Jump to content

Wikipedia talk:WikiProject Chess/Archive 31

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 25Archive 29Archive 30Archive 31Archive 32Archive 33Archive 35

Proposed changes to the Troitzky line article.

abcdefgh
8
h8 white knight
c7 black pawn
c6 white knight
d5 white king
g5 black king
8
77
66
55
44
33
22
11
abcdefgh
A. Black to play and draw.
abcdefgh
8
e8 white knight
g7 black pawn
g6 white knight
d5 black king
g5 white king
8
77
66
55
44
33
22
11
abcdefgh
B2. Black to play and draw.
abcdefgh
8
a8 white knight
h8 black king
c7 black pawn
c6 white knight
h6 white king
8
77
66
55
44
33
22
11
abcdefgh
C. White to play and win.
abcdefgh
8
a8 white knight
h8 black king
c7 black pawn
c6 white knight
f6 white king
8
77
66
55
44
33
22
11
abcdefgh
C'. Draw.


I'm intending to make changes to the Troitzky line article.

I think the changes are necessary on the grounds of accuracy and clarity.

For background see the discussions Draws on and behind the Troitsky line and Troitsky line.

The current statement of the Troitzky line rule is

For the position with White to move, Troitsky established that if a black pawn is securely blockaded (by one of the white knights) on a square no further forward than the line a4–b6–c5–d4–e4–f5–g6–h4, then White can win the resulting endgame (and similarly in reverse for Black), no matter where the other pieces are placed.

(I added the word "securely" myself, because I assumed from the discussion in Draws on and behind the Troitsky line, that Bubba73 had intended the addition, but omitted it by oversight.)

The drawn position C' shown is a clear exception to the statement following the words, "For the position with White to move, Troitsky established that", rendering the complete statement incorrect.

Note that because of the phrase, "with White to move", positions A and B2 must be taken after Black's first move to count as candidate exceptions. I think the phrase is redundant and would like to find a source that doesn't include it.

The statement of the Troitzky line rule is also unclear, because the phrase "securely blockaded" is not defined.

I have been soliciting the views of chess players on whether the current wording is adequate to cover various drawn positions. The response so far is unanimously that the wording needs to be changed to cover each of the positions, except that Bubba73, so far as I understand, would regard the wording as OK in respect of at least position A shown, possibly also position B2.

The reason for Bubba73's viewpoint hinges on the meaning of the phrase "securely blockaded". I have discovered that different people have different opinions on this.

I think in view of the ambiguity the phrase needs further definition. The form of such a definition would also determine if exceptions of the type I show here run to millions or just thousands.

The Troitzky line rule is given without any attribution. I think it needs to be a cited reference. Further explanation or definition of the phrase "securely blockaded", should this occur in the citation, needs to be taken from the same source, otherwise, in view of the ambiguity, it could express something different from what the author intended.

I think the best source would be from the originator of the rule. (See subsequent questions.)

If the cited source doesn't define the phrase "securely blockaded" then I think some comment needs to be included about what the likely intention was. So far I have come across five views of what the term means:

(1) The knight stands on the square immediately in front of the pawn and is not en prise.

(2) The knight stands on the square immediately in front of the pawn and is directly defended by another white piece.

(3) The knight stands on the square immediately in front of the pawn and either the black king can be prevented from attacking the blockading knight or the knight can be directly defended if it is attacked, without, in the process, allowing the black king to capture the remaining knight or fork both knights in such a way that one can be captured on the subsequent move.

(4) Bubba73's meaning, which is, I believe, strictly between (3) and, "The knight stands on the square immediately in front of the pawn and all lines of play for black lead to a loss". (Bubba73 has dropped out of the conversation and referred me here, so my apologies to Bubba73 if this is a misrepresentation.)

(5) The knight stands on a square forward of the pawn on the same file, the square behind the knight being understood as the point at which the pawn is blockaded, and either the black king can be prevented from attacking the blockading knight or the knight can be directly defended if it is attacked, without, in the process, allowing the black king to capture the remaining knight or fork both knights in such a way that one can be captured on the subsequent move.

The last of these was my own understanding until I started the discussion. I now find that it is incorrect in terms of standard chess terminology. I've included it because I suspect this was what Troitsky intended.

I would say that (1) cannot be the intended meaning in this case because there are then very many obvious exceptions.

I think (2) can also be discounted, because White would often not want to directly defend the blockading knight. For example C is won for White, but he draws if he directly defends the blockading knight.

I would expect (3) to be the majority understanding among readers without any further explanation in the text. With this definition all the drawn positions shown here would be exceptions.

Meaning (4) with appropriate formulation could possibly encompass positions A and B2, but a formulation would be difficult, and no reasonable formulation could encompass C', so is probably not worth while. Some exceptions would have to be noted in any case.

Meaning (5) would possibly be best if there were any reference for it. (I've previously understood "blocked" and "blockaded" in an English sense, i.e. if you can't immediately move where you want to go because there's something in the way, you're blocked, whereas if there's a blockade at the end of your cul-de-sac, you're blockaded even if you live at the other end of the cul-de-sac. Standard chess usage, if I now understand correctly, makes no distinction in the case of individual pawns and each implies there is a piece immediately in front of the pawn.)

I'm lacking sources, so I'm hoping someone can assist. At the moment I have only (Averbakh & Chekhover 1977) relating to the two knights v pawn ending and it doesn't mention the Troitsky line. All the related internet references I can find are either verbatim copies of the Wikipaedia article (in the form prior to the recent changes) or links back to it. I've ordered (Troitzky 1937) but it could be a month before it arrives.

In the meantime, it would be useful if anyone could answer any of the following questions.

(a) Did Troitzky propound the Troitzky line rule, or was this someone else based on Troitzky's analysis. ALREADY ANSWERED BY Bubba73

(b) If the answer to (a) is "someone else", who, and is there an available reference? NO LONGER RELEVANT

(c) If the answer to (a) is yes, is the rule to be found in (Troitzky 1937)? If not, where? I THINK ANSWERED AS YES BY CHESS CAFE ARTICLE

(d) Does anyone know of a source for the Troitzky line rule that implicitly or explicitly uses meaning (5) of the term "securely blockaded"?

(e) Can anyone provide a citation and the corresponding text for the Troitzky line rule? I believe, from what Bubba73 says that both (Fine 1941) and (Müller & Lamprecht 2001) should contain such. (A reference authored by Troitzky would be preferred.) ALREADY ANSWERED BY Bubba73 BUT (f) NOT AVAILABLE FROM SOURCE GIVEN

(f) Assuming the answer to (e) is yes, is there a definition or succinct explanation of "securely blockaded" or whatever equivalent is used? Can this text also be posted, please, if so?

(g) If I can find no source with a definition/explanation of "securely blockaded", which of (1)-(5) do you think should be the suggested meaning inserted in the article, or should it be something different from all of these?

I will post a draft when I have more information.

  • Averbakh, Yuri; Chekhover, Vitaly (1977), Knight Endings, Batsford, ISBN 0-7134-0552-X
  • Fine, Reuben; Benko, Pal (2003), Basic Chess Endings (1941) (2nd ed.), McKay, ISBN 0-8129-3493-8
  • Müller, Karsten; Lamprecht, Frank (2001), Fundamental Chess Endings, Gambit Publications, ISBN 1-901983-53-6
  • Rabinovich, Ilya (2012) [1927,1938], The Russian Endgame Handbook, Mongoose, ISBN 978-1-936277-39-1
  • Troitzky, Alexey (2006), Collection of Chess Studies (1937), Ishi Press, ISBN 0-923891-10-2 The last part (pages 197–257) is a supplement containing Troitzky's analysis of two knights versus pawns.

Martin Rattigan (talk) 18:58, 14 June 2015 (UTC)

The Russian Endgame Handbook, by Rabinovich, page 88, quotes Troitzky "... on any placement of the black king, White undoubtedly wins only against black pawns standing on a4, b6, c5, d4, e4, f5, g6, h4 and above". Bubba73 You talkin' to me? 23:32, 14 June 2015 (UTC)
Well that's certainly interesting!
The snippet would no doubt be preceded in the original by some specification of the positions of the white pieces (i.e. a blockade in some sense) but the reference to "... black pawns standing on ..." seems to kill off my idea that Troitsky may have been intending something corresponding to definition (5) above, otherwise he would have said "... black pawns stopped on ..." or some such phrase.
Most interesting though is the word "only". I think he is expressing a necessary, not a sufficient condition that White undoubtedly wins. That is, he is saying that White does not undoubtedly win if the pawns are past the Troitzky line, not that White does undoubtedly win if the pawns are on or behind the Troitzky line. (Undoubtedly here meaning, presumably, for any position of the black king.)
welcome back by the way. — Preceding unsigned comment added by 80.218.32.135 (talk) 00:21, 15 June 2015 (UTC)
I think he is stating a sufficient condition, because (1) he gives positions of the pawn past the line where White may or may not win, and (2) he says that White undoubtedly wins if the black pawn is on or behind the line. Bubba73 You talkin' to me? 00:30, 15 June 2015 (UTC)
He seems to me to be saying that there are always positions of the Black king where White doesn't necessarily win if the pawn is past the line (of the kind you mention under (1)).
But as for (2), he doesn't say (according to your quote) "White undoubtedly wins if the black pawn is on or behind the line", he says "White undoubtedly wins only against black pawns standing on a4, b6, c5, d4, e4, f5, g6, h4 and above", i.e., "White undoubtedly wins only if the black pawn is on or behind the line", the critical addition being the word "only".
There is a difference in meaning between "if" and "only if".
I think he leaves open the situation on or behind the line.
80.218.32.135 (talk) 01:09, 15 June 2015 (UTC)
Of course, if I have understood the quote correctly, none of the positions I gave would be exceptions to what it states. In fact I don't believe there would be any exceptions.
No, he is saying that White always wins if the black pawn is on or behind the line. Bubba73 You talkin' to me? 01:31, 15 June 2015 (UTC)
Is he? I still don't see it. If "always" is taken as a synonym of "undoubtedly", then the quote would be, "White always wins only if the black pawn is on or behind the line", not, "White always wins if the black pawn is on or behind the line", as you have just stated it. Martin Rattigan (talk) 02:12, 15 June 2015 (UTC)
What do you think undoubtedly means?? (Realize that this is a translation from Russian.)Bubba73 You talkin' to me? 02:14, 15 June 2015 (UTC)
As I said above, for any position of the black king, but I would be more sure if the snippet were complete. Does Rabinovich give anything preceding the snippet? As for the translation, who knows? Martin Rattigan (talk) 02:39, 15 June 2015 (UTC)
Having had some sleep meantime, I see that I have probably completely misunderstood your previous response. In fact your response simply means exactly what it says; what do I think is the intended meaning of the word translated as "undoubtedly". My response probably appeared a little strange. No sane translator would render "for any position of the black king" as "undoubtedly". If you understand the original meaning as "always" I would concur. I wasn't questioning this substitution in my previous comment.
My point was about the logical form of the statement Rabinovitch attributes to Troitzky. This statement is only if where is "White undoubtedly(always) wins" and is "the black pawn is on or behind the Troitsky line". (My first question here would be, "Do you agree so far?.) You have rendered this in your last few comments as if , i.e. "White undoubtedly(always) wins if the black pawn is on or behind the Troitzky line". The first would be formalized as whereas the second would be formalized as and these are not logically equivalent.
The whole paragraph is: "Troitzky showed that 'on any placement of the black king, White undoubtedly wins only against black pawns standing on a4, b6, c5, d4, e4, f5, g6, h4 and above (26 cases)'. Along with this, he demonstrated a great number of losing positions with further-advanced pawns; however, in the vast majority of such positions the win is forced only with very complicated maneuverings, the explanation of which would take up space in this manual." Bubba73 You talkin' to me? 03:19, 15 June 2015 (UTC)
OK, yet another, revision on my part regarding your question about the meaning of undoubtedly. If, as seems reasonable, this means "always", then it must mean for all positions of some description. Then the logic as I stated it can't be so simple. "Always" would imply a universal quantifier over positions and the scope of this would need to include both and . I will have to think about that.
On the other hand the full paragraph places no constraints on the positions of the white pieces (i.e. no mention of any sort of blockade), so on the face of it Troitsky must be here proposing a rule about drawing possibilities when the pawn is not behind the line (when no such constraints need to be mentioned) rather than winning possibilities when it is behind the line. Otherwise the statement would be ridiculous. (That is assuming the paragraph doesn't occur in a section headed "Positions with the pawn securely blockaded" or some similar context.)
Martin Rattigan (talk) 22:12, 15 June 2015 (UTC)

Fine and Benko, page 99, say "White wins if the pawn cannot cross the line." Bubba73 You talkin' to me? 02:27, 15 June 2015 (UTC)

Thanks again. Do they have an actual statement of the Troitzky line rule? I.e. a statement that could be used as a citation. Martin Rattigan (talk) 02:39, 15 June 2015 (UTC)
Muller and Lamprecht. Fundamental Chess Endings, pp. 19-20, "The Russian theoretician Triotsky made a detailed study of this endgame and discovered the following rule: If the pawn is securely blockaded no further down than the line, then Black loses, no matter where the kings are. If the pawn has advanced beyond the line, there is usually a drawing and a losing zone for the defending king, which were also analyzed by Triotsky." (emphasis in the original) John Nunn did a computer analysis and found Triotsky's analysis to be "surprisingly accurate". Bubba73 You talkin' to me? 04:41, 15 June 2015 (UTC)
And thanks again for the full quote. This could form a basis. — Preceding unsigned comment added by 84.73.201.231 (talk) 15:59, 15 June 2015 (UTC)
This also answers my question (a) in the affirmative. I would expect that Troitsky, at least, would have had some detailed description of what he meant by "securely blockaded", or some categorization of exceptions. It's not plausible that he was unaware of drawn positions such as C'.
From the discussion above, I suspect that Rabinovitch's quote was for a necessary precursor to the rule given in M&L. Troitzky would probably have located the Troitzky line position by finding drawing zones as the pawn position was moved back from the queening square. If the quote is from (Rabinovich 1927) then Troitzky would have been working on the ending for some years after this. Martin Rattigan (talk) 23:32, 15 June 2015 (UTC)

Bubba73's contributions have already answered about half of the questions (a)-(e) I posed above (see capitalized comments following the questions). I'm hoping the remainder will be answered in (Troitzky 1937), which I expect to receive in about five weeks, so I think the discussion is probably best left dormant until then, but feel free to contribute if you think it will be useful. Martin Rattigan (talk) 13:19, 16 June 2015 (UTC)

In the meantime, I have invested in the Kindle version of M&L FCE and confirmed that there is no further definition of "securely blockaded".

However the M&L statement includes the phrase "no matter where the kings are" but omits any reference to the knights. It is then possible that the intended meaning of the phrase "securely blockaded" in M&L is none of (1)-(5) I suggested above, but a new one viz:

(2) A knight stands on the square immediately in front of the pawn and is directly defended by the other knight.

With this meaning the statement in M&L is correct (to the best of my knowledge).

I think that it would be clearer to give definition (2) explicitly (both in Wikipedia and in M&L), since the reasons I give for inferring it are somewhat tenuous.

I am attempting to contact Dr. Müller for confirmation, but so far without any progress. Can anyone advise me how I might establish contact with either author?

I will wait a week or so for possible contradictory comments, then (if there are none) prepare a draft amendment.

Martin Rattigan (talk) 12:51, 18 June 2015 (UTC)

Again, I think if the pawn is "securely blocked" then you're not forced to un-block it. Similarly, if you have to give up a knight, it is no longer a two knights versus pawn endgame. Muller writes for chesscafe.com, so you could try contacting him there. If you do, let us know what he says. Bubba73 You talkin' to me? 02:51, 19 June 2015 (UTC)
As I said above regarding your understanding of the phrase "securely blockaded"

Meaning (4) with appropriate formulation could possibly encompass positions A and B2, but a formulation would be difficult, and no reasonable formulation could encompass C', so is probably not worth while. Some exceptions would have to be noted in any case.

White in C' is forced neither to unblock the pawn nor give up a knight, so I am assuming that you agree that C' is an exception to the Troitzky line rule as stated in Wikipedia, whatever is understood by "securely blockaded". (White is, strictly speaking, also forced to do neither in positions A and B2 shown above). But in any case I would expect readers to understand one of the definitions (1), (2) or (3) in this context.
From the rule in M&L, meaning (2) can be tenuously inferred. This because the M&L statement includes the phrase, "no matter where the kings are", not as in Wikipedia, "no matter where the other pieces are". If definition (2) were intended, I believe the statement in M&L becomes correct. (Notice that the blockading knight is defended by the other knight in none of the draws I gave above.)
I already sent an email to Chess Café asking for Dr. Müller's e-mail address but so far have no response.
Martin Rattigan (talk) 09:34, 19 June 2015 (UTC)
abcdefgh
8
h4 black pawn
b3 white king
h3 white knight
f2 white knight
b1 black king
8
77
66
55
44
33
22
11
abcdefgh
E. White to play and win

Bubba, you say above

I think if the pawn is "securely blocked" then you're not forced to un-block it. Similarly, if you have to give up a knight, it is no longer a two knights versus pawn endgame.

I think by this you intend to imply that the wording in the Troitzky line section is acceptable as it stands. Forgive me if I'm wrong.

If this is what you intended then I think your position is untenable.

If you consider the position E to the right, I would say this is a standard win under the Troitzky line rule as stated. However, to win white is forced to both unblock the h4 pawn and offer Black one of his knights.

According to what you say, it would seem you neither consider this position to be a two knights v pawn ending nor consider the h4 pawn to be securely blocked.

I don't think you could possibly expect readers of the article to concur.

Can we agree that the section is in need of revision?

Martin Rattigan (talk) 19:52, 19 June 2015 (UTC)

Bubba's point is that there's hardly a difference between calling positions A, B2 and B3 exceptions to the rule and calling a situation where Black is to move and one of White's knights is en prise an exception to the rule. Technically this is true, but nobody writes things like "king and queen can checkmate a lone king but only if the queen is not en prise". The difference in position E is that Black cannot force a draw by winning one of the knights, but White can force a win by giving one of them up. Only position C' presents a difficulty for the current phrasing, although it is such a contrived situation (how did the knight end up on a8? it would make a nice final position for a study) that it wouldn't surprise me if prior literature had not considered it, and if that were the case, amending the article to take this one situation into consideration would strictly speaking be WP:OR. Cobblet (talk) 00:31, 23 June 2015 (UTC)
The same thought about making it a line in an endgame study had occurred to me. But if the wording were changed to that in (Müller & Lamprecht 2001) and definition (2) added for the phrase "securely blockaded" (see Dr. Müller's comments below) then the statement becomes strictly correct. All exceptions, including "one of White's knights is en prise", would be covered by the statement. I am hopeful that I may find an equivalent formulation in (Troitzky 1937). I assume neither option would fall foul of WP:OR. Martin Rattigan (talk) 10:11, 23 June 2015 (UTC)
And of course nobody writes, "king and queen can checkmate a lone king but only if the queen is not en prise", but neither do they write "king and queen can checkmate a lone king no matter where the pieces are placed". Readers will discount the queen en prise situations so long as the italicized phrase is omitted, but otherwise they will just think it's incorrect. Also, in position E, though Black can't force a draw by taking the knight, he can force a draw if White doesn't give up the blockade on h4 (similarly in the drawn positions shown). The rule has no use if the same interpretation can't be used to distinguish between the situations it is intended to cover and the exceptions. Martin Rattigan (talk) 10:39, 23 June 2015 (UTC)
abcdefgh
8
c8 white knight
e7 black pawn
e6 white knight
b5 black king
e5 white king
8
77
66
55
44
33
22
11
abcdefgh
B3. Black to play and draw

The draws are also exceptions to the second Troitzky line rule (except that B2 would need to be moved two squares toward the queen side; B3 right). This section therefore also needs to be changed. References for a revised version could be problematic unless the result has been updated with the advent of the Syzygy DTM50 EGTBs.

Can anybody with a set of Syzygy EGTBs confirm that the line is in fact valid. (It would have been difficult and possibly prone to error to establish the result with the Nalimov DTM EGTBs.) I will download the Syzygy EGTBs myself and check, but I'm quite slow at that sort of thing and busy with other things at the moment.

If it isn't valid then it might be better to drop the section.

Martin Rattigan (talk) 19:17, 21 June 2015 (UTC)

Dr. Muller's article here strongly suggests that his interpretation of "securely blockaded" is not (2). This means I can't just cite M&L and give meaning (2) as the definition of the phrase. But with luck I can cite directly from (Troitzky 1937) - this is the source given in the article. (I think question (c) is answered in the affirmative.)

Martin Rattigan (talk) 21:37, 21 June 2015 (UTC)

For info, I wrote, "I have been soliciting the views of chess players on whether the current wording is adequate to cover various drawn positions. The response so far is unanimously that the wording needs to be changed". I now have two opposing views (not including Bubba and Cobblet, who I hope to convince; these from here) so it's standing at 6 in favour of change, 2 against to date. Martin Rattigan (talk) 23:59, 22 June 2015 (UTC)

Dr. Müller has responded to my email to Chess Café. He wrote:

You can use the quote from FCE. You can of course also add your definition of "securely blockaded" as the blockading knight is protected by the other knight. In FCE Frank Lamprecht and I indeed used "no matter, where the kings are" to deal with this problem. Of course it would be best, if an original quote from Troitzky can be found and used.

Martin Rattigan (talk) 09:16, 23 June 2015 (UTC)

Discussion as of June 26

Since I proposed to produce a draft (#Draft proposed), objections to changing the current wording have been raised by Bubba (#Bubba's objection) and Cobblet (#Cobblet's objection). I don't accept either of the objections as valid. I will deal with each in detail.

A. Bubba's objection.

(i) Bubba says, "I think if the pawn is "securely blocked" then you're not forced to un-block it". This I think means to say that the phrase "securely blocked" doesn't apply if Black can force White to unblock the pawn in subsequent play.
(a) If black plays Kf6 in A, Kc6 in B2, any in C' or Kc6 in B3, then the position is drawn.
(b) In a drawn position the side to move has no winning line. He is forced only to make moves which don't lose, maintaining the draw.
(c) In each of the positions, after Black's first move White can maintain the draw with or without the knight at the edge of the board. This knight is then of no value to White. If black attacks it then White is not forced to defend it in terms of the object of the game.
(d) In each case, if White does not defend the knight (if attacked) he can maintain the blockade. Therefore, according to Bubba's interpretation of the phrase, the pawn is "securely blockaded" in each of the draws I've shown.
(e) By contrast, in position E white is forced to abandon the blockade on h4, because if he does not he loses a half point.
(f) This also applies to any position won for White with the pawn blockaded on or behind the Troitzky line, except for a hundred odd trivial and unforceable mate or mate in 1 positions.
(g) The conclusion then is that if Bubba's contention (i) be taken seriously, the Troitzky line rule as stated applies only to draws of the kind I show and the trivial positions mentioned.
(h). This is obviously not what Troitzky intended.
(ii) Bubba says, "if you have to give up a knight, it is no longer a two knights versus pawn endgame"
(a) True. In many of the won endings with the pawn blockaded on or behind the Troitsky line the pawn will promote and in some variations with a rook's pawn Black will take a knight. These are then also no longer two knights versus pawn endgames.
(b) The Troitzky line rule as stated makes no mention of the material in following play, therefore with the current wording (ii) is irrelevant. The positions I've shown are all two knights versus pawn endings in the sense that White's material (apart from the king) is two knights and Black's material (apart from the king) is one pawn.

B. Cobblet's objection.

(i) Cobblet says, "Bubba's point is that there's hardly a difference between calling positions A, B2 and B3 exceptions to the rule and calling a situation where Black is to move and one of White's knights is en prise an exception to the rule. Technically this is true, but nobody writes things like "king and queen can checkmate a lone king but only if the queen is not en prise". "
abcdefgh
8
d5 white king
d4 white knight
g4 black pawn
d3 black king
g3 white knight
8
77
66
55
44
33
22
11
abcdefgh
F. Black to play and maintain the draw
(a) The draws I gave were deliberately kept simple. They were intended to make the point that a change in wording was necessary, not as endgame studies. Having said that, I don't see that a drawing line for Black in A, B2 or B3 is significantly simpler than than a drawing line for Black in the diagram F, right. (This is point at which the Topalov versus Karpov game mentioned became a two knights versus pawn endgame and Karpov failed to maintain the draw.)
I disagree with the third sentence. Cobblet (talk) 03:30, 27 June 2015 (UTC)
Then, in the absence of an opportunity to play the endings over the board we will just have to agree to disagree. I think Kf3-g2 in F just as obvious as the drawing lines in the draws I gave. 80.2.119.0 (talk) 19:41, 27 June 2015 (UTC)
So you're better at endgames than Karpov. I guess I'm out of my league here. Cobblet (talk) 00:49, 28 June 2015 (UTC)
"So you're better at endgames than Karpov."
Karpov is regarded by many as the greatest endgame player ever. The two knights versus pawn ending is not playable over the board, even by the strongest grandmasters, without them having studied it. Karpov freely admitted after the game that he had never looked at the ending. A player's skill in this ending is determined more by his level of study than by any innate talent for chess. Karpov felt that the time required to study the ending was not justifiable in view of its rarity of occurrence. Karpov is far from alone in this respect. See, for example here and here. (I believe Gurevich went on to win the tournament reported in the second link.)
"I guess I'm out of my league here."
Agreed.
Martin Rattigan (talk) 09:02, 30 June 2015 (UTC)
(b) Readers will automatically discount situations where the knight is en prise or the knights can be effectively forked ,even when, as with the current wording (i.e. "no matter where the other pieces are place") it specifically says they are not to be discounted. I think it is too much to expect them to automatically discount draws of the kind I've shown.
(c) Even assuming they do say to themselves, "but that is obviously not meant to include situations where the Black king can immediately take a knight or fork both, or the non blockading knight is in a corner and and the Black king can move to the adjacent or next square along the diagonal from the corner (excepting when the blockading knight already covers an exit the other knight), or the non blockading knight is in a corner on its eighth rank and the blockaded pawn is the nearest bishop's pawn on its starting square and the black king is not in the opposite corner on the same rank with the white king (to move) on its sixth rank either in opposition to the black king or one square nearer the corner knight, or the pieces are in the configuration shown in B2 with black to move and the pawn on either c6, d6 or f6 (or the reflections), or ...", then what exactly does Cobblet expect them to put in place of the dots?
"but that is obviously not meant to include situations where the defender can either force the win of one of the knights or the loss of his remaining pawn" is what I expect a reader to surmise. This covers all the exceptions you gave except C', does it not? Cobblet (talk) 03:30, 27 June 2015 (UTC)
No it does not. But I think you just proved my point about expecting too much of the reader.
If you were to add "or the removal of the blockade" you would rule out all the draws I've shown except C', but you would also rule out all the non trivial wins that the rule is intended to cover. (See section (i) of my response to Bubba's objection.) 80.2.119.0 (talk) 19:41, 27 June 2015 (UTC)
(d) The exceptions they are expected to tacitly assume in (c) would have to be a complete list of situations. They cannot' replace it by, "but that is obviously not meant to include situations where Black can force a draw", because then it would simply become a truism and they would finish with no Troitsky line rule at all. It would apply wherever the dots were in the Troitzky line diagram.
(e) In view of (b)-(d) I believe that either the rule as stated must be replaced by a correct version (e.g. as in (Müller & Lamprecht 2001)), or it should be explicitly stated that the rule given is incorrect and there are numerous exceptions (preferably the former).
(ii) Cobblet says, "The difference in position E is that Black cannot force a draw by winning one of the knights, but White can force a win by giving one of them up. Only position C' presents a difficulty for the current phrasing ..."
(a) I agree that the first sentence states a difference between A or B2/3 and E, but it doesn't seem to say anything significantly more than White can offer a knight in all the positions, but E is a win and the others are draws.
(b) There seems to be an implication that the portion of the second sentence here is a consequence of the first. I would say it is a non sequitur. (I may have assumed that an implication was intended when it was in fact not.)
(iii) Cobblet says, "Only position C' presents a difficulty for the current phrasing, although it is such a contrived situation (how did the knight end up on a8? it would make a nice final position for a study) that it wouldn't surprise me if prior literature had not considered it, and if that were the case, amending the article to take this one situation into consideration would strictly speaking be WP:OR"
abcdefgh
8
a8 white rook
e8 black rook
h8 black king
c7 black pawn
f7 black pawn
g7 black pawn
h7 black pawn
e6 black pawn
h6 black bishop
b5 white bishop
d5 black pawn
a4 white pawn
c4 white knight
d4 white pawn
b3 white pawn
h3 white pawn
b2 white king
d2 white pawn
e2 black queen
f2 white pawn
g2 white pawn
b1 white knight
h1 white rook
8
77
66
55
44
33
22
11
abcdefgh
G. Black to play.
(a) For the reasons given in (ii) and in response to #Bubba's objection, I think all the positions I gave represent a difficulty for the current phrasing.
(b) The Troitsky line rule would be invalidated by a single exceptional position. Since C' is a draw in all but two placements of the two kings, this position alone represents a large number of exceptions.
(c) The wording doesn't include, "so long as play leading up to the position is not silly". This is not a normal assumption in endgame positions.
The point is not whether the position is silly but whether previous literature has considered it. Cobblet (talk) 03:30, 27 June 2015 (UTC)
I can't find anything in WP:OR to support what you say. Clearly it would be impracticable to read all previously published literature to check that C' hasn't been previously published or find a position with different king positions that has never been published. Could you clarify the point you make with reference to WP:OR please? — Preceding unsigned comment added by 80.2.119.0 (talk) 20:45, 27 June 2015 (UTC)
(d) The knight could well end up in the corner from a position similar to G, right (though this position may also be regarded as unlikely and it's unlikely that it would finish up as two knights versus pawn). At any rate C' is not inconceivable.
(e) If the wording were changed to use the (Müller & Lamprecht 2001) formulation together with definition (2) of "securely blockaded" or a similar formulation from Troitzky, I don't see how WP:OR would be involved.
I have no problem with quoting FCE verbatim, but extrapolating definition (2) from the words "securely blockaded" would definitey constitute OR unless you have a reliable source directly supporting this very strict limitation on the placement of the second knight, which is most certainly not an immediately obvious interpretation of those two words. Per the FAQ on Wikipedia talk:Identifying reliable sources, personal communication from an expert does not constitute a reliable source. Cobblet (talk) 03:30, 27 June 2015 (UTC)
I didn't extrapolate (2 _ ) from the words "securely blockaded", but from the phrase, "no matter where the kings are". I took that as the "obvious" implication. How immediately obvious this is is possibly open to debate. I am not invoking Dr. Mueller's email to support this, I included the email above only because Bubba requested it. 80.2.119.0 (talk) 19:41, 27 June 2015 (UTC)
And it seems to me that he clarified what he meant by "securely blocked" and said that the exceptions were covered. Bubba73 You talkin' to me? 20:03, 27 June 2015 (UTC)
Yes, indeed; I have already pointed this out. The exceptions I gave are not exceptions to the rule as stated in (Müller & Lamprecht 2001). Neither are positions in which Black can capture a knight (there are none) or fork the knights.
But Muller said "... the blockading knight is protected by the other knight...", but the blockading knight is not protected by the other knight in C'. Bubba73 You talkin' to me? 22:52, 27 June 2015 (UTC)
Precisely. Nor in any of the other draws I showed. The statement in M&L is correct (and different from the incorrect one in Wikipedia). Martin Rattigan (talk) 23:50, 27 June 2015 (UTC)
Well, if that is what he meant by "securely blocked" then position C' doesn't matter. Bubba73 You talkin' to me? 23:56, 27 June 2015 (UTC)
All the positions are exceptions to the rule stated in Wikipedia. That is why it's incorrect. None are exceptions to the rule stated in(Müller & Lamprecht 2001). That's because it is correct. That's why I propose to replace the one in Wikipedia by the one in (Müller & Lamprecht 2001) (or an equivalent from Troitsky). Position C' and the others will then no longer matter. But they do as it stands. Martin Rattigan (talk) 00:19, 28 June 2015 (UTC)
I cited and checked references I have, and I have almost all endgame books written in English. I can't say something that isn't in a reference, see wp:cite, wp:RS, WP:VER, and wp:or. Bubba73 You talkin' to me? 00:40, 28 June 2015 (UTC)


It's the wording in Wikipedia that I'm questioning. I am planning to replace this with the version from (Müller & Lamprecht 2001) if I can't source it from Troitsky. 80.2.119.0 (talk) 20:58, 27 June 2015 (UTC)
Just to be clear: I have absolutely no problem with you replacing the current wording with M&L's. But I maintain that your interpretation of "no matter where the kings are" cannot be readily inferred from M&L and hence constitutes OR. It's not obvious to me that M&L meant to exclude positions like wKh8,Nd3,c7;bKg6,Pd4 (the first example he gives as a win after making the "securely blockaded" statement in the Chesscafe article) with his wording – he did according to your interpretation. (You said is, not can be.) Cobblet (talk) 00:49, 28 June 2015 (UTC)
I agree that, "the blockading knight is or can be guarded by the other knight" is a more correct interpretation of the rule stated in M&L than "the blockading knight is guarded by the other knight". The two are obviously equivalent, so I would be happy to include the former should I fail to get a suitable formulation from Troitzky. Martin Rattigan (talk) 02:48, 28 June 2015 (UTC)
They are obviously not equivalent or else I wouldn't have made my comments. This is the point of WP:OR – say what the source says, don't impose your own interpretation of it which may not be shared by everyone. We should not qualify M&L's statement unless M&L publish a qualification themselves. Cobblet (talk) 03:01, 28 June 2015 (UTC)
If won position B can be reached from position A then A is obviously also won. To say they are not equivalent is just nit picking. You are allowed to make obvious comments on the citations. Martin Rattigan (talk) 03:19, 28 June 2015 (UTC)
No, not if B can be reached from A - it has to be forced. Bubba73 You talkin' to me? 03:30, 28 June 2015 (UTC)
Obviously. Martin Rattigan (talk) 03:44, 28 June 2015 (UTC)
abcdefgh
8
a6 white king
h6 black pawn
d3 black king
a2 white knight
8
77
66
55
44
33
22
11
abcdefgh
H. White to play

So you're telling me that in position H, it is "obviously" correct to say that White's knight guards h2, because White can achieve this by force in eight moves. To me these are not equivalent statements – the second part is not obvious to all but the most sophisticated of endgame players. Your interpretation of the word "guards" is non-standard. Cobblet (talk) 04:06, 28 June 2015 (UTC)

By saying a piece "guards" another piece of the same colour, I mean it could move to the square of the piece it guards were the guarded piece not there. This should have been clear from my phrase "is or can be guarded". I believe this is a perfectly standard use of the word.
By saying the blockading knight can be guarded by the other knight I mean the other knight can move to guard the blockading knight.
By saying, "The two are obviously equivalent", I mean that the implication, "If won position B can be (forcibly) reached from position A then A is obviously also won" is obvious. It follows (just as obviously) that a position is won according to the M&L version of the Troitsky line rule with one definition of "securely blocked" if and only if it is won according to the other. I am not saying that if a position B can be forcibly reached from position A then the play required to reach it is obvious. That would be fatuous.
I would not say in the position you gave that h2 is guarded by the knight. For this the knight would have to be on one of the squares f1, f3 or g4. Neither would I say that h2 can be guarded by the knight because Black can ensure that White cannot move his knight to any of the squares mentioned and still maintain the draw.
80.2.119.0 (talk) 01:41, 29 June 2015 (UTC)
OK, looks like I misunderstood you. So in M&L's example I mentioned above (wKh8,Nd3,c7;bKg6,Pd4), is the pawn "securely blockaded" or not? It seems we agree that the knight on d3 is not guarded by the other knight – originally I thought you were implying otherwise. Cobblet (talk) 02:24, 29 June 2015 (UTC)
The knight is securely blockaded with the meaning "is or can be guarded" because Black cannot prevent Na6-c5, whoever has the move. With the meaning "is guarded", it is not guarded, therefore not securely blockaded. However a position which is securely blockaded can be reached (again because Black cannot prevent Na6-c5), therefore, that the position is won for White also follows from the Troitzky line rule stated with this meaning.
Note that, as correctly stated on the page, the Troitzky line rule does not take into account the 50 move rule and I would guess this particular position could not be won if the 50 move rule were enforced (though a DTM50 EGTB could possibly prove me wrong). 80.2.119.0 (talk) 08:19, 29 June 2015 (UTC)
Well I'm glad you recognize the distinction. M&L's statement was simply that the pawn had to be securely blockaded, not that a position where the pawn was securely blockaded could be reached by force. It is neither Wikipedia's job to to make the latter inference for the reader (as trivial as it may or may not be), nor even to interpret what "securely blockaded" means in the first place (before this discussion I had never seen anyone else express the idea that the knights had to be able to defend each other for the Troitzky line to apply). Cobblet (talk) 12:17, 29 June 2015 (UTC)
I recognize the distinction. It's totally trivial and hardly worth discussing. You can't inflate the importance of your point by repeating your arguments ad nauseam. I don't intend to continue discussion on this particular point. If the citation will come from Troitsky it will be in any case irrelevant.
It is acceptable to make obvious comments on quoted sources and it is Wikipedia's job to present its subject matter clearly and accurately.
"before this discussion I had never seen anyone else express the idea that the knights had to be able to defend each other for the Troitzky line to apply"
Neither had I. I became aware of it only after Bubba posted the quote from M&L. Prior to discussion I googled "Troitzky line" and found only verbatim copies of the Wikipedia article and links back to it. So it's probably not surprising that you haven't encountered it previously. Martin Rattigan (talk) 09:02, 30 June 2015 (UTC)
You're one to complain about repeating arguments ad nauseam. From WP:OR: "Wikipedia does not publish original thought: all material in Wikipedia must be attributable to a reliable, published source. Articles may not contain any new analysis or synthesis of published material that serves to reach or imply a conclusion not clearly stated by the sources themselves." From WP:V: "Even if you're sure something is true, it must be verifiable before you can add it." "In Wikipedia, verifiability means that anyone using the encyclopedia can check that the information comes from a reliable source." From WP:OR again: "Even with well-sourced material, if you use it out of context, or to reach or imply a conclusion not directly and explicitly supported by the source, you are engaging in original research." That is what you are doing here. Cobblet (talk) 10:05, 30 June 2015 (UTC)

I have removed the final sentence from the second paragraph of the "Troitzky line" section which was incorrect and unsourced and replaced it with the Rabinovich quote Bubba provided earlier, which is similar but correct. I have also inserted two "citation needed" flags pending deletion or replacement of sentences preceding the flags. The first sentence, "Therefore the ending is more of theoretical than practical interest.", is unsourced and appears to express the author's own view. The other sentences (starting the subsection "Pawn beyond the Troitzky line"), "If a pawn is beyond the Troitsky line, the result usually depends on the location of the defending king. Usually there is a "drawing area" and a "losing area" for the defending king, which was also analyzed by Troitsky.", are either unsourced, or, if the source is meant to be M&L as inserted at the end of the paragraph, both innaccurate and taken out of context rendering it grossly incorrect. — Preceding unsigned comment added by 80.2.119.0 (talk) 09:25, 1 July 2015 (UTC)

I have deleted the sentences referred to. Martin Rattigan (talk) 22:34, 7 July 2015 (UTC)

Unless further objections are raised, or objections are raised to the above reasons for dismissing the objections so far raised, I now plan to produce a draft after I have received (Troitzky 1937). I expect this to be ready near the end of July.

Martin Rattigan (talk) 20:53, 26 June 2015 (UTC)

I'll withhold further comments until said draft is produced. I have a fear that Troitzky will disappoint you – let's hope I'm wrong. Cobblet (talk) 03:30, 27 June 2015 (UTC)

I have just received (Troitzky 1937). Although it looks as if there is enough analysis to support a carefully constructed version of the Troitsky line rule (I haven't had time to check this) there appears to be no explicit mention of the Troitsky line or the rule. Moreover it wouldn't be feasible to construct a readily comprehensible version using only direct quotes without forming a conclusion from the quotes that is not explicitly contained in the original. Troitsky's version of the rule (assuming it exists) could well be from the same source as Rabinovich quotes, since the line was there explicitly mentioned. But from the paragraph given this could be (Troitzky 1937) if Rabinovich first included the paragraph in the 1938 edition, because he says, "Troitsky showed ..." rather than "Troitsky stated ...".

For the time being I will produce a version using (Müller & Lamprecht 2001). The reader will have to do without an explicit translation of "securely blockaded" (courtesy of Cobblet and the Wiki legal system), but the source contains further explanation which I will also include, so he should be left in no doubt.

This should be ready in a week or so. Martin Rattigan (talk) 23:24, 7 July 2015 (UTC)

Draft amendment to Troitzky line section

_______________________________________________________________________________________________________________________ (this line not for inclusion)

/*==Troitzky line==*/ Troitzky line

The Troitzky line
abcdefgh
8
b6 black circle
g6 black circle
c5 black circle
f5 black circle
a4 black circle
d4 black circle
e4 black circle
h4 black circle
8
77
66
55
44
33
22
11
abcdefgh
Troitzky line positions when White has the two knights and Black the pawn.
Müller and Lamprecht 2001
abcdefgh
8
g6 black pawn
g5 white knight
h3 white knight
8
77
66
55
44
33
22
11
abcdefgh
Black loses no matter where the kings are. (Müller & Lamprecht 2001)
Kling & Horwitz 1851.
abcdefgh
8
g3 white king
d2 white knight
f2 black pawn
f1 white knight
h1 black king
8
77
66
55
44
33
22
11
abcdefgh
White to play and win.

1. Kh4 Kg2 2. Kg4 Kg1 3. Kh3 Kh1 4. Ng3+ Kg1 5. Nf3#

The pawn does not move; it assists mate by blocking the king's escape.

Whilst two knights cannot force checkmate (with the help of their king) against a lone king, a decrease in material advantage allowing the defending king to have a pawn can actually cause his demise. The reason that checkmate can be forced is that the pawn gives the defender a piece to move and deprives him of a stalemate defense (Müller & Lamprecht 2001:19–20). Another reason is that the pawn can block its own king's path without necessarily moving (e.g. Kling & Horwitz position right).

The Troitzky (or Troitsky) line (or Troitzky position) is a key motif in chess endgame theory in the rare but theoretically interesting ending of two knights versus a pawn.

The line, assuming White has the two knights and Black the pawn, is shown left.

The Russian theoretician Troitsky made a detailed study of this endgame and discovered the following rule:

If the pawn is securely blockaded by a white knight no further down than the line, then Black loses, no matter where the kings are.

— Karsten Müller and Frank Lamprecht, Fundamental Chess Endings 2001

An example of the application of this rule is given in the diagram Müller and Lamprecht right; "... the position would be lost no matter where the kings are". (Müller & Lamprecht 2001)

However, the checkmate procedure is difficult and long. In fact, it can require up to 115 moves by White, so in competition often a draw by the fifty-move rule will occur first (but see this article and Second Troitzky line section for the zone where the win can be forced within fifty moves).

Troitsky showed that "on any placement of the black king, White undoubtedly wins only against black pawns standing on [The Troitzky line] and above" (Rabinovich 1927:88).

John Nunn analyzed the endgame of two knights versus a pawn with an endgame tablebase and stated that "the analysis of Troitsky and others is astonishingly accurate" (Nunn 1995:265).

Even when the position is a theoretical win, it is very complicated and difficult to play correctly. Even grandmasters fail to win it. Andor Lilienthal failed to win it twice in a six-year period, see Norman vs. Lilienthal and Smyslov vs. Lilienthal. But a fine win is in a game by Seitz, see Znosko-Borovsky vs. Seitz (Giddins 2012:26).

Two knights versus pawn is sometimes called the "Halley's Comet" endgame.[1]

_______________________________________________________________________________________________________________________ (this line not for inclusion)

notes

(i) I will make the change in a few days if there are no objections.
(ii) I have not checked existing references, I am assuming they are correct.
(iii) I have inserted (Müller & Lamprecht 2001) as a source in the Trotsky line rule, the following sentence and the corresponding diagram. I have the Kindle edition so I cannot insert page numbers in these. If anyone has the paper edition I would be grateful if they could insert these after the change. Kindle location is 403.
(iv) I have replaced the repetition of the Troitzky line position in the Rabinovich quote.
(v) Heading is intended to remain the same size. Altered here to avoid creating new section.
(vi) I have made a small deletion in the second sentence regarding practical importance, but left the less contentious wording to stand. The sentence is unsourced and represents author's opinion, but I will leave others to delete or change it if they feel this is necessary (but note that the wording "the line" in the following sentence would need to be changed to "the Troitzky line" if it were deleted).
(vii) The same point about unsourced material applies to parts of the other paragraphs I have left unchanged, but I haven't attempted any correction to this.

Martin Rattigan (talk) 17:33, 14 July 2015 (UTC)

Discussion as of 14 July 2015

Ratings bot

Is the ratings bot not working or is there a formatting problem with some articles? For example, at articles such as Watu Kobese, the rating is not automatically being filled in. Greenman (talk) 17:47, 10 July 2015 (UTC)

Noticed that with Viswanathan Anand though in that one it wasn't updating the rating. Ended up removing the FIDE ID and rating bits of the infobox. Jkmaskell (talk) 20:27, 10 July 2015 (UTC)
Does anyone know the bot/user that normally does this, so we can follow up? Greenman (talk) 20:49, 10 July 2015 (UTC)
I think you want to look at WT:WikiProject Chess/Archive 21#Automatic updated FIDE rating in infobox (by bot). I thought it was a nice hack. Quale (talk) 05:44, 13 July 2015 (UTC)
Hey, I think the bot in question is User:DrTrigonBot, who updated Template:Elo rating. It doesn't work anymore, the API changes and the labs outage last month probably broke it. Unfortunately DrTrigon has been inactive and unreachable for almost a year. I've started running his bot scripts in October on dewiki (and some smaller wikis) and I would be able to do the same here. However I don't have the time to implement new features or fix edge cases, I only make sure it runs in its current state.
Are there any plans by someone to add the FIDE elo ratings to wikidata (this would be much better, as quite a lot of wikis want to use it)? If not, I'll contact BAG to run DrTrigons script from my bot account. Sitic (talk) 15:28, 21 July 2015 (UTC)

Apparently there's a new game called "women's chess"

I'm not happy with all these new categories at all, e.g. Category:Women's chess players. MaxBrowne (talk) 15:42, 18 August 2015 (UTC)

I agree, and I think it's trouble. "Women's chess" isn't really distinct from "chess" in the same way as in many other sports, since the rules don't change and because women are welcome in all chess competitions. If we did have an article for the category, it would probably be called Women in chess or something similar, rather than Women's chess which wouldn't make much sense. On the other hand, there is a Women's Chess Championship and Women's Olympiad and other team events for women only, and separate titles for women.
I preferred things the way they were before introduction of those "women's chess" categories, and would rather not divide the chess biography categories by gender at all. Others have a different opinion, and I think about five years ago one or two chess editors said that they wished the women were in separate categories. One point in favor of the subcategories by gender is that they can be members of other parts of the category tree such as Category:Sportswomen by sport.
If separate categorization is to be done, I think "female chess players" would be a better term than "women's chess". And if gender-specific categories are really required, the men should be put in separate "male chess players" categories. Asymmetric treatment of men and women in the categories is blatantly sexist. Quale (talk) 23:26, 19 August 2015 (UTC)
What Quale said. How do we proceed? Cobblet (talk) 02:25, 20 August 2015 (UTC)
I've notified the editor who created the categories of this discussion and invited them to join this discussion. If we get some more input from that editor or others, we should try to decide what we think should be done. As I see it, the three main options are to 1) accept the new categories as they are, 2) return to the previous practice of having no gender-specific player categories, or 3) rename "women's chess players" to something such as "female chess players" and create corresponding "male chess player" categories. Everything except leaving things as they are now would require a lot of work, but the fine folks at Categories for Discussion (CFD) have automated tools that they use to do this efficiently. Once we think we know which option we prefer then a proposal and discussion at CFD is probably appropriate anyway. Participants there may have a broader view of the category system and usual practices than we do. Quale (talk) 06:00, 20 August 2015 (UTC)
Not so sure about "male chess players" categories, since there are no male only events that I know of. I've looked at other sports & activities to see how they're handled on wikipedia. The only major sport I know of where men and women compete on completely equal terms is equestrian/eventing; they have a Category:Equestrians; among the many sub-categories are two little-used categories Category:Male equestrians and Category:Female equestrians which should probably be deleted.
For something more comparable with chess, there are no gender-related subcategories to Category:Go players. As with chess professional players are predominantly men, but there are a few very strong female players who can compete at the highest levels, e.g. Rui Naiwei. There are no gender-related subcategories to Category:Bridge players. Poker has a category Category:Female poker players but this isn't divided up any further. There are no gender-related subcategories to Category:Backgammon players.
I'd suggest we etiher adopt the poker categories as a model (i.e. have national sub-categories, have a female sub-category, but don't combine them) or have no gender subcategories at all. MaxBrowne (talk) 06:18, 20 August 2015 (UTC)

I can answer the question "Why we need the female categories?". There are more and more players in many categories where are men and women together. I want find women only and this isn't easy. I waste time for searching a few women among hundreds of men. I think "adopt the poker categories" is good idea (Category:Women's chess players -> Category:Female chess players). Examples:

I think first part of separating can be doing by bot with using category Category:Chess woman grandmasters Mircea (talk) 06:49, 20 August 2015 (UTC)

As far as I can tell WP:Cat/gender is the relevant policy. It's interesting in that it suggests something like Category:Female heads of government "does not need to be balanced directly against a "Male heads of government" category, as historically the vast majority of political leaders have been male." Whether that's also true of chess players is possibly debatable. Otherwise the poker model (so no categories combining sex and ethnicity) seems fine to me. Cobblet (talk) 09:23, 20 August 2015 (UTC)
I think "combining sex and ethnicity" in chess players categories is important - see examples for American, French and Russian above. Otherwise in polish wikipedia is category for Polish female chess players with 104 pages! So, "adopt the poker categories" - yes, but only wording "Female" and use "combining sex and ethnicity". Mircea (talk) 09:43, 20 August 2015 (UTC)
I'm not really convinced there is a need for this. My main concern is of notability, which for some female articles is suspect. We would theoretically have a couple of hundred extra categories for which there would be for the majority a handful of players and for which simply playing chess is a qualification. Surely the female title categories are good enough, if only to confirm their notability? Jkmaskell (talk) 09:59, 20 August 2015 (UTC)
Interesting too that WP:OCEGRS suggests that "dedicated group-subject subcategories... should only be created where that combination is itself recognized as a distinct and unique cultural topic in its own right. If a substantial and encyclopedic head article (not just a list) cannot be written for such a category, then the category should not be created." Is there another (maybe less obscure) guideline that also articulates this philosophy behind category creation? At first blush it would seem that this rules out the idea of categories combining sex and ethnicity – I don't think Polish women in chess (or women in chess of any particular nationality) qualifies as a distinct and unique cultural topic. On the other hand, Polish Women's Chess Championship could likely be a substantial article – would that mean that Category: Polish women's chess champions is a reasonable category? Is that really a "better" category to have than Category:Polish female chess players? (Perhaps it is, I'm not 100% sure. Maybe the folks at CfD will know.) Cobblet (talk) 14:16, 20 August 2015 (UTC)
I've raised this at WP:CFD, see Wikipedia:Categories for discussion/Log/2015 August 21. This is probably the best place to continue the discussion. MaxBrowne (talk) 01:11, 21 August 2015 (UTC)

Have a look. Just created it. Probabaly needs more sources. -Koppapa (talk) 16:45, 4 September 2015 (UTC)

elo update

The Elo in the infoboxes unfortunately is still from July, i.e. 2 months old. Example: Ralf Åkesson. Who is updating this? Best regards, --Gereon K. (talk) 16:15, 10 September 2015 (UTC)

See the #Ratings bot section just above. It was pretty cool when it was working, so maybe someone can get it going again. Quale (talk) 02:36, 11 September 2015 (UTC)

World Cup results tables

How does everyone feel about the addition of World Cup results tables? This has been done over numerous articles in the past couple of days and I'm not keen on them, seeing as they duplicate information which should already be in the main section as prose. I'm also concerned that the average reader will not understand it. The editor has argued that the table is done in other sports. I suggested going to Talk but they have instead continued editing pages in this way. Jkmaskell (talk) 20:28, 16 September 2015 (UTC)

Singling out World Cup results in this way is a bad idea. It makes the World Cup look like it's the most important event on the chess calendar, which is simply not true (or else Anand and Carlsen would agree to play). Cobblet (talk) 20:43, 16 September 2015 (UTC)
I don't like it too and I agree with you, it also may lead the readers to think it's the World Championship. Sophia91 (talk) 20:56, 16 September 2015 (UTC)
Agreed. Confusing too. It should be noted that there were earlier incarnations of the 'Chess World Cup'; the 1982 televised event, and the short lived series that began in 1988, run by the GMA. An old note on the WikiProject Chess Project Page under 'Articles to create' explains more. The current Chess World Cup article is inadequate in its coverage. Brittle heaven (talk) 23:17, 16 September 2015 (UTC)
This type of table is in wide use for golf and tennis players, including GA-rated articles. My only concern here would be the points made about the stature of the event. The World Cup is a pretty big deal, but not the biggest, and most other major chess events wouldn't really translate to this type of table. So, presenting this one tournament in this context does seem to raise it above all others. It's a nice idea but these tables are probably just not a good fit for chess bio articles. --SubSeven (talk) 23:53, 16 September 2015 (UTC)

I think the lack of documentation of the South American chess scene in general may be the issue here, rather than his inherent lack of notability. He was a damn good player - check out his games for proof, look at his coolness under fire against Panno for example. He was probably stronger than many European players who have articles, and apparently was renowned as a coach too. But chess literature, like wikipedia, tends to centre on the Anglosphere and Europe. MaxBrowne (talk) 07:52, 28 October 2015 (UTC)

Non chess articles for review

Chess gets short shrift, it seems to me, in many articles not principally about chess where it deserves a sub-section, or a mention. I thought the page Knockout tournament was one such, and I wanted to call attention to it here, but couldn't find the right place to do so.

I looked at all the sections on the Project page for where to put my request, but there wasn't one, so maybe we need a new area on the Project page for "articles that are principally about something else, but have a section that touches on chess, or has a section that ought to touch on chess, but doesn't, or needs improvement." Well, the section needs a shorter name, how 'bout, Non-chess articles needing improvement or some such.

In particular, Knockout tournament is a REDIRECT to Single-elimination tournament, which mentions chess once, in passing. In my view, there ought to be a section or sub-section there, with a brief overview on how chess tournaments work since the 1990s, along with a link to some chess page for more detail. That specific suggestion is perhaps a topic for a new section on Talk:Single-elimination tournament and I may or may not add it, if I have time, but the point here is, Where on the WikiProject Chess should one call attention to a situation like this so interested Project members are aware of it? I think there's isn't one, and that we need such a section.

If somebody creates a section for it, please ping me, or just move my request there. This would have the benefit of generating more inlinks to chess project articles from articles in sports, game theory, history, bios of individuals not known primarily for chess, computation and artificial intelligence, mathematical puzzles, athletic training, broadcast sports and television, social and club activities, and other areas; with a hopefully salutary effect on page views of chess articles. Mathglot (talk) 02:54, 30 October 2015 (UTC)

Started the 2015 in chess article

Hi all, I've started the 2015 in chess article. I'd appreciate any help with adding tournaments, referencing tournaments, and the like! Thanks, /wia /tlk /cntrb 21:12, 2 November 2015 (UTC)

Someone else please deal with the POV-pushing edit warrior

I'm already on 3R so I could get in trouble if I engage any further. Defining chess as a "sport" in the opening sentence of the main chess article is clear POV pushing. I've tried to explain on the talk page but he just keeps reverting. Defining chess as a "mind sport" as another editor wants isn't much better - this is not a familiar term to the average reader. The article was fine just how it was, defining it as a "board game" in the opening sentence and noting that the IOC recognizes it as a sport a bit later on. This fairly reflects the prominence of the "chess is a sport" argument.

You wouldn't go to the Shogi article and expect to read "Shogi is a Japanese sport" or even "Shogi is a Japanese mind sport" in the opening paragraph, but Shogi is highly organized and has serious competitions, just as chess does. MaxBrowne (talk) 22:43, 3 November 2015 (UTC)

I set things right again, for the moment, but I encourage other chess editors to keep an eye on this. The arguments in favor of changing the lede are utterly stupid, but unfortunately having to deal with stupid arguments is part of the joy of contributing to Wikipedia. Quale (talk) 00:12, 4 November 2015 (UTC)

I'm a bit on the fence with this one. I removed the PROD to generate a discussion but I'm not really sure there's much to say about the topic. Maybe a description and illustration of chess fonts, like the old fashioned ornate font used by British publishers like Bell until the 60s, the Batsford-style font from the 60s-90s, and the modern fonts? Can't find much material on this topic on line either. MaxBrowne (talk) 01:53, 1 December 2015 (UTC)

It can't be any worse than chess box. You could talk about chess fonts; but one could also dig deeper into history – I wonder what illustrations of chess positions in the old Arab or European manuscripts looked like. Cobblet (talk) 03:24, 1 December 2015 (UTC)
Chess box was listed as top importance - I changed it to bottom importance. Bubba73 You talkin' to me? 03:30, 1 December 2015 (UTC)
1497 chess diagram: http://dioscorides.ucm.es/proyecto_digitalizacion/index.php?doc=b23374548&y=2009&p=7 MaxBrowne (talk) 03:51, 1 December 2015 (UTC)
Diagrams are a function of Chess fonts, which is really a huge & fascinating topic. I'm guessing that unfortunately there aren't any published WP:RSs on same. (A hobby of mine, I've contacted British and German printing houses. The true forms of several beautiful fonts were lost when printing methodologies changed from hard-type to digital [in the 1980s!?]. The original engineering artworks weren't accessed by new-book printers/publishers for the digitizations, that's why after a certain point we see only approximated versions in all books & magazines [that aren't photo image-plate reprints].) IHTS (talk) 14:19, 1 December 2015 (UTC)

This is a new article created by User:Dino for the sequence 1.Nf3 d5 2.g3. The implication of the title is that there's a variation of the Réti Opening which is known as the King's Indian Attack. If one defines the Réti as simply the move 1.Nf3, then this is not unreasonable, but we still shouldn't have separate articles on the KIA and the "KIA variation of the Réti". It also remains common to reserve the term "Réti" for systems involving a quick c2-c4 hitting a pawn on d5, and the KIA for systems that do without this move (if an early central break is played it's usually e2-e4). Under this convention, the Réti and KIA are different openings – one is not a variation of the other. I thought this was already explained in our articles on the two terms as well as in Zukertort Opening. I suggest redirecting this article to King's Indian Attack. Cobblet (talk) 09:44, 19 December 2015 (UTC)

---

I too wondered of the similarity. But kept finding Internet references to the "Réti Opening, King's Indian Attack," plus a few books about it, as [1].

dino (talk) 22:43, 19 December 2015 (UTC)

That is the very picture of an unreliable source. Please seek advice from more experienced editors if this subject area isn't your strong suit. Cobblet (talk) 22:53, 19 December 2015 (UTC)

Yochanan Afek has apparently reported the death of Zvulon Gofshtein on Facebook. I'm looking for an official source for the information but can't find one. Maybe someone who can read Hebrew will have better luck? Cobblet (talk) 03:35, 26 December 2015 (UTC)

I tried googling

זבולון גופשטיין שחמט

("Zvulon Gofshein chess"). Got some hits on the Netanya Chess club etc but nothing about his death yet. MaxBrowne (talk) 10:51, 26 December 2015 (UTC)
http://www.chess.org.il/ContentPages/ContentPage.aspx?Id=1644 Here's a source. MaxBrowne (talk) 11:49, 27 December 2015 (UTC)

"US Chess" branding vs. "USCF"

Could use a third party to weigh in at Talk:United States Chess Federation#"US Chess". I'm at 3RR. IP seems intent to edit war over the matter and hasn't yet used the talk page. It's possible I'm in the wrong here, but I don't think a common acronym should be dispatched with universally because an organization decides it's changing its marketing strategy. — Rhododendrites talk \\ 05:31, 10 January 2016 (UTC)

Troitsky vs Troitzky

I've started a thread at Talk:Alexey Troitsky#Troitsky vs Troitzky in the article that project members may be interested in given the amount of attention that went into /Archive 31#Proposed changes to the Troitzky line article. --Marc Kupper|talk 09:22, 4 February 2016 (UTC)

This article looks totally redundant to me. It started out as a redirect, then became a DAB page, then somehow it became an article. It has no coherent theme and there is nothing in the article that isn't covered better elsewhere in wikipedia. MaxBrowne (talk) 23:25, 5 February 2016 (UTC)

Redirect to the list seems fine now. -Koppapa (talk) 16:15, 14 February 2016 (UTC)

World Chamionships

Created articles for the next two women's editions 2016 (knock-out), 2017. I won't move the 2016 match article for now, guess it's fine now. Suggestions to improve? -Koppapa (talk) 16:15, 14 February 2016 (UTC)

Thanks, Koppapa but when you create a new chess-related article remember to add it to Wikipedia:WikiProject Chess/Index of chess articles. Sophia91 (talk) 05:19, 20 February 2016 (UTC)

proposal about alekhine-capablanca rematch controversy

There is a widely debated topic on chess history, it is not clear whether it is based on facts or rumours/speculations. the topic is: "Alekhine-Capablanca world championship rematch." Some believe Alekhine avoided Capablanca and some believe contrary, ALekhine gave him a chance but Capa did not use it. can wikipedia enlighten all of us on this matter? p.s: I am not good at wikipedia, I cant add references and e.t.c. --Sir artur (talk) 19:53, 5 February 2016 (UTC)

Good sourcing is critical for issues such as this. Edward Winter is my go-to for chess history, but even better is to look at the original sources he refers to, e.g. contemporary newspapers and chess magazines. Some writers (e.g. Reuben Fine) are known to be unreliable on historical matters. MaxBrowne (talk) 23:29, 5 February 2016 (UTC)
good sourcing is why I asked this question here. if any of us knows something reliable about this issue, then why not to enlighten the rest as it is done on Staunton-morphy controversy ?! --Sir artur (talk) 12:13, 5 March 2016 (UTC)

"Incomplete" complete lists

Looking at the cleanup listings for the WikiProject and coming across the incomplete list section which as you would expect contains a lot of national championships. British Chess Championship is on the list but looking at the article I don't see any missing events. Given we can't add this years yet because it hasn't happened is there really a need for it to be tagged? Strictly speaking it should have a maintenance tag on there saying it is complete as of whenever, but then it'll appear in the "potentially dated" category. Jkmaskell (talk) 18:00, 6 April 2016 (UTC)

If all cells in a table are filled, the tag is no longer needed. That is, if all years are included of cause. -Koppapa (talk) 05:17, 12 April 2016 (UTC)

This draft is at AFC for review, we need some advice about its notability and whether the sourcing is good. Please post your comments to the draft's talk page. Roger (Dodger67) (talk) 15:42, 13 April 2016 (UTC)

Omega chess diagram

A discussion started in template talk page, because I'm testing a new version in order to fix some displaying problems of current diagram. Please, join. --Francois-Pier (talk) 09:19, 15 April 2016 (UTC)

Use of annotation symbols

The recent edit and reversion at Two Knights Defense has got me thinking. The annotation symbols "!?" and "?!" are inherently subjective and reflect the opinion of the annotator; as such is it ever appropriate to use them in a wikipedia article? Maybe Harding says 4...Nxe4 is "?!", but other writers might give it a "!?". There's no particular reason to prefer Harding over Estrin or Beliavsky or the numerous others who have written about the opening. Maybe we should take the Hübner approach and award only question marks to objectively bad moves, leaving the "!"s and "!?"s and "?!"s to writers who are under no obligation to be neutral? MaxBrowne (talk) 03:44, 12 April 2016 (UTC)

I'll have to take a look at that article – I know for a fact Harding has written about that line since (there's an early Kibitzer/ChessCafe column), and likely others have as well: there's no doubt 4...Nxe4 deserves "?!" at least.
You have to realize subjectivity is a big part of annotating chess moves: for instance, any evaluation of a position that's not "mate in X" necessarily contains some element of subjectivity. There are two ways in which one uses "!?" and "?!": one is "I think this move is good/bad but I haven't analyzed the position extensively enough to prove it," but with modern engines there's less excuse for this and we should try to minimize this usage. On the other hand, I don't think the other sense of "!?" and "?!" – moves that don't objectively change the evaluation of a position, but make your or your opponent's task (or both!) harder in a practical sense – should be censured. Cobblet (talk) 22:19, 12 April 2016 (UTC)
Some would argue that 4...Nxe4 deserves "?", some would argue for "!?", and any sign at all is a reflection of the writer's opinion, a shorthand for "dubious move", "bad move", "interesting move" just as if they'd written it in so many words. My point is that if writers do not all agree on the assessment of a move and it is not a clear blunder, wikipedia should not take a position on it by adding an annotation symbol. (4...Nxe4 was just an example, I'm talking about a broader issue here). MaxBrowne (talk) 22:50, 12 April 2016 (UTC)
I'd say it has to reflect the citation. If the quoted moves include symbols, then adding them is staying true to the cite, which ought to be the aim. Anything else is original research, and/or a distortion of the cited work. If an editor strongly disagrees with the assessment, he/she is free to add an alternative citation, with an accompanying note that states e.g. "GM Smith disagrees, and awards this move a !? because of ...". Of course this could get messy, so we'd probably take a vote here in the Wikiproject if it got out of hand, and reflect what the majority of sources say. Brittle heaven (talk) 23:22, 12 April 2016 (UTC)
If the writers don't agree on the assessment of a move, compare their analysis. Usually someone will be right and someone will be wrong, because one author saw something the other didn't. If it comes down to conflicting assessments of a complex position somewhere down the line, then I'd agree with you that we should usually not add a symbol; but I doubt this happens much anymore. Cobblet (talk) 23:50, 12 April 2016 (UTC)
If writers disagree, we are not in a position to rule on which writer is right and which is wrong. Engines might be able to answer the question but that is OR. The main source given (Harding 1977) is out of date and like just about anything else published back then it almost certainly contains errors that an engine would expose. Yet this appears to be our primary source for a subjective annotation symbol. To reiterate - attaching an annotation symbol to a move is equivalent to writing the comment "this move is dubious", "this move is interesting" (or whatever !? is supposed to mean), "this move is bad" etc just as if it was written in so many words, and in wikipedia's voice. MaxBrowne (talk) 01:27, 13 April 2016 (UTC)
Not sure I understand how "in WP's voice". Is it a question of markup context/confusion where the annotation symbols aren't attributed clearly enough to the analysis source? 4...Nxe4?! (and the other symbols) are from Harding/Botterill's pen of course. (The book defines in Symbols after the Preface "!? Interesting move" and "?! Dubious move", equivalent to ECO's "!? a move deserving attention" and "?! a dubious move", and WP's [short] defs "!? Interesting move" and "?! Dubious move". [Korn MCO 12 says "!? Good move—but open to further research" and "?! Speculative attempt to complicate".]) IHTS (talk) 04:40, 13 April 2016 (UTC)

I added the 4...Nxe4?! material more than 3 yrs ago. Since it was difficult see who was quoting whom, I made some adjustments (added sources for Gligoric, Korn; everything not from Harding/Botterill is sourced individually now, perhaps that improves). Probably the 5.Nxf7? line can be eliminated as too detailed (overkill). (Cobblet, maybe the entire 4...Nxe4 line is just a footnote/overkill, too!? If so feel free to delete.) Almost all the Harding/Botterill material after 5.Bxf7+ is also found in ECO C2 (1981, Gligoric), and that material has been carried over without change, to ECO C3 (1997, Matulovic). Of course anyone w/ more recent analysis should go ahead and improve the article. (Ya gotta start somewhere, right!? [Or the line can be deleted, as above.]) Cheers, IHTS (talk) 07:36, 13 April 2016 (UTC)

Addendum: Oh! I see the point re "WP's voice", I think: 4...Nxe4?!, unattributed, was in the TOC (which might convey WP's voice). Instead of adding attribution to the TOC, I removed the symbol, so now only the article body has it. (If still unclear the symbol is attributed to Harding/Botterill, I can copyedit some more, just let know.) BTW, Gligoric also gives 4...Nxe4?!, and as mentioned Matulovic copies that, but Korn gives 4...Nxe4?) IHTS (talk) 07:57, 13 April 2016 (UTC)

The later Harding article I'm thinking of is this. Harding highlights the importance of 6...h6 in the 5.Bxf7+ line – note his assessment of 8.dxe5 compared to 8.d5. I'd certainly support a further pruning of the lines – it would render moot a couple of quibbles I'd have (citation needed for the line beginning 5.Nxf7? Qh4! 6.g3 Qh3; and are all the attributions to Gligoric/ECO following Lopukhin's 7.Nc3 correct, or was ECO quoting Lopukhin's analysis without attribution?). Cobblet (talk) 13:39, 13 April 2016 (UTC)
That's weird Harding says he didn't come across 4...Nxe4 until 1993 but his book is copyright 1977. He says ECO gives 6...h6?, but neither my ECO C *2* (1981) nor *3* (1997) show that.

Anyway I've trimmed out the 8.d5 line per Maric's 9...Kg8! discovery. And I've re-placed couple supporting refs for better clarity. (My only src for 5.Nxf7? line is Harding/Botterill book. Everything after 7.Nc3! is contained in ECO C *2* and *3*, except the attribution of 7.Nc3! to Lopukhin, the attribution of 10.Bg8! to Estrin, and the subvar 8...e4 9.f3!, and those three things I'd found only in Harding/Botterill. [How much of that analysis is Lopukhin's I don't know; re attribution what I did was draw on Gligoric's assignment re C57 in ECO C *2*, p. 247.]) IHTS (talk) 09:08, 14 April 2016 (UTC)

Update: Turns out I have Estrin's 1971 English ed., and essentially all the Harding/Botterill and Gligorich/ECO lines replicate that; I've updated the article refs accordingly. IHTS (talk) 04:48, 19 April 2016 (UTC)