The Chess Variant Pages

Short explanations to Pocket pieces TT

Thanks to Jim Aikin (JA), PBA and Fergus (F) for giving early comments that after some explanation can help others to understand the language and ideas on main Pocket Pieces TT announcement page.

I should first point that my English is far from perfect, nevertheless I hope my explanations would be understandable.

PBA: "The more general problem with the examples is that they are written in Chess-problemese, not English as they appear. For example, #2 is shorthand for mate in 2, white to move."

F: "I can't make heads or tails what these problems are about. It is all Greek to me. How about a primer on what the problemist codes actually mean?"

It is true that that problemists have its own jargon and abbreviations with precisely defined meanings that may be hard to understand. Good explanation of standard abbreviations like h#, #2 and so on may be found in webzine Problemesis by Christian Poisson (choose the appropriate link in lower window):

I recently came across very good series of articles explaining various terms from chess composition by renowned master C. G. S. Narayanan, it may be of help:

I am very sorry, but the tourney itself is aimed not only to standard CV pages readers, but also at normal fairy chess composers - Hans planned to send the announcement into Infoblatt, bulletin of announced tourneys around the world. The composers need very little further explanation and I, beeing the composer for a long time, cannot easily find the right level of explaining the things to people form outside of composition. I was prepared to answer the questions and stil I am.

JA: " Maybe I'm missing something, but the example problems given make very little sense to me. In Problem 1, is "S" being used as an abbreviation for "bishop"? ..."

Well, it is standard of official magazine of British Problem Chess Society, The Problemist, to abbreviate normal chess pieces king, queen, rook, bishop, knight, pawn as K, Q, R, B, S, P respectively. It is not standard worldwide, naturally. Not even English-speaking-countries-wide :-) as new American chess problem magazine StrateGems uses N for knight instead of S. In spite of all difficulties with signs for pieces, I hoped it will be clear from notation that abbreviations are as given above. It is also the standard of Chess Composition Microweb (with exception of pages dedicated to originals of Pat a Mat, Slovak chess composition magazine, where Slovak notation KDVSJP is used).

PBA: "First, I believe S is a Knight, since the moves make sense. Isn't the Knight called a Springer in some languages?"

Indeed, it is. In German and e.g. in Swedish it is very similar (if not the same), Swedish magazine for chess composition is called Springaren.

JA: "How can 'the defending pocket S' counter any threat at all, until it's placed on the board?"

I guess this question touches the comment after key and threats of #2. Pocket piece can defend the threat only by placing it on the board, it seems to be true in given circumstances. The comment was aimed to show other important thing - that after any defence in solution of #2 there must be mate available for white. And if black has too many defences, it could be too difficult to find reasons that would lead to mate - or as problemists say, to find harmful motivation of black defences. Example: Before key in #2 white cannot mate in 1 by 1.pSe6+? as black has at his disposition 1...Qxe6! Only after 1.Sxc5! and black defence 1...Qd3 black withdraws guard of e6 (this is called black error or harmful motive of black defence) and white can mate by 2.pSe6#

So far I was speaking only about harmful motivation. But every black defence here (and generally in all #2 with threats) has two "sides" - defence motivation and harmful motivation. Defence motivation is strong part of the defence from the Black's point of view. It is the reason why the move defends the threat. Here 1...Qd3 defends because after it white cannot mate 2.Sxb3+? nor 2.Se6+? because of 2...Ke4! It turns out that 1...Qd3 gives bK prospective flight e4, unblocks e4 as problemists say.

After little thinking it is clear that any move of bQ carries this defence motive - unblocking - but here are listed only 5 of them. Why? Because other are weaker, they allow more than 1 white mating move and as such are ignored, they add nothing new to the problem. Multiple mates after defence are called duals and generally they are considered to be bad, lowering the value of the problem, they should be avoided if possible. But if it's impossible and they don't disturb the idea, they may be accepted as the judge of The Problemist competition did with this problem.

(Btw, it is very useful to study in deep motivation of all defences to find out how the problem works. E.g. 1...pSb5 defends by closing Qa6-d3, giving prospective flight d3 after any of threat checks, but th error is deploying the pocket and white can mate 2.Qa4# without danger of pocket interference on b4 or c4.)

I return to the reasoning of the comment. If white had only one threat, say 2.Sxb3#, black would have many more defences by pocket piece simply threating b3 and it would be difficult to find suitable mating continuation on all of these. So the comment (by author Ronald Turnbull) may be considered to be the advice of a man experienced with the pocket pieces to the authors of problems for this tourney. It is not axiom - only advice.

PBA: "Problem two is more mysterious. I've done some delving on the web, and I some idea. White starts by undoing a previous legal move (this is a form of retrograde analysis problem), then it's a helpmate I think in 5 (the 1, I think applies to the number of retracted moves). A help mate is where both black and white are cooperating to mate white."

Not exactly, but close. The stipulation in long is: "White retracts one halfmove to get the position where he can checkmate Black with his help in 3 halfmoves. " It is not usual to formulate the stipulation in terms of halfmoves (sometimes called single moves), but here I tried to be precise in view of the fact that CV readers aren't so familiar with numbering of moves in problems.

In direct problems like "mate in 2" everything is as usually. White starts and mates in 2 moves, whichever is the black defence. But in plain helpmates, say in helpmates in 2 moves (abbreviated h#2), there are in fact 4 halfmoves - Black and White cooperate to checkmate Black and moves are - black, white, black, white. Also the notation is reversed - see e.g. page with some helpmates - trace the manoeuvres of sides to grasp the idea of helpmate:

(Btw, helpmates are nowadays very popular, if you didn't see any yet, try to solve some of them, you can find tons of them at CCM, there is the chance they will catch you :-))

But! Here the stipulation contains h#1,5 - it simply indicates, that 1st black move is missing, it means white moves, black moves and white mates after retraction.

JA: "In Problem 2, what does "retract" mean? The white king isn't on a7, so how can its move to a7 be retracted? If the notation "(+R)" means a pocket rook is being deployed, where is it being deployed? And when? The text includes the line "2.0-0-0," but that's a move by white, and white's king is not in position to castle. When did black deploy his pR?"

This comment points that the explanation of the solution was too short. Sorry. I hope the following will help.

a) White retracts (= takes back) move Ka7xRa8.

(In shortened notation it is indicated by moving wK to a7 and adding bR to a8 (+Ra8). Deploying of pocket piece is unified in notation by indicating "p" before piece abbrevation.)

Now there are two possibilities. White is in check, last move was chek by black rook to a8. But it could have arrived from d8, c8, b8 (1st possibility) or the last move was pRa8+ (2nd possibility). Only in 2nd case bRa8 can castle. Both White and Black know that and the cooperate to checkmate Black in 3 halfmoves (or as problemist abbreviate that, to h#1,5). White plays Kb6 and now... Black castles! This proves that bRa8 was pocket one, this in turn shows Black had already deployed his pocket and thus white can mate by pRa8# as black cannot interfere on b8.

b) The logic is similar. Retracting Kb7xRa8 sets Black into position where are two principal possibilities - 1st - his last move was with R or K, 2nd - his last move was pRa8. Later castling proves 1st impossible => 2nd is sure and the rest is the same.

b) Here there are no castling tricks.

If anything remains unclear, don't hesitate to ask.

JA: "After thinking about it some more, I'm not sure a "pocket piece" problem would differ in any major way from an ordinary chess problem. I mean, the main difference between dropping a piece onto the board and moving it from one square to another is that in the former case you can't capture an enemy piece. Only in very constricted board conditions might you encounter a situation where a piece could be dropped onto a square that it couldn't reach by being moved from another square -- and in this case, the solution to the problem is likely to be obvious."

It seems not to be true, yet it is interesting view.

I see the interest of the view in the fact that is valid generally in fairy chess. Fairy chess problems shouldn't show the ideas that may be well showed in usual (orthodox) chess problems. They should show something new, unusual, impossible, ... and it is right to point out that if there were only usual ideas possible in Pocket pieces variants, it would be nonsense to announce the tourney for this kind of problems.

Fortunately, in my opinion, first example problem is the clear counterexample for this. White pocket knight mates from 8 squares around bK - it is clearly impossible in standard chess problem. The point is not in the fact that these 8 squares are in any way unreachable for knight, usual knight could enter these from many directions. The idea is in the fact that pocket knight can be on many places in the same time - much like Kwisatz Haderach in Herbert's Duna :-).

JA: "With respect to the defending (losing) player dropping a piece, this is a little more interesting. By definition nothing the defending player can do is going to avoid checkmate, so the fact that the player has a queen in hand, for instance, is relevant in the sense that the solution must be one where dropping the pocket piece won't help, no matter where it is dropped."

... and I can only add that this reflection can be turned into valuable problem in itself. Everything depends on author's imagination. There is more than slight chance that good problems shall appear.

Juraj Lorinc, judge of Pocket Pieces TT

Written by Juraj Lorinc.
WWW page created: August 7, 2001. (Some delay due to late posting on the web by Hans Bodlaender.)