[Retros] Retro solving competition Messigny

Pascal Wassong pascal.wassong at free.fr
Wed Jul 7 10:42:07 EDT 2004

Hello,

>>>>> "Mario" == mri_two at t-online.de (Mario Richter)

>>>>> "Joost" == Joost de Heer <ildjarn666 at linuxmail.org> writes:

Mario> Regarding the results table it seems possible to loose some
Mario> points by giving incomplete(?) solutions.

Joost> Probably gain some points by giving partial solutions. Or
Joost> perhaps the same bonus as in the Dutch retro solving
Joost> competition was used: Guess the composer(s).

No, there are no points for guessing the composers. But I am sure
that some solvers guessed half of them.

Mario> So my question - especially with respect to the SPGs and
Mario> the "game determining final move problem" by François
Mario> Labelle that were used in that contest - is:

Mario> What constitutes a correct AND complete solution of such
Mario> problems?

Joost> In a solving contest, a correct solution is a solution
Joost> which leads to the goal. No proof needed, cooks are equal
Joost> to author's solution. An incorrect solution is a solution
Joost> with a writing error, or a partial solution. Some points
Joost> can be given for this. (but no points for 'cook schemes',
Joost> Michel :)

Joost is right. It is enough to find a solution leading to the
diagram to gain maximum points for a problem. So a cook gives as much
points as the intended solution. Hopefully, no problem is cooked.

It is possible to gain (not lose!) some points by giving only partial
solutions. This is seldom the case for SPGs, but often for classical
retros. Though I would have given some points for partial proof
games.

Problem 7, by Tom Volet, has the stipulation "Order and details of
captures". They are 6 captures, and a solver only gave 4 correct
captures. I awarded him 3 points. (Giving him 2 or 4 points would
not change his ranking, so this had no importance).

Problem 8 has 2 questions : "Last move" and "Give last moves as long
as they are completely determined." 2 solvers found "half" of the
last completely determined moves (this is were Michel Caillaud "lost"
2 points), and Thierry Le Gleuher is the only one who found the
complete solution (he didn't find the first problem by François
Labelle).

Finding half of a solution is possible, because when a solver finds
some unique last moves, he loses nothing writing this down, even if he
didn't completely solve the problem.

I added the question "Last move" to this problem to allow most solvers
to gain at least 1 point on this problem. Only 3 of them took this
opportunity.

For these kind of problems, it is not necessary to prove the solution.
It would take too much time to write it down. I choose the
stipulation so that it is easy to verify the solution. The
stipulation would probably not be the same if the problem had been
published in a retro column.

I also gave 2 points to a young boy, 10 years old, which gave a
(dualistic) proof game leading to the diagram, except that he didn't
see that wK and wQ changed there places. He passed 2 hours trying to
solve the problems, competing against a world champion, and giving him
no points would have been too discouraging.

Pascal

PS : I can probably send the Word document (or in text format)
containing the problems privately to people requesting it, so they
won't be published publicly on the mailing list.