[Retros] Retro-strategy and new AP-type

Rol, Guus G.A.Rol at umcutrecht.nl
Tue Jun 26 08:35:30 EDT 2007

Dear Valery,

Thank you for your comments! Had there been just one "refutation" for
4... Rd8+? I would have included it in my annotation but unfortunately
there are more after 5.Kxb7! Rd7+ and now either 6.Kc8 Bxg7 7.b7! Ke7
illegal mate, or 6.Ka8 Bxg7 7.Pxf7+ Kxf7 illegal mate. Btw, I wouldn't
mind dropping the repetititon rule for this composition. What could be
more fun than see black string together an infinity of checks which he
must all take back in the end because he can never justify them by

Guus Rol.

-----Oorspronkelijk bericht-----
Van: Valery Liskovets [mailto:liskov at im.bas-net.by]
Verzonden: dinsdag 26 juni 2007 13:30
Aan: The Retrograde Analysis Mailing List; Rol, Guus
Onderwerp: Re: [Retros] Retro-strategy and new AP-type

Dear Guus,

The intention of your problem is fresh, beautiful and, I believe,
correct in the framework of the AP-logic: indeed, we see a posteriori
why B couldn't (or had the right not to) mate on the 2nd move. A
noticeable (although reflexmate restricted) contribution into the

However, I have doubts as to the soundness of the problem as is: in
order to postpone castling, instead of immediate 4... Bxg7, Black could
play first 4... Rd8+!, and only after 5.Kc7(xb7) Rd7+ 6.Kc8 take 6...
Bxg7, etc. If I'm right, maybe merely the repetition rule saves your
position (with r#8)!?

Btw, the additional thematic nuance mentioned by you which is concerned
with the forced illegal reflexmate Ke7# (instead of compulsory 0-0)
would have been very interesting, particularly if it could arise on
greater than n move (in r#n) as the only way to disqualify some B's

Valery L.

"Rol, Guus" wrote:

> Dear retro-friends,The solution to the reflexmate problem I posted

> last week is not difficult once you get the right AP (a posteriori)

> idea. Look at the equivalence with e.p in a formal logical context:A

> Posteriori:1. If "castling right" proves "e.p. right" in all proof

> games then "playing e.p". can be justified a posteriori by

> "castling"2. If "the right to 0-0" proves "no right to 0-0-0" in all

> proof games then "denying the right to play 0-0-0" can be justified a

> posteriori by "0-0" "Denying the right to play 0-0-0" in a reflexmate

> problem is demonstrated by refusing to mate through long castling and

> needs a posteriori justification as shown in this

> composition:Problem:White: Kd6, Bg7, Sd7, P:b6,c4,f5,g6Black: Ke8,

> Ra8, Rh8, Bd8, P:b7,c6,f6,f7.R#6 (AP = A Posteriori).Solution:Note

> that one of the black castlings is illegal.1. Sxf6+ Bxf62. c5 Rd8+!

> (black refuses to 0-0-0 and must justify that by playing 0-0 later)3.

> Kc7 Rd7+4. Kc8 (Kb8? Be5+! Kc8 wastes a move) Bxg7Now white is in

> zugzwang; black on move would be forced to play Ke7 illegal mate!5.

> Pg6xf7+ Rxf76. Pf6 0-0 reflexmate!Black is happy to accomplish his 0-0

> justification and white to get his timely reflexmate. Note that black

> cannot successfully delay the mate with additional checks!Guus Rol.


> ------------------------------------------------------------------

> _______________________________________________

> Retros mailing list

> Retros at janko.at http://www.pairlist.net/mailman/listinfo/retros


More information about the Retros mailing list