Jan-Mar 2007

R320 - Guus Rol

Probleemblad Jan-Mar 2007

Guus Rol wrote on the Retro Mailing List, Apr 4th, 2007: My composition is an illustration of the appreciation of "available game state information". Two well-known retractor-problems by R. Kofman (also -1 and #3) use contradicting principles on the treatment of retraction information. I have combined elements from both of them in 2 of the 4 variants of my problem to demonstrate the impact of rule-consistency. In my view one of Kofman's originals is a misdirected interpretation of the mutex (mutually exclusive castling) convention. Which in turn shown that the mutex convention is not just redundant but also dangerous when it causes confusion amongst retro-composers.

a) Retract -1. Rd5xPg5. The black pawns captured 5 times, so the g- and h-pawn didn't cross-capture, so the second rook had to enter the northeast corner through e8, and black may not castle. 1. a7 ed5 2. a8=Q Bc8 3. Qc8#; 1...Kf8 2. Rd7 Ke8 3. Rd8#

b) Retract -1. b5xa6ep!. Now castling is mutually exclusive. If white may castle, Rd4 is a promoted rook, and it must have escaped the 8th row through d8, or promoted on e8. 1. OOO! ~ 2. Rd7 ~ 3. Rd8#

c) Retract Bg7xPf6!. This proves nothing about castling rights. 1. Rd7 Rg7 2. Rd8 Ke7 3. R1d7#; 1...e5 2. a7 ~ a8=Q/R#

d) Retract 1. OOO!. Rd4 is promoted and it promoted either on e8, or it had to leave the 8th row through d8. 1. OOO? Rh5! 1. c6! Rh5 2. cd7 Kf8 3. d8=Q#

R321 - Klaus Wenda

Probleemblad Jan-Mar 2007

-1. Kh4-g4 Q~8-h8 -2. g5xh6ep[h2] h7-h5 -3. Bg3xRh2[Bc1] Qh8-~ -4. Kh3-h4 Qa8-h8 -5. a4-a5 (retro-zugzwang) b5-b4 -6. Be1-g3 and 1. Bb4#

R322 - Dirk Borst

Probleemblad Jan-Mar 2007

a) 1. Nf3 h5 2. Ne5 h4 3. Nxd7 Qxd7 4. g3 Qh3 5. Bg2 Bf5 6. O-O e6 7. Bf3 Bd6 8. Bh5 Ne7 9. Bg6 O-O 10. Bh7+ Kh8 11. Bg8 Rxg8

b) 1. g3 h5 2. Bh3 h4 3. Bxd7+ Qxd7 4. Nh3 Qxh3 5. Rg1 Bf5 6. Rg2 e6 7. Kf1 Bd6 8. Kg1 Ne7 9. Kh1 O-O 10. Rg1 Kh8 11. Rf1 Rg8 12. Kg1

To castle or not to castle...