[Retros] No occupied white square

[Francois Labelle]

On 12/06/2013 12:28 PM, Noam Elkies wrote:

> Heinonen's problem was still the first of its kind


Well technically there were earlier realizations with massacre proof 
games like 
http://pdb.dieschwalbe.de/search.jsp?expression=PROBID=%27P0004246%27 .


> And you've already published some

> of your earlier computer discoveries in problem magazines, so why not these?


Maybe because that one wasn't particularly hard to program and to 
compute, but if the result is interesting then I guess that's what 
counts.  I'll try to get them published.

Some extra facts:
- 10.0 is also the minimum length for a non-unique proof game (for both 
white squares and black squares). This was Eric's original question.
- There are two unique proof games of length 10.0 with men on white 
squares, but the solutions are almost the same (one has 1 more capture, 
leaving 20 men standing).
- There's only one unique proof game of length 10.0 with men on black 
squares.

So my two PGs are essentially unique.


> P.S. Is it feasible to compute how many such proof games there are in

> 10.5 or even 11.0 and beyond?  It would be interesting to see what

> the range of possibilities is (and how much choice Heinonen had for

> his 11.0-mover).


The search in 10.5 moves just completed. There are 177 and 98 unique 
proof games with men on white squares and black squares (respectively). 
I extrapolate that 11.0 moves will take 2 months, so I'm not sure if 
I'll do it.

On 12/07/2013 02:23 AM, "Bernd Gräfrath" wrote:

>

> I would also like to see your two problems published in a magazine, 

> preferably side by side; and the pioneer problem by Unto could be 

> mentioned in the comments.


By the way Olli informed me privately that Unto found a 10.5-mover 
(before he was informed of my 10.0 movers). If I publish my problems I 
could also mention any of Unto's problems (if Unto is interested).


> By the way: I think that computer-helped discoveries can also qualify 

> for awards, and even for the FIDE-Album.


I think so too, although such problems might be harder to judge. A 
problem that is impressive to a human composer might have been easy to 
generate by computer. Is this unfair? On the other hand, treating all 
computer-helped discoveries as "easy to generate" would be wrong, 
because they might not be.


> (I hope that you have sent your recent 2-King-PG to the FIDE-Album?)


Yes I did send it. Fingers crossed!

     François

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.pairlist.net/pipermail/retros/attachments/20131209/abe5dc3c/attachment.htm>