[Retros] Difficult

[Николай Белухов]

 Hi,
If we are speaking of retro problems, then, yes, Breyer's problem is very hard; this is evident at least from the fact that the cook in the original version went unnoticed for several decades. The "Chinese puzzle" type of retros (I think the term was suggested by Guus Rol) seems to produce the most difficult retro problems. The famous "The new bee-hive"
http://www.softdecc.com/pdb/search.pdb?expression=PROBID='P0008419'
was once called "one of the most difficult retros ever" by Joost de Heer. Harry Goldsteen later composed other extremely difficult problems based on the same matrix; personally, I have not tried to solve them and I cannot compare them to Breyer's masterpiece. They can be found on the CPDB using the search criterion, say, 
a='goldsteen' and year>='2008'
Andrey Kornilov composed other puzzle-type problems that require a labourous solving (preferably computer-helped), such as
http://www.softdecc.com/pdb/search.pdb?expression=PROBID='P1017683'
These are the first few that I can think of. I suppose that most of the SPGs with a large number of moves would be extremely difficult to solve by a human, but (as opposed to retros) they are computer-treatable. Unfortunately, I understand almost nothing of SPGs, so here's just one very long:
http://www.softdecc.com/pdb/search.pdb?expression=PROBID='P0000136'
As for orthodox composition, creating extremely difficult problems there is not that difficult when one is equipped with a proper computer program; I've seen many 5 or 6-piece positions with a stipulation similar to "Mate in 370". As for human-composed human-solvable problems, I think that "Mate in 257" by O. Blathy was the record, but length does not actually imply difficulty here (the length is usually pumped up by some multiply repeated simple maneuver)! 
Best wishes,
Nikolai


 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.pairlist.net/pipermail/retros/attachments/20100222/d3e9eca0/attachment.htm>