[Retros] Accidental retro

Noam Elkies elkies at math.harvard.edu
Sun Feb 5 23:18:46 EST 2006

Mario writes:

> Noam wrote:

>> Leafing through a copy of Laszlo Polgar's _5334 Problems, Combinations,

>> and Games_ that I saw on a coffeetable earlier today, I noticed

>> the following miniature in the Mates In Two section (#2940, p.547):

>> W Kg6 Rf6 Nc8 Pe6 Pf7 = 5

>> B Kf8 Pe7 = 2

>> The retroanalytic content is surely not intentional ...

> I think this position is misprinted, the correct place

> of the white king might be: Kh8.

Do you have a source for this correction? While this change
would indeed preserve the actual play and make the position legal WTM,
it would eliminate the setplay, making it a pointless problem
(and out of place in this section of the book).

> Btw., the book also contains some problems, where the retroanalytic

> content is intended, like: [#1942: r3k1K1/1RR3p1/7p/8/8/8/8;

> #1747: 8/8/8/8/8/8/kPP5/1nQ1K2R]

Evidently you've spent much more time with Polgar's book than I.
By now the coffeetable where I saw the book is several time-zones away,
and I have not been able to locate a copy in the Harvard library system,
nor in either of the main nearby bookstores. Does the book mention
the retroanalytic content in the solutions section? For that matter,
what does it give as the solution of #2940? Since the chapter name
includes "White to move", it would be strange to see a BTM problem,
but I suppose it might have been included as a trick question,
unlikely as that seems in a beginners' book. A beginner might well
not even consider Castling in positions like #1747 and #1942 that are
so far removed from the opening.

> Most interesting for me was

> No. 1513 (and similar No. 1517)

> W: Nb8 Kc8 Rb7 Pa6

> B: Ka8 Ra7

> #2

> _________________

> | |

> | k N K . . . . . | 8

> | r R . . . . . . | 7

> | P . . . . . . . | 6

> | . . . . . . . . | 5


> According to the book, the composer of this problem is

> Mrs. W.J.Baird, 1907.

> Is it really likely that Mrs. W.J. "Twentieth Century Retractor" Baird

> overlooked such a simple retroanalytical fact?

It seems particularly unlikely here: there's no BTM solution,
and the absence of a legal last move is the only nontrivial thing
about the problem. Perhaps she constructed it as a schematic example
of an illegal problem. Here too I ask: what do L.Polgar's published
solutions for problems #1513 and #1517 say?


More information about the Retros mailing list