Log in

No account? Create an account
25 May 2007 @ 09:29 am
Today's Nikoli Championship Puzzle  
So it was a good morning on Nikoli.com. To be honest, I woke up about a quarter-hour before the contest and was not fully awake when it started. I was sure I'd lost when I got to 6 minutes and began to force things with nishios (particularly after I realized I'd taken minutes to place some fairly easy digits as my solving video clearly shows). But apparently the tough parts were tough enough to slow everyone. I am now the World, and Nikoli, Sudoku Champion. The latter comes with a nice t-shirt. W00t!

As an aside, I just realized that character limits on comments (you get 100 on any Nikoli puzzle) is a challenge for English speakers on a Japanese site. A hundred Japanese characters can form many more words than 100 English characters. I often wonder, when I see a long stretch of japanese from another competitor, if they are waxing far more poetically on a puzzle than I can in my language under the character constraint.
THrpipuzzleguy on May 25th, 2007 01:50 pm (UTC)
Congrats! Though I'd suggest taking out the minor spoiler in your comment, and maybe LJ-cutting this entry as well.
motrismotris on May 25th, 2007 02:16 pm (UTC)
I believe I took care of both - as there aren't daily puzzles of my types that I comment on, I was a bit careless with the standard etiquette.
(Deleted comment)
motrismotris on May 25th, 2007 02:24 pm (UTC)
Nishio is my way of saying I looked at some fairly locked digits and assumed one was true to see if it led to a contradiction. To be honest, as its typically over just one digit, it could also be called "painting," although this is not always the case.

I don't really have any recommended resources as the "generic" sudoku vary somewhat in the skills they require based on the source. If this is a championship that involves Will Shortz puzzles, his books may be the best place. I wish there were more data on particular sites about solving times, but you might start to catalog some of your times. Mine aren't the best to compare to, but as a very lower bound you could solve the UK championship puzzles I linked to here and compare with their champion (who was solving on a poster), and me (who was solving on paper).

(Deleted comment)
motrismotris on May 25th, 2007 06:44 pm (UTC)
As you suggest, I agree that for most top sudoku solvers, a puzzle being a variant isn't really much of a ripple. Putting in a diagonal constraint? No problem.... There might be a subset that is not as good at some of the mathematical ones like killer, but a geometric variant is not, itself, going to throw people who are good at visualizing placements for the numbers 1-9 already.

The toroidal at the first WSC was very difficult and I'm amazed I finished it (no one else got very far at all - just one digit in to be exact after fifteen minutes). I think the difficulty stemmed more from the lack of really any singles/pairs to use as work-ins as you'd see in a regular sudoku. It didn't matter that the rows/columns had wrapping properties when you couldn't force a placement. There is one digit most mortals can place, and then a lot of staring. The best proof though is to try it yourself so look here for that puzzle.

I think I'm actually using "generic" not for classic sudoku, but for mass-produced computer-generated classic sudoku. While Will's books sell well, I find the puzzles which are on lower quality paper and with no symmetry to speak of, to be less than ideal. Even Wayne Gould's puzzles seem a step forward, with a good difficulty rating, although I tend to only solve Nikoli sudoku if I have a choice as sometimes they interest me with hidden themes. It is probably a point of contention I have as a puzzle constructor myself, and I addressed it two blog posts ago, but "generic" as I'm using it is definitely a subset of "classic" or "vanilla" sudoku.
(Deleted comment)
(Deleted comment)
motrismotris on May 30th, 2007 06:34 pm (UTC)
[attempted to be as spoiler free as possible] it was 15 minutes total, and it was on stage on big placards. I think its impervious to much guesswork as you won't reach contradictions early. To solve it, I recall finding a property about the shape that mattered, with my eyes drawn to it by two of the digits. I could "test" a possible contradiction and that got me about 6 digits. Using the same kind of thinking, but in a harder sense, gave me another digit that then cracked enough of it to get me to about 12 digits. Time was running short, so I decided, as patterns were arising, to assume something about the puzzle. That assumption proved true, and I finished just before time was up.
(Deleted comment)