Chess Problem Database Server

167 problem(s) found in 1208 milliseconds (displaying 100 problem(s)). [A='Kemp' and FIRSTNAME='%Charles Edward'] [download as LaTeX]

1 - P0000848
Denisson Nixon
Charles Edward Kemp
183 Die Schwalbe 04/1970

(2+3)
ser-h#14
Wieviele Lösungen?

Eine mögliche Lösung: 1. a5 2. a4 3. a3 4. a2 5. Ke5 6. Kd4 7. Kc3 8. Kb2 9. Ka1 10. b5 11. b4 12. b3 13. b2 14. b1=L Lg7#.
play all

Es gibt 112,112 Lösungen, wobei jede spielgerechte Abweichung zählt.

Forget about aP for now. Number of paths for bP 7 bK = c(5,4) - collisions.
c(5,5) = 10!/5!/5! = 252
Collisions of two types: when bK moves onto bP & vice versa. Can't have both.
For each the number is the same: one moves 4 times and other moves 3. Then the critical move when one treads on the other. Finally either can move off first.
So that's c(4,3)x2x2 in total = 140
Difference is 112.
Now blend in aP moves c(10,4) ways to do this = 1,001, hence result

more ...
comment

Keywords: Path enumeration, Seriesmover
Genre: Mathematics, Fairies
FEN: 5B1K/pp6/5k2/8/8/8/8/8
Input: Gerd Wilts, 1995-06-03
Last update: A.Buchanan, 2017-09-22 more...

2 - P0001722
Thomas R. Dawson
Charles Edward Kemp
5677 The Fairy Chess Review 8 10/1943

(8+12)
h=Retropatt in 3

R: 1. Kf7xSe7 Sc8-e7 2. Kg8xTf7 Tc7-f7! 3. 0-0 Td7-c7 Retropatt
play all

Henrik Juel: Stipulation should be 'help-retrostalemate in 3'. -0... Kf7xSe7 -1.Sc8 Kg8xTf7 -2.Tc7! 0-0 -3.Td7 and Black is retrostalemate. Not a defensive retractor, but a help retractor with a weird aim. (2004-03-18)
Sally: Zurück:
1. Kd7xSe7? - Sg8-e7,
2. Kc8xTd7? - Tg7-d7,
3. O O O??? - Tfg7!
In der Verführung wäre die Rücknahme der langen Rochade illegal, da der w. Bauer a2 als Schlagobjekt im Osten gebraucht wurde und sich auf "a8" umwandelte. Somit musste der
schwarze Turm auf "a8" bereits gezogen haben! (2018-06-05)
comment

Keywords: Help retractor, Symmetrical position, Asymmetrical solution, Castling
Genre: Retro
FEN: 3r1r2/1p2k2p/8/2PPPPP1/2PpppP1/3pqp2/4p3/4K3
Reprints: 336 FIDE Album 1914-1944/III 1975
H28 ASymmetrie 2013
Input: Gerd Wilts, 1995-06-03
Last update: Alfred Pfeiffer, 2018-06-05 more...

3 - P0001723
Charles Edward Kemp
The Fairy Chess Review 1944

(7+12)
Retro-Hilfspatt in 6

R: 1. Kf7xTg8 Td2-d8 2. Ke7xDf7 Dc4xBf7+ 3. Kd8xSe7 Sd5xBe7+ 4. Kc8-d8 Tg3-g8 5. Td8xSf8 Sd7-f8 6. 0-0-0 Kc5-c6
play all

Rosalie Fay: Retract -1 ... Kf7xRg8 -2 Rd2-d8 Ke7xQf7 -3 Qc4xPf7+ Kd8xNe7 -4 Nd5xPe7+ Kc8-d8 -5 Rg3-g8 Rd8xNf8 -6 Nd7-f8 0-0-0 -7 Kc5-c6 and Black is in retrostalemate. Black may not play Na6-c7, Ne6-c7 or Bf2-e1; nor g7xf6 because [bBf8] is on e1. Black's 5 captures are g7xf6, dxexfxg and cxb. Therefore Black may not play axb or h3xg2. (2021-05-22)
A.Buchanan: Thanks for the solution, Rosalie: not easy to find. What exactly is a retropat? It's defined in the V&V Encyclopaedia as: "A position where the side whose turn it is to retract has no valid last move." But how is that different from just "illegal position"? If here White retracts e.g. 4. ... Tg3xTg8, then the position is suddenly illegal, as White needs two pawn captures dxc and also hxg (which then must promote btw, as Black only captured 1 unit on g file). The position is illegal, so Black has no retractions to a legal position. Why is this not considered a short solution? Thanks! (2021-05-23)
A.Buchanan: All I can think is that maybe a retropat is a position which would be legal with the other player to retract? But what's the real answer? (2021-05-23)
Henrik Juel: Niels Høeg defined his 'Retromate' in 1923, now termed Defensive retractor, type Høeg, with forward stipulation #1
Around 1938 he defined 'Help-retrostalemate' but soon realized the intrinsic inconsistency (as you note, Andrew) and abandoned this genre
He tried to save his idea by defining 'Help-sub-retrostalemate', where Black only has one possible retraction in the final position, but this turned out to require eliminating unchecks and the double pawn step before an ep capture from the list of 'possible retractions', and the Helpsubretrostalemate was not a succes (2021-05-23)
A.Buchanan: Thanks Henrik - I can't help thinking that there is something to be salvaged here, and V&V's mention of the PTM might allude to this. Or maybe not. Aside from the stipulation, the term also refers to a retro try which leads to a cul-de-sac. And there are a few barely-used undocumented PDB keywords in German on the subject. (2021-05-23)
Henrik Juel: Høeg defined retrostalemate as
An illegal position that would be legal, if the side to retract could avoid his obligation to do so
As you point out, the concept is useful when talking about tries in a retroplay
I consider retrostalemate as a dynamic form of illegality, to be distinguished from static forms like impossible pawn structure, too many captures, etc.
Perpetual unchecks is another type of illegality, where both sides are forced to continue unchecking, so there is no way to reach the initial array (2021-05-23)
A.Buchanan: Thanks Henrik. Høeg's definition as you've just given it seems rock-solid. Mathematically, chess can be viewed as a graph where each node is a position, and each edge is a legal move. The legal positions are those with the game-array position as an ancestor. There is then an invertible mapping call it "flip" which changes the player to move. This underpins threats, set play, seriesmovers, passes, parity, and now it seems retropat too. A retropat is an illegal position whose flip is legal. What's the issue that caused him to abandon this? (2021-05-24)
Henrik Juel: From 'Høegs skakproblemer' (1982) p.147 I quote in translation:
As the aim of a helpretrostalemate (hRS) problem is an illegal position, there arose some criticism against the genre, and Høeg modified it by defining the helpsubretrostalemate genre...
I think the hRS genre can be saved easily simply by allowing White to make an illegal retraction (illegal only because it leads to RS) as his last move
A similar trick is use to 'save' the doublemate aim, ## (2021-05-24)
Henrik Juel: In Aarskrift 1938 p.6 the editor describes the October 1938 DSK meeting, where Høeg presented his recent invention, hRS. The editor writes (partly translated):
The interested reader is referred to Skakbladet 01/1939, The Fairy Chess Review, and the chess column in Eskilstuna Kurieren; the retro expert TR Dawson calls the hRS 'a complete novelty and a delightful one'.
There is also an original problem by Høeg
dia: wKe2, sKe5, sSh4 (1+2)
stip: hRS3
R: 1.Kf6xDe5 De3-e5+ 2.Kg6xDf6 Dd4-f6+ 3.Kf5xTg6 Ke1-e2!
(note that black retractions are written first, because it is a help problem)
I think that we should consider the hRS genre resurrected, despite the 'some criticism' that led Høeg away from hRS and to the less fruitful h-sub-RS (2021-05-24)
Henrik Juel: 'Der Blick zurück' (2006) discusses the concept of retrostalemate in great detail, but has no mention of Høeg's hRS genre, as far as I can tell (2021-05-24)
A.Buchanan: Thanks Henrik: so the usual "illegalphobia" muddied the waters yet again. I think your repair is a good idea. But I'm not sure about the phrase "illegal retraction" though. I distinguish between illegal moves and positions. The rules just deal with illegal moves. They say nothing about illegal positions, which is a concept from the Codex. For any legal or well-formed illegal position, we can specify the set of legal moves. (An ill-formed position has e.g. wPa1 If the rules are sound, all ill-formed positions are also illegal.) In retro terms, the legal moves may come from legal or illegal positions. But I would like "illegal retraction" to be the inverse of a legal move, and not imply anything about the position. What language I would like to use instead, I don't know yet! What do you think? (2021-05-24)
Henrik Juel: no comment... (2021-05-24)
A.Buchanan: That has the effect of a comment, Henrik, alas. And a dampening one too, alas :) (2021-05-24)
Henrik Juel: I did not mean to dampen your spirits, but to me a normal retraction connects two legal positions
What is the Garden and jungle principle? Google tells you how to make a forest garden! (2021-05-24)
A.Buchanan: I will make 1 vital point as briefly as I can. I am not trying to be “legalistic” Henrik :-) but we need a mathematical model for chess which can be extended to the fairy domain. It’s explicit in the rules that move legality is independent of position legality - which is not even mentioned in the rules. For example, checkmate requires only that the last move be legal. This is perfect because we want to be able to play chess without having to validate whether the position is legal. In fairy chess or on bigger boards, legality may be undecidable. So we really need this FIDE principle for fairy chess. The Codex then defines legality of position in terms of whether the game-array is an ancestor. Again perfect. Now for retro-moves. A retro move is the inverse of a move, and legality should be defined based simply on the FIDE concept. We want to be able to retract moves even in positions which do not derive from the game-array. There are formats e.g. Horizontal Cylinder where there are no legal positions apart from the game array itself! We can still do retractions in such a fairy format. So we can’t define legal retro-move in terms of legal position. Ok so far? (2021-05-25)
A.Buchanan: The concept of legal retro-move grounded in legal move has been over-shadowed for generations by the interpretation that Henrik describes. But this idea would very useful in its own terms. Look at all the Baird-style retractor #2s, or the myriad Wenigsteiner help-retractors. It's ridiculous that there is no (public) engine today which can solve them, when all you need to do is apply the movement rules backwards to the current position. Yes you will need to check manually that there is no "strategic" factor which might invalidate a solution, but we are getting more comfortable with HC+ these days. I bet no such concern applies to any Baird retractor. Agreed, Henrik? (2021-05-25)
A.Buchanan: So local retraction is fine for Baird (where everything is legal) and the fairy world (where we everything is illegal or we don't care about positional legality anyway), but there are these problems which are looking at the border between legal and illegal positions, and we local retraction is not sufficient. These two concepts have historically been conflated, and share a single term. In the formal chess Garden, concepts can and must be made solid, but sometimes naming them is hard, while in the Jungle of what is actually out there, we have the names (e.g. "fairy", "version", "tempo"...) but defining them is impossible. So how would you want to name them distinctly. Maybe we should say that "legal move" is redundant - we should just say "move". And then "legal move" would be used in cases where the state (i.e. position) prevents us from castling. This would mean that local retraction is just "retraction". And then we can keep "illegal retraction" for the sense that the world is used to. What do you think Henrik? What words would you like for these concepts? (2021-05-25)
Henrik Juel: You are way ahead of me, and I really do not know what to say, Andrew
I think I will just stay in the Jungle... (2021-05-25)
comment

Keywords: Help retractor, Castling (sg)
Genre: Retro
FEN: 3R1rk1/p1n4p/2K2p2/1B6/1p6/PpP2P2/b1P3p1/1n2b3
Reprints: 338 FIDE Album 1914-1944/III 1975
Input: Gerd Wilts, 1995-06-03
Last update: A.Buchanan, 2021-05-23 more...

4 - P0004325
Charles Edward Kemp
London Evening News 1943

(7+9) cooked
Schwarz nimmt 1 Zug zurück, dann h#1

R: 1. g5xSf4, dann 1. Le5 Sd5#
play all

Cook: R: 1. e7xSd6, dann 1. Le5 Se4#

Henrik Juel: Diagram error? -1... g5xSf4, 0... Le5 1.Sd5#. But the symmetrical solution -1... e7xSd6 etc. seems legal too. (2004-03-22)
A.Buchanan: Yes bBf8 would have died at home but bPc could capture twice to promote on c1 or e1 (2023-03-31)
A.Buchanan: Odd. The diagram is thematically symmetric, so not just missing a unit. WinChloe has the same diagram. I wondering if we are missing something from the stipulation. Does anyone have access to an archive for this? It's also odd that it was reprinted: surely there would be a comment there? I suppose it can be rendered sound by placing wTTSS somewhere symmetric. Then Black runs out of pawn captures if sLd4 is promoted. In any case, wPh7xg8=S is affordable. (2023-04-01)
Mario Richter: The reprint in 'Problem' says: "Ne e7xSd6 radi nepravilne postave" ("Not e7xSd6 because of illegal Position.") The diagram in 'Problem' is the same as here in the PDB.
How about adding black pawns on b7 and g2? (2023-04-01)
A.Buchanan: Hi Mario. I considered adding sBb5e2. That gives 10 needed captures by Black and 6 by White, including sLf8. So the try fails for *two* reasons. sBb7g2 fails even more badly. I thought that also adding wTTSS although more uneconomical in pieces defeats the try for only one reason. Also I liked the trick h7xg8=S. And it must make it harder to find the solution. Does Rawbats do simple retractors? Can you try wTa0h1 wSa7g1 as a risky initial bid, and then back off until there are no cooks? (2023-04-01)
Mario Richter: Hi Andrew, I was well aware that with adding sBb5e2 or sBb7g2 the try fails for two reasons (for [subjective] optical reasons I preferred sBb7g2 to sBb5e2]). The reason I suggested the addition of 2 black pawns was the following consideration: I cannot believe that such an experienced Retro Composer like C.E. Kemp could make such a big mistake. It's much more likely that the printer made an error by omitting two pieces.
On the other hand, if we want to correct the problem including a unique reason for the try to fail, your suggestion of adding wTa8h1+wSa7g1 (FEN=R7/N4pp1/1P1pnkp1/5n2/3b1p2/1PP5/1PP2P2/K5NR) works - rawbats gives exactly R: 1. ... Bg5xSf4, dann 1. Ld4-e5 Sf4-d5# (and says about R: Be7xSd6 - "Illegal: Mehr schwarze Bauern-Schlaege notwendig als weisse Steine fehlen") (2023-04-01)
A.Buchanan: Hi Mario, thanks for this, I agree with everything you say. It's curious that the reprint in Problem did not pick up the error. Any overlooked black units are more likely to have been lurking on dark squares, maybe sBb4d2. This forces the try to only fail for 1 reason, albeit rather banal. Does this work. Or the +wTTSS mentioned forces the solver to do a little more counting, and there is a lot more solving space to search for the h#1. Rawbats' conversational skills seem to have improved: have you integrated it with ChatGPT maybe? (2023-04-02)
comment

Keywords: Help retractor, Symmetrical position, Asymmetrical solution
Genre: Retro
FEN: 8/5pp1/1P1pnkp1/5n2/3b1p2/1PP5/1PP2P2/K7
Reprints: (24) Problem 83-86 09/1962
Input: Gerd Wilts, 1995-06-03
Last update: A.Buchanan, 2023-04-02 more...

5 - P0004828
Charles Edward Kemp
The Fairy Chess Review 1936

(7+11)
Schwarz nimmt 1 Zug zurück, dann h#1

R: 1. ... exd3ep, dann 1. Kxd4 Td2#
play all