[Retros] more prior art for GD50JT

[Noam Elkies]

Andrew Buchanan writes:


> Another example of a "Hashimoto-style" promotion

> that I came across recently in a very different setting is:



> rnb1kbnr/1ppppp1p/8/8/8/8/P1PPPPP1/1N1QKBNR

> Jacques Rotenberg, Thèmes-64 1983

> (12+13) SPG in 9.0


Thanks for posting this one.  I'm not sure I'd call this realization
"Hashimoto-style", though; from the first two examples it seems best
to use that name only when none of the promotee's moves are captures.
(Disclaimer: I have yet to go through the new batch that Hashimoto
kindly posted, so I don't know whether each of them also satisfies
that constraint.)  "GD50JT-style", perhaps?

At any rate, I'm not surprised to see the theme combined with
a deletion diagram including a Phoenix impostor.  In fact this
does not require as many as 9.0 moves.  Any realization of the
GD50JT theme with capture of the promotee requires at least
7.0 moves, and this bound is attained (thus also barely qualifying
as a "Shorty" for Mark Kirtley's series) by the following SPG:

delete Rh1, Pawns c2,g2, pawns b7,e7,f7 in 7.0 moves
[NDE 4/4/02, C+ Popeye 3.41 in about an hour].

There must be several minor variations that are also sound;
I stopped after having Popeye verify that deleting a7
instead of e7 also works in the same way.

--Noam D. Elkies