Coincidence?!?

Discuss all puzzle-related subject here
Post Reply
detuned
Posts: 1672
Joined: Mon 21 Jun, 2010 2:25 pm
Location: London, UK
Contact:

Coincidence?!?

Post by detuned » Thu 24 Mar, 2011 1:17 pm

So the latest nikoli.com Masyu puzzle - published March 24 2011 at 03.00 and authored by Guten - didn't feel quite right:
seenbefore.png
seenbefore.png (19.95 KiB) Viewed 5042 times
Why - well it's identical (modulo symmetries of the square) to a puzzle I put on my blog in July 2010. here.
Image

I think I'm inclined to believe this is something of a coincidence. I really want to believe that nikoli authors aren't plagiarists. The thing is, trying to write a Masyu puzzle with twisted symmetry is quite a thankless task as the aesthetic constraint is really quite strong. I can sort of believe any author might play with the idea of having diagonal lines of givens and stumble upon the configuration of these three particular lines. For one thing, they look pretty cool, and another they actually yield quite a nice solve. Having got these in place, forcing things to work with effectively 3 extra circles, again placed in a fairly aesthetically pleasing way whilst forcing the unique solution doesn't seem such an extra leap, given the intial configuration of the three lines.

But...come on, it's identical!!! Seriously, what are the odds?!

What are people's thoughts on this one?

GarethMoore
Posts: 1176
Joined: Thu 24 Jun, 2010 9:27 pm
Contact:

Re: Coincidence?!?

Post by GarethMoore » Thu 24 Mar, 2011 1:31 pm

I'd say it was coincidence, since as you say it's a very constrained problem to finish off a puzzle symmetrically (let alone with twisted symmetry), and you might reasonably imagine that someone else will try starting a puzzle with those central diagonal lines at some point. I know from trying how incredibly difficult it is to make a non-trivial perfect twisted symmetry Masyu in any case. I haven't tested this one, but I guess the question is if you draw the diagonals in then how forced are the remaining placements? If there are only one or two options then coincidence seems perfectly credible. Especially from an author who's been writing puzzles on the site for a very long time.

Mephisto
Posts: 60
Joined: Tue 18 Jan, 2011 12:28 pm

Re: Coincidence?!?

Post by Mephisto » Thu 24 Mar, 2011 2:29 pm

In composing chess problems it is relatively frequently the case that the same chess puzzle is composed twice by different authors at different times – not seldom by composers which are above reproach. In this case the later problems are marked as "anticipated by <original-author>".

For example: http://www.janko.at/Retros/Strategems/N51.htm#R168

~ÔttÔ~

detuned
Posts: 1672
Joined: Mon 21 Jun, 2010 2:25 pm
Location: London, UK
Contact:

Re: Coincidence?!?

Post by detuned » Fri 25 Mar, 2011 2:04 am

Thanks for the input guys. I think all things considered you are probably right. Still, it's very strange, and it definitely took me aback.

I suppose looking at it optimistically, my puzzles can now be recognised as being of nikoli.com standards :roll:

Post Reply