X-rays, skewers, dogs and cats

[munich]

Can an X-ray attack ever go through a pawn or a bishop along a rank or a file, or a rook diagonally?

Well, I guess yes (I am not sure if I understood the question correctly): Just look at the two examples provided by Alvarofrota. 2 cases where the x-ray goes through a pawn.

Can an X-ray attack ever go through a king?

No, not if you mean the X-ray by CT. Because to go through the king and take the piece behind the king means that you the king would be in check. The king cant be in check unless the position is an illegal position.
In the previous move, the king would have needed to go out of check. If he didnt, but moved s.th. else, then he leaves his king in check. Only then can the king be in check and instead of taking the king you would take the piece behind the king (which is of course much weaker than taking the king). You can only take a king if you are in a blitz game.
In OTB, if the opponent overlooked that he is in check he needs to take his move back and take care about the check (but the "moved-touched" rule would still apply).
So there cant be an example other than derived from an illegal position or a blitz game where you overlooked that you can take the king.
Example (white to move): if white wins the black queen with the x-ray attack QxQh8+ then white just missed that the king is already in check.
And how could such a position legally happen? Blacks last move must have ignored the check.

( ) vs ( )

Date:

Event:

Site:

Round:

Result:

[SetUp "1"] [FEN "7q/6k1/8/8/8/2B5/7Q/K7 w - - 0 1"]  *

However, if you give the king a check, he moves to the side, and then you take a piece behind him --> then the skewer  tag applies. (see CT definitions). So if it is black's move in the above position, then it is a skewer.

Or more generally, can equal valued pieces ever be pinned to each other?

Yes, this is possible (bishop pinned by a bishop, white to move):

( ) vs ( )

Date:

Event:

Site:

Round:

Result:

[SetUp "1"] [FEN "7k/6b1/8/8/8/2B5/3Q4/K7 w - - 0 1"]  *

[interlist]

Munich - you changed your answer!

[Update]

Munich, I believe you originally answered yes, that an X-ray Attack can go through a bishop along a rank or file, and through a rook along a diagonal.  The question is exact and as plain as possible in its wording, if you want I can diagram it, but that really shouldn't be necessary.

If it is confusing, it is because it of the "extra" conditions attached, but these aren't my doings!

I will reread your and Alvaro's answers wrt (with respect to) X-ray Attacks through the king, and will post an interesting example from a blitz problem I did today.  At least, I hope you will find it interesting and, I think, illustrative.

I am mulling over my next post - we might be reaching the half-way point in the discussion - ha!  Let me take a few moments to make sure I haven't missed a question or two before I commit to writing this next installment.  I definitely am interested in Alvaro's opinion on pins for right now...



 

[alvarofrota]

Rio de Janeiro, 14 de setembro de 2012.

Prezado interlist:

There is not possible an X-ray attack of a Rook through a Bishop or of a Bishop through a Rook as the line of action of this two pieces are always different.

In the same way, is impossible an X-ray attack through a Knight as a Knight has not a line of action.

And, finally, respectfully disagreeing of my friend munich, of course is possible an X-ray action through a King, as a King do has a line of action, despite a restrict one.

An example of this is problem number 75452:

( ) vs ( )

Date:

Event:

Site:

Round:

Result:

[Event "Problem nº 75452"] [Site "www.chesstempo.com"] [White "White plays"] [Black "And wins"] [WhiteElo "-"] [BlackElo "-"] [Result "1-0"] [FEN "1k5r/RP3pp1/4p2p/3q2b1/1P1r1n2/7P/1PQ2P2/4R1K1 w - - 0 1"] { A very beautiful  mate in six moves: } 1.Ra8+  $3 { ... "probably the 'worst' first move ever played", according davaniel } Kxb7 2.Ra7+ $1 {An *attraction/distraction sacrifice* of Rook} Kxa7 3.Qc7+ Ka6 4.Ra1+ Kb5 5.Ra5+ Kxb4  { And last, but not the least: } 6.Qc3# $3 { the White Queen checkmates Black King *X-ray defending* his Rook!. } 1-0

Another example should be saw in my first post in this tread:

( ) vs ( )

Date:

Event:

Site:

Round:

Result:

[Event "Example nº 3"] [Site "www.chesstempo.com"] [White "White plays"] [Black "And wins"] [WhiteElo "-"] [BlackElo "-"] [Result "1-0"] [FEN "8/8/6R1/5k2/Q7/3K4/8/8 w - - 0 0"] {White delivery a checkmate} 1. Qe4# {using an **X-ray defense** through the Black King.} 1-0

There are a number of other problems with this same tactical motif.

Aquele abraço!

Álvaro Frota

[interlist]

Greetings Alvaro!

And for the pin question... what say you?

--interlist

[interlist]

Hi Álvaro,

Yes, I'm sorry, you did indeed show that "X-ray of a king with queen" example - that and the other one both utilize a favorite pattern of yours (and with a little more practice, a favorite of mine as well!).  You must forgive me, I had a different example in mind and should have noted your previous example.  But heck, we got another great example as a result - muchos gracias!

What I'm after is precision - which is implicit in the second part of my question "can you define the criterion"?  

Basically, I believe your examples suggest the king can be "X-ray Attacked" through, but only if the king is itself attacking the target square.  I think your examples only show "X-ray Defenses", if I'm not mistaken.  We shall, for the moment, keep the skewering of a king in a separate category.

Am I correct then, that the king can only be "X-ray attacked" on a square adjacent to the king, according to your definition of course?

Aquele abraço!

--interlist

[alvarofrota]

Rio de Janeiro, 15 de setembro de 2012.

Prezado interlist:

In my opinion, a pin of an undefended Knight against an undefended Knight is the very same thing as a skewer of an undefended Knight against an undefended Knight.

To be truly through a piece, an X-ray attack (or defense, whatever) must be on a square where this piece is defending (or attacking). This implies that your piece and the adversary piece must have a common line of action.

A King has four lines of action, two stretches of diagonal, a stretch of a file and a stretch of a rank, all of them with only three squares.  

A Pawn that is not on the border of the board has two lines of action, the two stretches of diagonal in front of him with only two squares. An X-ray attacker (a Queen or a Bishop) must be placed behind the Paw, in order to X-ray attack the square of the common line of action.

A Pawn on the border has only one line of action but it is impossible to X-ray attack this kind of Paw, as the X-ray attacker (Queen or Bishop) must be placed out of the board.

Finaly, lets see problem number 58458, a very typical example of an X-ray attack that leads to a back rank mate:

( ) vs ( )

Date:

Event:

Site:

Round:

Result:

[Event "Problem nº 58458"] [Site "www.chesstempo.com"] [White "White plays"] [Black "And win"] [WhiteElo "-"] [BlackElo "-"] [Result "1-0"] [FEN "3r2k1/pp3pp1/1q2p2p/7P/5P2/3R4/PPP1Q1P1/1K6 b - - 1 1"] { Black delivers the first check } 1...Qg1+  { and White defends the back rank with his Queen } 2.Qd1  { However, the Black Rook *X-ray attacks* the White Queen through the White Rook, allowing  } Qxd1+ 3.Rxd1 Rxd1#  { a *back rank mate*. } 1-0

Here it is interesting to note another example of an X-ray attack throug a King. Indeed, in the final position, the Black Rook X-ray attacks the "a1" square through the White King and, as the White King can not move to this X-ray attacked square, he is checkmated.

However, I strongly do not recommend tagging this kind of tactical motif as *X-ray attack* or we will mess up our Tag System.

Aquele abraço!

Álvaro Frota

[interlist]

In my opinion, a pin of an undefended Knight against an undefended Knight is the very same thing as a skewer of an undefended Knight against an undefended Knight.

Thanks Alvaro, for the reply.  I appreciate you taking the time to write out your definition so well. As you know I use a different definition, and will contrast our definitions in a future post.  I prefer to have your exact definitions rather than relying on some paraphrase.

The pin issue is involved in understanding each system's definitions.  Therefore, if you don't mind, for further clarification I'd like your opinion on the following position:

( ) vs ( )

Date:

Event:

Site:

Round:

Result:

[Event "Pin/Skewer/X-ray question"] [Site "Chess Tempo"] [Date "2012.09.15"] [Round "-"] [White "White"] [Black "Black"] [Result "*"] [EventDate "2012.09.15"] [FEN "1n4k1/1p6/8/6p1/5n2/2N3B1/1P2P3/7K w - - 0 1"]     1.e3 Nd3 2.Bxb8 Nxb2 3.Kg2 *

The Black knight at b1 is at-risk, protected from the X-ray attack (or pressure) of the bishop by the screening (shielding/blocking/etc) knight on f4. This is not a skewer in the normal sense - firstly the two pieces are of equal value, and secondly, the knight on f4 need not move (the built-in exploit inherent in most definitions of a skewer). Obviously, if Black were to move the play would be 1...Nc6 and the at-risk knight would flee to safety.

But with White to play, the simple pawn move, e3 subjects Black to the loss of material.  So, two questions-

   1. In the original position, is the f4-knight pinned?

   2. What are the tags you would use for this problem?

Many thanks,

--interlist  

PS- Although addressed to Alvaro, I would welcome comments from others as well.

[interlist]

I asked about pins in the previous post, and since nobody has volunteered an answer I suppose the onus is upon me to continue the thread.  I asked about the position because it seems to demonstrate a small conundrum with the working definitions of tags vs. the official canonical definitions.

Let's start with position in question from the previous post.  One might suggest "Pin" as a tag, since the knight on f4 cannot move without exposing the knight on b8 to capture from the bishop (conversion of X-ray attack into direct attack). This seems reasonable, except when checked against the canonical definition on the CT Tactical Motifs manual  (CTTM):

Pin
A pin occurs when an attacked piece cannot move without exposing a more valuable piece or square behind it to attack.

Here the two pieces are of equal value, so it is not a pin according to this strict definition.  Perhaps then, the tag might be "Distraction":

Distraction
Distraction (sometimes called deflection) involves forcing the opponent to move a piece that was previously guarding important squares or pieces.

One might argue that the f4-knight is "guarding" the b8-knight.  But this is somewhat unnaturally overloading the meaning of word "guarding", and has no illustration in the CTTM. Consider the same position with the bishop removed, nobody would say the knights are guarding each other.  It is natural to say that the bishop is X-ray attacking the b8-knight, and directly attacking the f4-knight however.  Moreover, the f4-knight need not move, as Black can choose which knight to lose. This also argues against the use of the tag "Distraction".

The latter point can also be made against the argument that the "Clearance" tag (2nd meaning) should be used:

Clearance
[...] The second form of clearance occurs when one player forces a piece away from a diagonal, rank or file (often using a sacrifice) to make way for another piece to utilise or attack the cleared path. [...]

This might be a good tag, except that Black is not forced to move the f4-knight.  What tag would apply in that case?

I'll dismiss the use of the tag "Skewer" out of hand (but will mention that it also has a similar condition as "Pin", i.e. that the two pieces must be of unequal value).  The rest of the tags can be quickly dismissed as well, save for this one:

X-Ray Attack
An X-Ray attack occurs when one piece attacks a square or piece through another piece. Note that this is not the same as a skewer as the relative value of the piece being attacked through is irrelevant.

The original position involves an X-ray attack, as the term (not the tag) is commonly used.  The potential of the attack is realized by a very simple exploit, a pawn advance that attacks the knight which is screening the second piece. One or the other knight must fall - so it qualifies as a tactic.  

What other tag(s) could be used?  

[interlist]

Since the cat is out of bag and I've pinned myself down in the quest for historical accuracy, I've worked like a dog digging various references of the terms discussed in this thread.  But enough punning around else I'll risk being skewered...

Today's post concerns the topic of "The Pin".  The following information can be found in Edward Winter's book "Chess Facts and Fables"

Chess Facts And Fables    (4.3/5 stars - 3 reviewers) Edward Winter McFarland & Company (November 21, 2005) 0786423102

The Pin

Sub-standard definitions of the term "pin" are to be found in many chess books. For instance, the glossary on page 180 of "Chess The Easy Way" by Reuben Fine (Philadelphia, 1942) offered a one-sentence explanation: "A pin occurs when a man screens a unit of higher value".

From page 200 of "The Everything Chess Basics Book" by the US Chess Federation and Peter Kurzdorfer (Avon, 2003) comes a description which is also everything but helpful:

Pin. This is a weapon that requires two enemy pieces on the same line with a friendly long-range piece. Instead of two good guys and one bad guy on the line, as in a discovered attack, we have one good guy holding two bad guys hostage. Well, only one of them is actually held hostage, but they both have to be there. The pin is more akin to a wrestling pin than to a sewing pin. In it, one friendly long range piece looks at a powerful enemy piece with a less powerful enemy piece shielding it.

Reference books shun the question of when "pin" became part of chess terminology. On page xii of his "Introduction to Chess Studies" (London, 1844) George Walker wrote:

"Of course I consider that all players for whom I have made up these Chess Studies are acquainted with the ordinary chess terms, as bishop 'pins' knight, and similar conventional phrases."

How much further back can "pin" be traced in chess literature?

(CN - 3298)

Winter also has information about the first occurrence/origins of various chess terms here:

http://www.chesshistory.com/winter/extra/earliest.html

where we find the following information:

Pin

    OED: ‘Removing his queen to obviate the “pin”.’ The Book of Chess by G.H. Selkirk (London, 1868), page 72.

    ‘If he move Q. to K. fourth, you answer with Q.B. to B. fourth, and then pin Kt., if he return with Q. to Q.B. sixth.’ A New Treatise on Chess by G. Walker (London, 1833), page 88.

Now, we all agree that pin is a venerable term, which has been part of the chess vernacular for well over a hundred years, if not far longer.  But a reexamine of the use of this term is necessitated in order to disentangle the various uses, some non-standard, of the term X-ray.

 Regards,

--interlist

‘In chess the “pin” is mightier than the sword!’ - attribution unknown (See CN 3599)

[interlist]

=====================================
The Winter of our discontent - a sidenote
=====================================

Well, I admit that's stretching a little to make a pun.  But I've relied quite heavily on Edward Winter's writings, especially in discussing the historical origins of various chess terms.  This is not without cause:

In the English magazine Chess Winter is always referred to as The Omniscient.
[...]
Winter is a just but stern supervisor of chess literature. Every chess writer in the
English language knows: when he makes a mistake in a date, overlooks a mate
in an analysis, or sins against the King’s English, he will be flogged by Winter,
whose eyes see everything.

As found in "The Hermit of Geneva", an article written by Hans Ree for the Skittles Room feature, available as a pdf at chesscafe.com.

An interesting and informative thread entitled Edward Winter does not exist is available at chess.com.  

Another well know procurer of chess information, Bill Wall, has this to say:

billwall
GM Keene thinks he knows who he is.   I don't think Edward Winter is a chess master and probably has not played in a chess tournament in many, many years.  He does make mistakes and typos as well (I have found several errors in his books), but does pounce on anyone else who does (I know.  He has pounced on me for years).  He just doesn't do it in a friendly way and he rarely says anything positive about anyone.  I have also read that Winter, or his works, are researched and written by more than one person.  He must have one of the best chess libraries or access to the major chess libraries in Europe.

I'll close by quoting from a comment Estragon made in that thread:

No matter who he is, he has performed a great service for us and for chess in authenticating its history and debunking its myths.

OK, that wraps up this little sidenote.

[interlist]

====================================
A Positional Puzzle, by way of Introduction
====================================

Now, I could add a word or two as an introduction, on this introduction to a word or two. But that might give the puzzle away (and the palindrome, of a sorts).  I will just say that it is White to move, and the challenge is to find the "best" move in the following position:



Now "best" is determined not by an engine (see below), but rather by how Capablanca, as White, moved.  And there is probably little need to sing the praises of Capablanca amongst this crowd, eh?

As a happy, or perhaps sad, confluence of themes, this position might also be worthy of inclusion in the STS suite of positions used to test engines.  In fact, consider the following comparison of Stockfish 2.2.2 JA vs. Fruit 2.3.1 in evaluations of the 3 leading candidate moves, both engines run to 16-ply:

#	Stockfish	Fruit
1	+1.29	+0.24
2	+0.64	+0.23
3	+0.00	+0.23

I'll let you guess which engine agrees with Capablanca's move choice.

Unlike asterion's challenge, this is a very unbalanced, dynamic position.  A tad more interesting, eh?

Bonus Credit to those who can guess the word or two associated with the puzzle!

[munich]

I wanted to say, that I agree with alvarofrota about me being wrong. An X-ray can happen through a king (his 2 examples of x-ray checkmating the king are very good).

For the anecdotal part of who defined a x-ray this way and who an skewer that way and somebody a pin his own way might be of interest for a historian.
But from a practical point of view, we want to have tags to tag puzzles in order to filter for them later.
Why would somebody like to filter for x-rays?
Well, because he wants to train the patterns of x-rays. What he certainly is not interested in is to know who said first what an x-ray is, when he filters for typical check mate patterns like alvarofrota presented them.

Chesstempos description might differ from historical quotes, but this is of no interest anyway (not when using the tags to filter for puzzles). That somebody said s.th. first in a different way - so what? These historic quotes were not tailored for chesstempo tagging.
Y. Seirawan said in his book, that at the end almost all tactics are somehow double attacks in one way or the other. But of course, just because some GM Seirawan said this, this is not going to be very useful in our CT tagging, since we want to tag puzzle by their idea. The more we can isolate a pattern the better the tag. "umbrella" tags are not good filters.

So, interlist, what is your intention then? you want to tag all skewers into x-rays? Even though you are not even using the tagging system to filter for puzzles?
A tag sorted set is a very powerful training. Of all over 30++ year olds who were already A-class players, I am the only one I know of who improved his rating considerably (Blitz and real rating of course! Not Standard nonsense).
It can be done by breaking down the complexity of chess. Going from the simple to the complex, breaking down the sheer amount of tactical ideas into set that contain more often the same ideas: by filtering for tags.
You wont get better in chess if you look for historical evidence who the first guy was who said what an x-ray is.
But you will get better in chess if you filter for puzzles. I recommend: do some skewer puzzles, do some x-ray puzzles, and some pin puzzles. But dont mix them. Stay for a while with one tag before learning some new patterns.
Then you will improve. And you will find out, how frustrating it is, to make people understand what these tags are for, and that it is not useful to tag according to historic quotes. I guess I made my point thoroughly clear now.
  
Edit:* having said these harsh words, I need to admit, that I think the historical quotes are interesting on their own. I am only worried, that these findings might lure people into tagging puzzles in a way that is not useful for training purposes.
So nothing wrong with these anectdotal findings, etc, as long as the tag-training isnt influenced by these findings.

[interlist]

munich, please have some patience, i hope to explain and answer your questions in the next few posts. I don't find your words particularly harsh either - you have your concerns which I've acknowledged as very legitimate and valid.  And your post asked a couple of reasonable questions, in your inimitable style of course.

I'm glad you did ultimately find some of the posts interesting - i've put a lot of work into collecting this information and presenting it. I also find it interesting, wanted to share, plus it keeps with the timbre of my previous posts.  I hope to switch to a more practical bent (and less historical) soon. 

 

[interlist]

Following up on the Capablanca position recently posted, and blending in the two-knight X-ray tagging question...

Here is a position from a composition prize winner, with three examples of the motif in question:

( ) vs ( )

Date:

Event:

Site:

Round:

Result:

[FEN "B3R3/3N1N2/2n1r3/2pnr3/Rbb1k1P1/2P1ppK1/2P2P2/4Q3 w - - 0 1"]

White to move, can you find the M-2?

Also, among various other attacks implicit or explicit in the position, note that the White bishop at a8 is X-ray attacking the king according to the current writer's nomenclature.  In fact, Black's king is much more vulnerable than initially it appears, four of White's pieces are X-ray attacking it. And every one of White's pieces will have a role in the M-2.

The theme present in this problem is, according to B.P. Barnes in Advanced Chess Problems and How to Solve Them, "one of the most fruitful in the history of problem chess".  It is also a sub-variant of the X-ray attack family.

[interlist]

This seems like a good point in the discussion to again review the concepts and definitions of what I refer to as the "Family of X-ray Attacks/Defenses", or X-ray attack for short.  Of course, utilizing Harkness' definition, and generalizing to include both piece colors, leads to the large family pictured above.  

I've prepared the graphic in order to clearly present the various definitions and their inter-relationships.

(Aside- There are still some issues which can be mentioned immediately. First, pins can be generalized and abstracted to mean "a piece tied down and prevented from moving by some defensive role".  If that is the case, the set of pins would extend beyond the X-ray family, and Venn diagrams would be appropriate. Next, I've introduced a few new terms such as half-pin and battery, that are utilized in the literature.  There is some overlap between the terms yet, but our discussion need not be concerned with this, as it needs not be so exacting at the moment.)

The main purpose of the discussion is to alert the reader to the existence of this family, and to introduce the new term "Collinear X-ray Attack/Defense".  The latter could also be referred to as "Alvaro X-ray Attack" or "Arrow X-ray Attack", AXA.  It essentially matches the definition he has outlined in previous posts.  His definition is unique in the literature and too restrictive to be generically referred to as "X-ray Attack".  It is similar to that used in the chessprogramming wikipedia pages, but more general to include "attacks" through pawns. As such, I believe it is the better definition of an important subclass of the family. It is a coherent and useful definition, and we will come back to this in a moment.

Now a simple working definition -- tactics are the exploitation of weaknesses in the enemy position through combinations in order to gain material.  The fundamental currency of a tactic is a combination - i.e. the collection of variations allowing the gain.  We like to characterize the combination with tags, or some name offering insight into the weakness exploited. Generally we refer to them as motifs.

If one looks at the X-ray family it is apparent that many of the categories are motifs that are self-exploiting, or rather, whose exploitation is obvious.  A skewer or a discovered attack has the exploit built in -- typically the very next move is the capture of a piece for gain.  A discovered check suggests an opportunity associated with the move delivering check, etc.

Some of the motifs in the X-ray family are among the most ancient definitions in chess, for instance, Pin. It existed long before X-rays were even discovered, let alone incorporated into chess terminology.  This motif doesn't have an exploit that is immediately apparent.  But it is a theme most every player learns to exploit almost as soon as they learn the moves to the game. We all know pins and how to exploit them. And we also refer to them, without an exploit to turn them into a tactic.

And this is true of the generic "X-ray attack".  It is a general term, and we shave off the various categories illustrated above, in order to suggest to how to exploit them.  It is better to refer to the specific sub-classes, because we better understand how to exploit them.  I refer to this as the "Law of Specificity".  It is a subtractive process, starting from an somewhat abstract, general class of positions.  And indeed it is a large family as the graphic shows (did I leave any terms out?).

As we subtract off the simpler, more familiar categories, we enter more advanced territory.   Subtract off pins, skewers, discovered check/attacks and what remains are typically harder to exploit, and maybe seem more mysterious. But they are some of the more enjoyable tactical challenges.  

Still, it might be nice to further categorize and select out additional subclasses. Some authors have done so, and talk about batteries and half-pins.  Perhaps in another post I will talk more about these.  Here, I want to introduce a definition of a new subclass in the X-ray family - Collinear X-ray:

A "Collinear X-ray Attack/Defense" involves a sliding attacking piece (Q,R,B) whose action is screened by a defending piece such that the defending pieces action overlaps with the attacking piece's action.  Thus, the collinear label.

This category or subclass that seems to be a CT favorite, and it's strongest advocate seems to be Alvaro. It is a slightly involved definition, but I admire Alvaro's efforts to bring it to people's attention, and educate them in how to exploit it tactically.  It is a coherent, powerful and practical definition which selects a category of X-ray attacks whose exploitation is inherent in its definition.  This is why it is so appealing (despite its complicated definition).  There are many great problems involving  AXA's.
 
But It does not encompass all the positions where the X-ray attacks can be exploited. Nor does it replace the generic, and rather easy to understand, definition of what an X-ray attack is (the definition given in the CT Tactical Motifs btw). Presently, most problems tagged with "X-ray Attack" on CT belong to this X-ray subclass. Subtract out this category, and the remaining part of the family I label "Terra Incognita" in the diagram above.  The exploitation of these positions is typically harder still, just as going from pins and skewers to AXA's was.

[more later...]