Comments | motris: Friday Puzzle #8

motris

Friday Puzzle #8 - Color Nikoli Sampler

Jul 31, 2009 00:01

Today is noteworthy in that Nikoli.com is running 3 contest puzzles at the same time, and even better, the contests are at 9 PM local time for me (not the usual 5 or 6 AM). So, given my lively Friday evenings, assuming I'm home from work, I'll be competing. In increasing likelihood, I'll start with Slitherlink << Shikaku < Nurikabe. It would be ( Read more... )

nikoli, slitherlink, fridaypuzzle, nurikabe, shikaku

Comments 10

thedan July 31 2009, 11:28:07 UTC

I am all for you not doing the Shikaku first.

rpipuzzleguy July 31 2009, 17:02:23 UTC

Wait, they're only giving out one prize? They had a winner for each of the three puzzles last time, as I recall. I'd be surprised if they didn't do that again; I think maybe their wording on the website may be imprecise.

motris July 31 2009, 17:13:54 UTC

No, there are three prizes. I'm saying the most likely way I'd win is if there were a three puzzle final, as I'm sure I'd be ~2-3 on nurikabe, ~3-4 on shikaku, about 5-6 on slitherlink

stigant July 31 2009, 22:19:16 UTC

Are we allowed to assume that the numbers in each grid generate a uniquely solvable puzzle of the appropriate type? (Note that this is slightly different than assuming that the puzzle as a whole has a unique solution)

motris July 31 2009, 23:16:44 UTC

This has to be the case. There is nothing else constraining how you solve the individual puzzles so if you have the right numbers, those puzzles alone would solve completely.

ertchin August 1 2009, 02:21:24 UTC

I did better on these than I expected, until I got hung up right before the end. Stared for minutes before I spotted the extremely obvious portion of the Nurikabe that I hadn't filled in. Fun!

grandpascorpion August 4 2009, 00:28:12 UTC

Very nice idea and execution. It would be nice to see something like this on a larger board (although I know from personal experience how labor-intensive these combination puzzles can be).

motris August 4 2009, 00:43:06 UTC

It didn't take too long to make an 8x8 puzzle. It did take a very long time to make it solve well, with a lot of tweaking.

I could imagine, if I was actually interested in doing it, making a 15x15 color mix. I would drop Slitherlink for Four Winds or Corral though so that puzzles with numbers in similar ranges would occur. It also seems likely you could add the step of puzzle identification to such a construction (ie which of these is the nurikabe?) as an extra challenge.

grandpascorpion August 4 2009, 03:12:54 UTC

Sure, kind of follows the 80-20 principle (or is it 90-10).

That would be interesting. I guess the big challenge would be reining in the difficulty level.

Never has one man worked so hard to achieve so little ... anonymous September 21 2009, 18:43:22 UTC

I continue to work backwards through these, and I must admit that this one took me far too long. Loved every minute of it as I knew that there would be a light if I kept slogging, but I suspect that I must have missed far easier paths than I found to get there.

Never hit a mistake, just took too long to push through.

Thanks for this novel twist on three of my favorite puzzle styles.

Ken

*** "How many could have created it? ... how many could have solved it? How wonderful for both."