## May 22, 2010

### Friday Puzzles #46

So my attempt to get my numbering back up to where it should be continues.

I was originally going to stick this hitori puzzle in yesterday – it is a much superior puzzle to Friday’s – but I felt it wouldn’t be doing this puzzle justice. I want to talk about this a little bit and that seemed out of place with yesterday’s nostalgic waffle. Suffice to say, some of the logic is quite involved in this puzzle, but what really interested me whilst constructing this puzzle was the whole idea of having so-called “minimal hitori”. The thinking here is that if the numbers in the grid which at no stage feature in any meaningful deduction weren’t even printed in then it might make for a puzzle with a nicer solving flow.

Being the sadistic contrarian that I am, I’ve decided to take a different approach. Most nikoli.com hitori usually feature an “ascending solve” – that is the first deduction will usually be with 1’s, then 2’s and so on. This is true of this puzzle – however what I’ve tried to do with superfluous clues is to disguise them (albeit on a fairly superficial level) amongst what you should really be looking for. Let me reiterate again, this is a tough solve.

Instead of writing “Sorry, Thomas”, I’ll instead go with “Sorry Thomas”. Or perhaps that should simply be: “Enjoy!” :)

#056 Hitori – rated hard

All puzzles © Tom Collyer 2009-10

### 4 comments by 1 or more people

1. #### Jack

I can’t say I found this puzzle much more obfuscated than usual for Hitori. For one thing, a pretty high percentage of this puzzle is signal rather than noise, so trying to make the noise more distracting may not have as big an impact.
I did enjoy the overall logic—starting with a Heyawake-like set of “ladders” rather than the usual 3-in-a-rows is nice, as is the communication across the puzzle with matched sets of ladders. The logic about keeping the puzzle connected in the lower-right is really the major aha moment, and unless somebody never really notices that, it’s not THAT rough a puzzle.

22 May 2010, 02:27

2. Hmmm – well I’m glad you enjoyed this! l I thought it was certainly tougher than a lot of nikoli 12×12 Hitori I’ve seen, bearing in mind there’s a bit more global – rather than localised – thinking to be doing than usual.

It certainly looked more obscure on paper, but perhaps that says something about my laziness in using rulers… :)

22 May 2010, 03:07

3. #### mathgrant

It took me about 7 minutes to complete this one in MS Paint. A little rougher than Nikoli’s roughest 12×12 Hitori, but not as rough as their roughest 17×17 Hitori, and very very elegant. Devious, but also elegant.

24 May 2010, 16:54

4. #### Thomas Snyder

About 3:49. I liked the start (and only missed one signal square I needed) and then just went into a “try to connect as much as possible” mode which worked out. The left/right and top/bottom connections were still nicely executed, but nothing you do is going to prevent the feeling of swimming in extra numbers from coming up when I do a Hitori.

30 May 2010, 17:39

#### Add a comment

You are not allowed to comment on this entry as it has restricted commenting permissions.

## Information

Welcome to the blog of current UKPA sudoku champion, two-time Times national sudoku champion and general logic puzzle fan Tom Collyer.

Home of the original Friday Puzzles, each Friday I publish a 100% original and handmade logic puzzle, inspired by the world-famous Nikoli company.

## May 2010

Mo Tu We Th Fr Sa Su
Apr |  Today  | Jun
1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30
31

## Search

• I like the appearance of the solution. :P Nice one. by Prasanna Seshadri on this entry
• I think I've seen something vaguely similar in some of Palmer's puzzles as well. To be honest I've a… by on this entry
• That's two puzzles in a row where I find something I had used before. I know, not plagiarism; just g… by Bram on this entry
• Kota, that's not really true. I had made my puzzle before going to the WPC already. Also I don't hav… by Bram on this entry
• About 20 minutes, after restarting from a mistaken conclusion. by Bryce Herdt on this entry