 Apr 2005 (UTC)

[edit] Richardcavell's edit

I'm generally reluctant to do wholesale reverts of edits made in good faith, which is why I'm explaining myself here. I think Richardcavell's edit of 20-Nov-2006 actually confuses the article. First, I don't see "what a player gains from analysing the games of a particular player" has to do with greatest player of all time. An instructive player need not be a great player, and vice versa. Second, I don't think the Fischer/Kasparov example is accurate, and will inevitably lead to people with different opinions editing it. So I'm undoing the entire edit. Perhaps better would be to produce an example ("Author X once evaluated player greatness judging by how instructive their games were"), but to my knowledge this has never been done. Rocksong 02:39, 20 November 2006 (UTC)

Fair enough. I take your reversion in good faith. I'd still like to see something there about subjective factors playing a role in determining who was the best chess player. - Richardcavell 03:43, 20 November 2006 (UTC)
Thanks for taking my revert in right spirit! In general, I think would be good to list any "quality" subjective lists out there. I know of at least two which deserve mention: The prolific author Irving Chernev wrote a book called "The Golden Dozen" in which he ranks his top 12 (up to Fischer, i.e. Karpov not considered). It was a fairly conventional list: 1 Capa, 2 Alekhine, 3 Lasker, 4 Fischer etc. And Fischer himself did a top 10 in the 60s, claiming to have done it on analysis of play only. His list was a bit more controversial (surprise!) - he left out Lasker and I think Morphy was top. Any other, more modern, subjective lists could be added also. I don't mean by average Joe on the internet, I mean published by reputable players or authors. Rocksong 04:43, 20 November 2006 (UTC)
Such a list can be assembled by collating individual comments over the course of chess literature. The deleted list above seems to be "List of World Champions with high win percentages over peers". For example, Euwe beat Alekhine when the world understood that human factors had weakened his chess skills. The Dutch GM won the title fairly, but mostly because the leading light was playing second tier chess, and it wasn't an age of 3-4 potential champions. If we are actively seeking subjective material, there's an effect from Morphy to Lasker that "the concepts of the game were still developing". They were playing on native ability, and arguably with an incomplete set of chess maturity. In the 1920's, the "Hypermodern" movement put the ideas of indirect control and weakening overextended centers into world view. Right afterward, a consecutive pair of "greatest ever" champions emerged - Capablanca & Alekhine. As a combined team, no one could seriously threaten them for about 25 years from 1920-1945. Individuals like Tal & Fischer achieved even higher lights in isolated years, but then could not sustain that level. The next "dominant" champions would be Karpov & Kasparov, who formed another combined team no one could touch for another 25 years from 1975-2000. These four players will always show up on "Greatest Ever" lists over a 5 year period or longer. The exact differentiation becomes difficult because of the fundamental vagueness of the concept of "great". TaoPhoenix 01:22, 7 September 2007 (UTC)

[edit] Title

As came out in the AfD the title is a major part of the problem. If we get the title right then some rewriting to meet the new title will get rid of much of the OR. My suggestion is: Historical ranking of chess players. Revised suggestion in the light of discussion: Comparing the chess greats. Please indicate below. BlueValour 19:17, 14 January 2007 (UTC)

[edit] Support

  1. BlueValour 23:35, 11 January 2007 (UTC)

[edit] Oppose (keep existing title)

  1. Slightly oppose - what people really discuss/debate is who is the greatest (or try to rank the greatest). Historical ranking of chess players sounds a bit like ranking all players, i.e. am I better than a club player from the 1950s, or are today's GMs stronger than those of the 1950s. That is not the question this page tries to answer. It tries to answer (or, to be more precise, document other peoples' answers) to how to rank the very greatest. So I kind of like the current title, though my feelings on it aren't strong. BTW, I don't see how changing the title will affect the OR. The OR in the first section could just as easily be applied to Historical ranking of chess players. The OR parts can be rewritten whatever the title is. Rocksong 23:59, 11 January 2007 (UTC)
Sorry, but I like Comparing the chess greats even less. Because the title contains no direct indication that we are attempting compare different eras. Rocksong 23:36, 14 January 2007 (UTC)
How about "comparing top chess players throughout history"? Bubba73 (talk), 23:58, 14 January 2007 (UTC)
OK, we need to move on, I agree. BlueValour 00:32, 15 January 2007 (UTC)

[edit] Change to different title (specify)

  1. I am in favor of changing the title and I think the proposed title is better than the current title. However, I think there should be a better title but I don't know what to suggest at the moment. I think the proposed new title sounds too narrow and doesn't accurately reflect the intent of the article. Bubba73 (talk), 01:06, 12 January 2007 (UTC) Perhaps "Criteria for top chess players" or "Measurements of top chess players", just a couple of ideas. Bubba73 (talk), 01:08, 12 January 2007 (UTC)
I like the revised "Comparing chess greats", or something similar such as "Comparing great chessplayers". Bubba73 (talk), 20:03, 14 January 2007 (UTC)

[edit] Absence of Fischer and Karpov from Elo's 1978 list

The chessbase.com reference for Elo's 1978 list - http://www.chessbase.com/newsdetail.asp?newsid=1160 - is a bit weird. It was certainly done around 1978, because it contains 1970s players like Portisch (born 1951, the same year as Karpov) and Mecking (born 1952). Chessbase notes, "Fischer and Karpov were at the top of the list, although they were 2780 and 2775 on the January 1st 1978 FIDE ratings list, since these numbers did not represent a five-year average for the players." But this doesn't explain their complete absence from the top 47, because Fischer (certainly) and Karpov (almost certainly) would have had higher 5-year averages than Portisch or Mecking in 1978. So either Elo mucked up, or Chessbase did. So that's why I've modified the note to say that Elo left Fischer and Karpov off his list, and left it at that. Rocksong 11:09, 3 March 2007 (UTC)

My guess is that he left Fischer off because he didn't have a five-year period of activity when he was at his peak. He had been inactive before playing again at his peak in 1970 to 1972. I also speculate that perhaps he left Karpov off because he had not yet reached his peak. Well, as you point out - others about the age of KArpov are listed, so perhaps omiting Karpov was an error. On the other hand, in the chart on page 88-89, it looks pretty clear that Portish hes leveled off and is near his peak. Bubba73 (talk), 04:39, 14 January 2008 (UTC)
Well, Fischer wouldn't be on the 1978 list, would he? He gave up chess around 1975 when he didn't defend his World Champion Title. Karpov's absence is somewhat interesting, however, since 1978 was the first year he defended his title against V. Korchnoi. —Preceding unsigned comment added by 137.122.45.230 (talk) 15:00, 7 May 2008 (UTC)

[edit] Advances over time

This article doesn't appear to even mention the fact that chess itself advances over time. For example, if you were to send Kramnik back in a time machine to play Staunton, or Morphy, or Steinitz in their primes, he would probably achieve technically winning positions in every game just by his modern knowledge of openings alone. Though Lasker is listed as "greater" than Kramnik (I agree with this), this doesn't mean he was STRONGER than Kramnik (which I think would be a completely unrealistic claim); in other words, what Lasker did with the state of chess in his time impresses me more than what Kramnik has done, but I know that if pitted directly against each other at their peaks, Lasker would be utterly and completely outgunned in a match. His early 20th century openings alone, taking nothing else into account, would lead to an embarassingly lopsided result. And, like it or not, openings are a part of chess, so that cannot be discounted in these discussions. And yet, this does not even seem to be mentioned in this article. Smyslov 16:43, 13 August 2007 (UTC)

The article used to say this, but it was in a rather unencyclopedic style and was unreferenced. So it got deleted. By all means add something along those lines, with references. Peter Ballard 23:52, 13 August 2007 (UTC)
What references? In any case, I don't think I'm well-suited to writing such things. Nevertheless, this is something that absolutely has to be in this article, if the article should exist at all. Without this crucial element, it is exceptionally misleading, and might honestly cause non-chess folks to believe that top players of the late 1800s were actually literally equal in strength to those of today ... which of course indicates that chess has not advanced in the last 100 years! That is, of course, absurd, but that's the impression I would get from this article as it stands. Smyslov 20:47, 14 August 2007 (UTC)
Lasker publicly announced that he was not interested in becoming a detailed student of the openings. Regardless of absolute knowledge, his attitude alone would knock off some points from a comparison with the modern era. Today he would be a "typical mid-level GM". Only a few years later, Alekhine did deeply study the openings, and posted comparable results. Anyone who wants to source this need only quote Laker's own "Manual of Chess". TaoPhoenix 01:36, 7 September 2007 (UTC)
To suggest that Lasker would only be a "typical, mid-level GM" indicates that you either greatly overestimate the strength of a mid-level GM or you greatly underestimate Lasker's strength, or both. Quale 20:40, 25 September 2007 (UTC)
I completely agree that Lasker was proportionally dominant in his time. My comment referred to the slippery advantage that all modern GM's share, which is the computerized spread of information. They all have the lineage of legacy to stand upon; "Lasker didn't have Lasker to study from". This is what the modern usage of the Novelty is about; even the best geniuses will struggle while contending live against something the opponent has looked at for weeks at home. TaoPhoenix (talk) 20:44, 24 December 2007 (UTC)

I strongly disagree. I think you are all greatly underestimating Lasker's playing strength. Give him a month or so to catch up and he would be a very dangerous opponent to any person alive. —Preceding unsigned comment added by 75.111.197.14 (talk) 04:19, 15 September 2007 (UTC)

Lasker "from what year?" Capablanca did not lose a single game from their match in 1921, and Schlecter held him dead even until the very end. His one legacy flaw was that even in his day, he *didn't want* to catch up with theory. TaoPhoenix (talk) 20:44, 24 December 2007 (UTC)

The greatest of the great chess players have a genius for chess that far transcends the evolution of chess and the body of opening theory. Morphy, Capablanca, Alekhine, Fischer- modern-day 2700-rated grandmasters wouldn't stand a chance against those guys. Do you really think Kramnik or Anand or Topalov could beat Capablanca? Tsk tsk. —Preceding unsigned comment added by 75.111.197.14 (talk) 07:07, 15 September 2007 (UTC)

Actually, yes, I do think that a modern 2700+ would beat any of those older players in a 24-game match. You mention earlier giving Lasker a month to catch up on opening theory. That's the whole point. You're talking about comparing the REAL Lasker, who did NOT have a month of modern openings, with modern players. There is no month to catch up, it's only Lasker and what was in his actual brain in that time, comparing with players of today and what's in their brains. Lasker would land lost positions out of the opening in most (almost all?) games. If top players of today aren't better, than those players of the past added nothing to chess. Smyslov 21:21, 24 September 2007 (UTC)
I thought you were talking about comparing the skill of the players in different eras, not the overall level of chess theory in different eras. The question is whether the players are better, not whether more is known about the openings. (In any event this argument is essentially impossible to resolve.) I think you're barking up the wrong article for the argument you want to make, which to me isn't very interesting anyway. I expect that the theory would improve over time. I don't see anything in this article that suggests that it hasn't. Quale 22:00, 24 September 2007 (UTC)
Er, the skill of the player is increased by this more advanced theory. How can that not be a part of their skillset? This is precisely the #1 (but hardly only) reason they are stronger. Smyslov 15:27, 25 September 2007 (UTC)
To be honest, I have no idea what you're arguing. According to you, any chess player today should be stronger than Lasker or Capablanca because of the greater amount of theory available to the modern player. Except that clearly isn't the case. Quale 20:42, 25 September 2007 (UTC)
According to me, NOTHING of the sort. I never said or thought any such thing. As a 1700 player, I would lose to Lasker or Capablanca, and never said or argued otherwise. Don't attribute that to me when it is utterly untrue, please. Top players are another matter from "any chess player today". There is a certain degree of strength on top that is gained by the advances in theory. This is improvement. Chess, as with most everything, improves over time. If today's top weren't better, that would mean no advances had happened at all. Clearly, even the most cursory examination indicates that advances have occured. Smyslov 15:14, 1 October 2007 (UTC)
I think the anon is basically right, and I disagree strongly that Lasker would be a "typical mid-level GM". In fact I don't think that's a consensus view. Lasker was an elite player, champion for 27 years and the very best in the world until Capablanca. Even the "typical mid-level GMs" would probably tell you that Lasker was better than they are. Quale 07:40, 15 September 2007 (UTC)
"Dominant in the best years of their day" is different from "transporting the man in a time machine". We had an example of exactly this concept when Bobby Fischer surfaced to play Spassky. Drawing from another Wikipedia article, "In the book Mortal Games, Garry Kasparov is quoted: "He is playing OK. Around 2600 or 2650. It wouldn't be close between us."" TaoPhoenix (talk) 20:44, 24 December 2007 (UTC)

[edit] Deep Blue

Why is there no mention of Deep Blue on this page? —Preceding unsigned comment added by 24.9.50.166 (talk) 18:44, 2 September 2007 (UTC)

  • It would be difficult to add much about IBM Deep Blue here for a bunch of reasons:
  1. Too few games played.
  2. Too few opponents. The only elite player Deep Blue played was Kasparov, and they only played a total of 12 games. In historical chess terms, this would only be a single short match.
  3. Deep Blue has no official Elo rating.
  4. Deep Blue no longer exists, so there's no hope of fixing 1–3.

There's a better opportunity to say something about the current programs (Fritz, Rybka, Hydra), but comparisons between humans and computer programs are fraught with difficulties. It is clear that the strongest chess programs have been better than all but very elite human players for several years, and in fact most observers believe that currently the strongest programs are better than the strongest humans. This will probably become more clear in the next few years, although I suspect that the very best human chess players will start to avoid playing strong programs (this is probably already happening), making a direct comparison harder. Quale 02:08, 3 September 2007 (UTC)

[edit] Elo Table

The Elo table's size is out of proportion to the rest of the article. I suggest limiting it to the top 10. None of the modern players outside the top 10 could even conceivably be called Greatest of all time. Peter Ballard 07:57, 10 October 2007 (UTC)

  • I like the list, although I have a clear bias because I'm the one who originally added it to Grandmaster some time ago based on a Chessbase News article. It didn't fit very well in that article (it was in the section discussing the ill-defined "super-GM"), so it was moved here. Trimming it to the top 10 would be sad, as Tal would drop off the list and his 2705 in 1980 is a lot more impressive than Morezovich's 2762 in 2007, in my opinion. The list provides more information than just the top Elo ratings, it also gives a timeline of top results. Quale 15:01, 10 October 2007 (UTC)
The list is now up to 50, and will probably grow exponentially in the future. It will soon get too long to manage. It is already too long compared to the rest of the article. I say we cut it to top 10, or maybe 20. Don't worry about Tal dropping off the list - his place in history is assured. Alternatively, fork the list out to a different article. Peter Ballard (talk) 04:57, 26 January 2008 (UTC)
No debate, perhaps my post was missed because of the discussion below? I repeat: if we list every player over 2700 then the list will grow exponentially in future. So (a) can we agree that we limit the list to the top X rated players ever? (b) What is a suitable value of X? I like X=10, but X=20 might be a reasonable compromise. Peter Ballard (talk) 07:05, 29 January 2008 (UTC)
I agree the table's size has to be limited, but "top X rated players ever" is not a good method - at present Capablanca, Botvinnik and Lasker are in positions 27-29, Tal is at 40, I can't see Steinitz, Alekhine, Euwe, Smyslov, Petrosian or Spassky (and I haven't mentioned any of the challengers or credible contenders from before 1984). I'd prefer to see the current top 10, plus all world champs, their challengers and a maximum of 1 other credible contender from each reign (e.g. Rubinstein, Keres); with the "Rank" column showing where they would appear in a full table, and a notes column saying e.g. "world champion xxxx-yyyy" of "challenged for world championship in xxxx". That would be a twofer, as it would also illustrate ratings inflation. Philcha (talk) 08:08, 29 January 2008 (UTC)
I think doing both would make sense, but in separate tables. Trim the peak Elo ratings to a top 10. Put the world champions in a separate table with their historical Elo ratings. I wouldn't include a credible contender from each reign, although if this were desired instead I would suggest just including all the world championship challengers. We could also list all of the FIDE world #1s. Since Kasparov had the top rating for such a long time (and Fischer and Karpov were the only #1s before him), Anand became only the sixth player ever to top the ratings list in April 2007. Actually the best way to show ratings inflation is not the top rating (this would actually suggest deflation since Kasparov's 1999 peak), but rather the average of the top 100 or the floor of the top 100. (The rating required to be in the top 100 is currently 2627. Apparently in Jan 2001, 2595 was enough.) Quale (talk) 09:09, 29 January 2008 (UTC)
What Quale suggests (2 tables) is actually in the article already: (1) Elo's original list, and (2) the highest Elos ever. I read somewhere that ratings inflation began around 1985, so if that is correct, those two tables should be pretty well sufficient. Peter Ballard (talk) 10:57, 29 January 2008 (UTC)
I think it is clear that there has been rating inflation. It is also a lot easier to get the GM title these days. I was looking at the GMs by year, and there were very few by 1970 compared to now. Bubba73 (talk), 15:10, 29 January 2008 (UTC)
Speaking of Fischer and Karpov being the only FIDE #1 rated players before Kasparov, Elo published retroactive ratings back to something like 1850. Bubba73 (talk), 15:35, 29 January 2008 (UTC)
Right. We should have more about that at grandmaster. I have seen a few scattered counts of the number of grandmasters, but unfortunately I don't have a complete count for every year. If we had counts for just the decades (1950, 1960, 1970, 1980, 1990, 2000) it would probably be interesting and useful information. I think that would show an explosion in the number of GMs in the last 20 years. Quale (talk) 15:33, 29 January 2008 (UTC)
I got Anne Sunnucks's Encyclopedia of Chess yesterday (used 1970 edition), and it has a list of all of the grandmasters by year through 1968. Some years only one GM title was awarded. There are more names there than I care to type in, but it would be easy to count them for each year. Bubba73 (talk), 18:30, 29 January 2008 (UTC)
I wonder if the 2nd edition has any more GM data. Unfortunately I only have the 1st edition, and I haven't seen any used 2nd editions listed for sale. Kenneth Harkness also lists GMs in his Chess Handbook, and I have a couple of editions of that. I'll look to see if it is helpful. Quale (talk) 21:53, 29 January 2008 (UTC)
It probably has a little more data. I found used copies of the second editon for sale, but the price was too high for the condition of the book. Bubba73 (talk), 22:13, 29 January 2008 (UTC)

I trimmed the table to 20 today, while updating the list based on April 2008 ratings. I was tempted to trim it to 10, but thought that was a bit too radical. Maybe we can trim it to 10 later. Peter Ballard (talk) 00:18, 2 April 2008 (UTC)

[edit] Warriors of the Mind - reviews?

I've found a few dismissive references en passant to Keene & Divinsky's "Warriors of the Mind" but no genuine reviews (only the usual guff at online bookshoops). Does any one know of any credible reviews? Personally I don't trust the book because it gives Steinitz a pretty low ranking, below several obscure Soviet GMs of the 1950s to 1970s. I suspect that Steinitz' rating is depressed by 2 factors that Divinsky's algorithm doesn't allow for: Steinitz' "career break" from the mid-1870s to mid-1880s; and the related fact that there were a lot more opportunities for competition after the break (Steinitz in his 50s; modern players decline after 40) than before it. Philcha (talk) 21:45, 1 January 2008 (UTC)

Warriors of the Mind is comprehensively demolished by Edward G. Winter in a scathing review that can be found in his book Chess Explorations. I don't think it deserves a place in this article. Pawnkingthree (talk) 16:28, 20 February 2008 (UTC)
I haven't seen either Warriors or the critique you mention. But my feeling is that Warriors should be discussed, but the critique of it should also be there. Bubba73 (talk), 16:42, 20 February 2008 (UTC)
Fair enough. I'll put in something a bit later. Pawnkingthree (talk) 16:56, 20 February 2008 (UTC)
By all means reference + link to Winter's criticism, but bear in mind Winter is no friend of Keene. Peter Ballard (talk) 01:00, 21 February 2008 (UTC)
True, but his arguments in this case are difficult to refute. Pawnkingthree (talk) 09:26, 21 February 2008 (UTC)

[edit] Title - another suggestion

The main problem with this title is that it is not a noun - all Wikipedia titles that I've ever seen, other than this one, are nouns. I propose this be renamed "Comparison of top chess players throughout history" - that's not perfect, but it's a lot better. --209.78.3.114 (talk) 20:28, 25 January 2008 (UTC)

I agree. Bubba73 (talk), 00:22, 26 January 2008 (UTC)
I had never thought about it before, but I also agree. Maybe "Comparison of top chess players in history" would be shorter than "throughout". Quale (talk) 02:46, 26 January 2008 (UTC)
Or perhaps "Comparison of historic chess players", to make it shorter. Bubba73 (talk), 02:47, 26 January 2008 (UTC)
Maybe we don't need "comparison". Just "Top chess players in history" ? I suspect that some would complain that "comparison" is needed, because there is no absolute way to judge the top player in history. I think comparison is implicit in the title, so perhaps it could be left out. Quale (talk) 03:10, 26 January 2008 (UTC)
Yea, maybe "comparison" should go. Bubba73 (talk), 05:33, 26 January 2008 (UTC)
I don't have a strong opinion, but note that the article has already been renamed once. It used to be called Greatest chess player of all time. See discussion above at #Title. Peter Ballard (talk) 05:00, 26 January 2008 (UTC)
Yes, but that title is sort of subjective and POV. Bubba73 (talk), 05:32, 26 January 2008 (UTC)
Precedent suggests that we might want to call it "Historical rankings of top chess players" or perhaps just "Rankings of top chess players." --66.214.221.166 (talk) 00:01, 28 January 2008 (UTC)

[edit] A different criteria

I've thought of a different criteria to use, using ChessMetrics. Much of the time, one player is far ahead of the field - 50 or more points. What about tabulating the players that were, at some point, at least 50 points above the second best player, and then sort these according to that difference. Any thoughts? Bubba73 (talk), 01:27, 27 January 2008 (UTC)

Would be OR if you or I did it. But Jeff Sonas did it in an earlier part of The Greatest Chess Player of All Time – Part IV (see links at foot of page). IIRC the top player by that criterion was Steinitz, the only time the poor old guy got a look in. Philcha (talk) 08:30, 27 January 2008 (UTC)
That's not quite what I had in mind, but that could be put into the article. Bubba73 (talk), 15:54, 27 January 2008 (UTC)
Whoops, what I had in mind is actually in Part 1. Bubba73 (talk), 15:56, 27 January 2008 (UTC)

[edit] Fischer's List

Should it be mentioned about Fischer's list that it was published in 1964, and so it couldn't possibly include players like Kasparov, Krammnik and Karpov, or Fischer himself? —Preceding unsigned comment added by 137.122.45.230 (talk) 15:05, 7 May 2008 (UTC)