August 19, 2011

Friday Puzzles #118

So I’m very glad to be able to say that once more I’m a contributor to Gareth Moore’s Sudoku Xtra magazine! If you aren;t aware of this little gem, then permit me a moment to tell you all about it! The Sudoku part of the title should tell you that a lot of the puzzles are indeed Sudoku related, but there are plenty of Sudoku variants included to keep every enthusiast’s appetite satisfied. A personal favourite of mine is odd-pair Sudoku, which I believe is Gareth’s own invention.

However, those who are underwhelmed by the whole Sudoku craze should still pay attention as the Xtra part of the title holds equal prominence. There are several nikoli-style puzzles in each issue, along with some other WPC favourites. And not to forget arguably the most interesting bit, the community section. (Of course I can’t claim to be totally impartial here!) This features a really rich and diverse collection of logic puzzles sourced from professionals, amateurs and experts alike.

OK, that’s quite enough plugging for now! This week’s puzzle is another Suraromu, inspired the puzzle that appeared in the latest issue of Sudoku Xtra, which was about as much fun I’ve had with any logic puzzle I’ve done in a while. I’m not sure this one rises to those standards, but It’s certainly non-trivial. Actually, Suraromu as a puzzle type intrigues me somewhat because much like Numberlink, as the difficulty of a given puzzle rises, so does that puzzle’s susceptibility to uniqueness strategies. This is certainly true of this puzzle, but unlike with numberlink I think I’d suggest to the puzzler wanting most satisfaction out of the puzzle to resist temptation this time around. And I think I’ve babbled on for long enough now, so all that remains to be said is, Enjoy!

#148 Suraromu – rated medium

[blockquote mark-up still apparently not working]

P.S. I should remark that I’m not entirely happy with the presentation of this puzzle. There are some squares which share gates, and I thought that during the test solving this might be a little misleading, even if it is at least implicit that every numbered gate has both of its “posts” labelled. My solution previously had always been to employ a convention that there should be no squares sharing gates, but I suppose I got a bit lazy this week. I’d be interested to know people’s thoughts on this one…


- 9 comments by 1 or more people Not publicly viewable

[Skip to the latest comment]
  1. Warren

    The shared posts certainly didn’t detract from the enjoyment of the puzzle. I might have to include a book of one of these on my next Nikoli order after you’ve now got me into them!

    19 Aug 2011, 12:39

  2. TheSubro

    Nice puzzle, very Numberlink-ish due to the singularity issues. Probably would not have focused on that aspect as much without your intro mention of it, but allowed for a much quicker solve. I definitely enjoy these, but have a bad taste in my mouth about them due to the mistakes I made trying to solve the great one you constructed for your recent LMI test. Keep em coming. Its all good.

    TheSubro

    19 Aug 2011, 21:07

  3. Ours brun

    Nice puzzle. I find this puzzle type pretty easy to deal with, which is quite fun since I am totally ridiculous at numberlinks.

    Regarding the presentation, I must say that I prefer the one used by Palmer Mebane. Much less confusing.

    21 Aug 2011, 11:11

  4. Agree that Palmer’s notation is very functional and works well, but doesn’t look quite right to me from an aesthetic point of view. I was vaguely toying with the idea of splitting the cells down the diagonal, a la kakuro clues, but didn’t want to do the extra fiddling around in inkscape. I’m not totally convinced that’ll look 100% right either, but I might give it a go for the next puzzle.

    I must say, I’m generally a fan of loop puzzles in general, but suraromu is fast becoming one of my favourites. They aren’t so much work to produce relatively interesting puzzles either, which is a bit of a bonus!

    22 Aug 2011, 13:52

  5. David

    Suraromu rocks.

    For me, the only real problems with shared gates are (1) when two shared gates both need to be numbered, and (2) when a numbered gate only has one side marked because the other side is the edge of the grid.

    I first read your comment as though you were thinking about having diagonal gates, which actually seems like it would be an interesting variant - it would allow a cell to be passed through horizontally OR vertically - but I’m not sure how it would work in actuality.

    24 Aug 2011, 01:16

  6. furudo.erika

    Suraromu is a very underated puzzle type IMO and judging by the couple you have here I’d like to see a lot more from you.

    Regarding the presentation, I’d say if it’s good enough for Nikoli it’s good enough for everyone. Generally you have the cleanest presentation I’ve seen but you should lose the grid lines between the black squares; they’re just ugly.

    24 Aug 2011, 18:07

  7. Rohan Rao

    Very nice puzzle. Not as good as the one in Nikoli2 IMO, but well-made. I’m not sure if there are many ‘solving approaches’ for this puzzle type and the solving path might tend to get repetitive.
    Looking forward for more…

    24 Aug 2011, 18:50

  8. MellowMelon

    Allow me to provide a counterpoint: I do not like Suraromu as a type anymore. This comes from the perspective of someone who generally looks for new, innovative logical steps to employ in types. The one here and the one from the LMI test were definitely nice puzzles, but just as a consequence of having so much space they suffer from plenty of uniqueness issues too. As a result neither took me any time beyond putting lines down.

    In line with what I said above about new logic, http://mellowmelon.wordpress.com/2010/04/03/puzzle-234/ is probably my preferred example of the ones I have made. Even then, and as I expected would happen, a commenter missed the new step and slummed through with uniqueness logic (no fault to them, the type is to blame). As the design of the type goes, Suraromu suffers from fundamental problems most of which relate to having too little constraints available to the constructor. The loop is not allowed to do any wiggling for the sake of uniqueness, and yet the only real conditions on it are gates, whose requirements to be straight lines bounded between two walls make them rather difficult to put down in places they’re needed, and where every one you add makes the puzzle substantially easier. I’ve run into very few instances of this type that take me any thinking if I allow myself uniqueness logic, many of which are my own constructions. And I don’t think it has anywhere near the charm or challenge of Numberlink to make this excusable.

    I came up with a variation awhile ago in attempt to save the type, although I have yet to get anything I’d be willing to post. My tweaks are (a) to add a new kind of gate that is not required to be passed through, with the circled number telling the total number of gates (required or not) the loop uses, and (b) to relax uniqueness on the loop, so that the puzzle is solved as soon as you find which gates to pass through and the order and construct at least one loop that does so.

    26 Aug 2011, 17:24

  9. So I liked that puzzle, and smiled to myself as I realised that what I was trying to ignore as something forced by uniqueness was forced anyway. I must confess to not being the hugest of numberlink fans (although the whole thing with trying to “prove” uniqueness does still intrigue me), but I sort of see Suraromu as being basically consecutive numberlink. I like that having blacked out cells provides an extra constraint. I like that the constraints that the gates provide, and I find it interesting you view these as being quite a weak constraint.

    I suppose I’ve taken quite a meta-approach to construction with the puzzles I’ve constructed recently, in that I will vaguely sketch out solutions and tweak them, adding in the gates as I go along, making sure to erase any possibility of patterns that will give me uniqueness issues. From my point of view, the gates suddenly become quite strongly constraining, particularly because they force parallel strands to diverge.

    I think there are plenty of possibilities for novel tricks here (certainly more than there are with numberlink) – and Gareth’s example I was originally talking about is a wonderful case in point. Although that brings me back to my other original point – the susceptibility to uniqueness. I’m sure you can do exactly the same thing to Gareth’s puzzle. I guess what I don’t understand is why you’d single out numberlink as being any different from Suraromu here. From that particular point of view, all Suraromu does is make things easier by providing additional constraints to fiddle out a solution.

    I’ll be keeping an eye out for your variation. I’ll also try to cook up a puzzle that can’t be so naively approached…I have a few ideas already…

    26 Aug 2011, 21:46


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.


How to play:
AkariFillominoHashiHeyawakeHitoriKakuroLITSMasyuNumberlinkNurikabeRipple EffectShikakuSlitherlinkSudokuSuraromuYajilin.

Tags

August 2011

Mo Tu We Th Fr Sa Su
Jul |  Today  | Sep
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

Latest comments

  • 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

Copyright Statement

Puzzles published on this blog are licensed under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 Unported License. Please leave a comment or send an email for more information.

Creative Commons License
Not signed in
Sign in

Powered by BlogBuilder
© MMXIX