From: Colin Paul Adams <colin <at> COLINA <dot> DEMON <dot> CO <dot> UK>
Date: 10 jan 1999
Subject: Re: More on chu shogi mating problems

>>>>> "Steve" == Steve Evans <trout <at> NETSPACE <dot> NET <dot> AU> writes:

    Steve> I have now tried my Zillion-of-Games based chu mating
    Steve> program out on some of the ancient problems without
    Steve> solutions in George Hodges' Middle Shogi Manual.

    Steve> The first problem I attempted was by Ito Sokan III from his
    Steve> 1746 collection of 50 problems. The problem is C11 in the
    Steve> Middle Shogi Manual and is as follows:

    Steve> 5, LN, 6 / 2, S, DE, 8 / 1, C, 1, BT, K, 7 / 1, FL, 3, G,
    Steve> 1, ln, +SM, 3 / 1, P, P, 2, fl, 6 / 3, p, 2, +B, 5 / 2,
    Steve> +dk, 1, ph, 7 / 12 / 7, +RC, 4 / 12 / 12 / 12

    Steve> This is referred to in the Middle Shogi Manual as a flawed
    Steve> problem with no solution. My program appears to have found
    Steve> a 5 move result. It is:

    Steve> 1. FLx7d=, K-9d 2. Ln-7c, K-10c 3. FL-8e+, K-10d 4. Ln-9e,
    Steve> +RCx9e 5. Ph-8f mate.

    Steve> I can't find a flaw with this solution (but then I am a
    Steve> rather poor chu player).  I was hoping someone on the list
    Steve> may care to provide independent comment on this solution
    Steve> before I send the details to George Hodges.

It's a matter of the rules - I sent this in a letter to GFH on
27/11/96:

    === start of extract ===

1. +DK-8e (this threatens Ln-6c mate, so:), Gx8e;
2. Fl-7d= , K-9d;
3. Ln-7c, K-10c; (or K-10d; 4. Lnx8e mate)
4. FLx8e+, K-10d;
5. Ln-9e, +RCx9e;
6. Ph-8f mate.

The triple themes of sacrificing the Soaring Eagle, so as to delay the promotion of the
Ferocious Leopard, and finally the smothered mate after removing
the Whale's coverage of 8f make this a beautiful problem, and convince me that this
was IS III's intended solution. However, this line appears to be flawed, since 1. \dots{}, BTx8e
appears to offer a strong defence. But before going into the analysis of this line,
I want to make a diversion to the subject of Rule 3.

On page 8 of MSM, I read:

``Should a piece not be promoted, it cannot do so on the next move, unless it makes
a capture.''

Now the relevant phrase is ``on the next move''. I interpreted this as (naturally, in
my opinion) ``on the next move made by the player'' - a bit like the KO rule in Go
(perhaps it is because I am a Go player that I find this the natural interpretation?).
However, when coming to problem (B27), after much puzzling over the given solution
(I had worked out one which was 4 moves shorter),
I finally realised that the correct interpretation
should be: ``on the next move by that piece''. This is confirmed also by (C11), as otherwise
there would be a trivial solution:

1.FLx7d=, K-9d; 2. Ln-7c, K-10c; 3.FL-8e+(illegal move), K-10d; 4. Ln-9e, +RCx9e;
Ph-8f mate; (and the Dragon King has taken no part in the procedings, proving
that IS III did not intend this, and therefore understood Rule 3. in the same way
as Ito Sokan I (B27)).

Now despite all this, I still find the natural interpretation of Rule 3. to be my
original interpretation. It may be that some other people will also so
interpret it. Therefore I propose the following reading:

``Should a piece not be promoted, it cannot do so on its next move, unless it makes
a capture.''
   === end of extract ===

George agreed with me. (in fact, I think a similar analysis is in one
of the supplements - I had missed it).

So you have to recode the rules accordingly. (Unfortunately, George
did not include my proposed word changing (along with several others
which I persuaded him to agree with) in supplement number 3 (most of
which is due to me)). Then you will find C11 is indeed flawed.
--
Colin Paul Adams
Preston Lancashire