[Retros] rights & ocassions / not answering Andrew anymore
grol33 at gmail.com
Mon May 26 13:03:36 EDT 2014
The 2003 SPG problem is "correct" under the stipulation it was published
under, which was "drawn", or "remisiert" in german. This is much better
than "draw" which is commonly perceived as an instruction to "play and
draw". As a fait accompli, "drawn" clearly delivers extra information to
the solver needed to make the solution unambiguous. One of the hidden retro
conventions is that players cannot surrender and cannot agree to a draw. An
automat of some kind is required to do the job.
Is it computer-checked? The first 18 moves were already computer-checked in
2003, the rep-cycle was too complicated then and apparently still is today.
I am sure though Michel Caillaud checked it out before he withdrew his own
version on the same theme. We were working on the same material in that
time. Fortunately, I had a head start on Michel of about one year. You
can't expect to beat him in a an even contest!
How I composed it? The way one expects from retro composers, back to front.
First I designed the long repetition cycle in several stages. As there was
still lots of freedom and material unused, I then attempted to setup an SPG
around it. I ended up with 3 versions one of which was refuted by the
computer. I chose the cleanest one remainng to put all emphasis on the rep
cycle. If I am not mistaken, this is now also the SPG with the highest
(possible?) ratio of repetition versus non-repetition moves. There are some
very short SPGs with repetition cycles but I couldn't find one with better
statistics than mine. You can take that as a challenge if you like!
Was the composition a success? Yes and No. It showed good content and
scored 2nd prize in Probleemblad. On the down side, I also intended it to
demonstrate a "premature repetition" try - see my other problems in that
period - but couldn't find an SPG to match. And then there is the little
bug nobody noticed and which I trust somebody will find out about some day.
Why spoil the fun by giving it away now?
Will it be in my book? Probably as an introduction to automatic draws. It
is a good example but without much theoretical significance. "Premature
repetitions" is a far more interesting and controversial subject for
treatment and debate. Ask Andrew about that!
On Sat, May 24, 2014 at 10:26 AM, per olin <per.olin at luukku.com> wrote:
> Dear all,
> The discussion here has centered around situations, where the subsequent
> play makes only one start of a game possible from many possibilities. The
> problem from 2003 is probably not of this type. The move sequence up to
> 18.Txg8+ is unique (C+ Euclide); the repetition moves are separate from
> this. I recall some time ago testing the repetition phase with Popeye
> a=>b. I am not sure did I get the whole sequence tested, could be that I
> stopped after having only the major part tested. Perhaps somebody with
> better computer abilities or the author can verify the uniqueness of the
> Best wishes
> Nicolas Dupont kirjoitti 24.05.2014 kello 00:49:
> > Dear all,
> > > P1011937,
> > > Guus Rol
> > > R222 Probleemblad 12/2003
> > >
> > > (13+13) C+
> > > Die Stellung ist Remis. BP in 41,5
> > > 1. h4 g5 2. hxg5 d5 3. Th6 d4 4. Tg6 Lh6 5. Tg7 d3 6. g6 dxc2 7. d4
> > > Dd5 8. Lg5 Dxg2 9. Kd2 Dd5 10. e4 Dxa2 11. Df3 De6 12. d5 Sf6 13.
> > > dxe6 Sd5 14. Df6 Sc6 15. f3 Sd8 16. Lb5+ c6 17. Se2 Tg8 18. Txg8+
> > > Lf8 19. Df4 Sf6 20. Th8 Sg8 21. Lf6 Lh6 22. Lg7 Lg5 23. Lf8 Sf6 24.
> > > Tg8 Sd7 25. Lg7+ Sf8 26. Lf6 Lh6 27. Tg7 Sd7 28. Lg5 Sb6 29. Df6 Sd5
> > > 30. Tg8+ Lf8 31. Df4 Sf6 32. Th8 Sg8 33. Lf6 Lh6 34. Lg7 Lg5 35.
> > > Lf8 Sf6 36. Tg8 Sd7 37. Lg7+ Sf8 38. Lf6 Lh6 39. Tg7 Sd7 40. Lg5
> > > Sb6 41. Df6 Sd5 42. Tg8+ remis
> > If I understand, the stipulation means that there are plenty of
> > solutions (for example moves by the wQ are not uniquely determined),
> > but only one where the diagram position already occured 2 times during
> > the solution.
> > It seems curious that PDB asserts it is computer tested with Natch,
> > although shorter games with same stipulation are not (e.g. P1003993 in
> > 11.5). Is it a mistake or is there really a way to check such
> > problems? Looking at hand each solution to detect the only one
> > satisfying the hypothesis?
> > Best,
> > Nicolas.
> > _______________________________________________
> > Retros mailing list
> > Retros at janko.at
> > http://www.pairlist.net/mailman/listinfo/retros
> Luukku Plus -paketilla pääset eroon tila- ja turvallisuusongelmista.
> Hanki Luukku Plus ja helpotat elämääsi. http://www.mtv3.fi/luukku
> Retros mailing list
> Retros at janko.at
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Retros