Friday, 30 November 2012

Friday Puzzles #185

Round 10 of the WPC was anthology, but I'm going to use this as an excuse to make a Corral puzzle.  Should I persist with these, dearest reader, bearing in mind Corral is apparently a nikoli puzzle, albeit one with no set of rules I can poach from here?

For now, I'll do what I've done before, and slavishly copy from the WPC instructions.  Enjoy!

Rules: (as taken from the WPC instructions)
Draw a single closed loop, going along the gridlines.  The loop cannot touch or cross itself.  The area enclosed by the loop represents a cave.  All numbers inside the grid must be inside the cave.  A number in a grid cell indicates how many cells (including the cell with that number) can be seen from that cell to the nearest wall in four directions, horizontally and vertically.

N.B. For anyone who hasn't solved these before, don't treat this as your average loop puzzle.  Shading in squares which are either inside or outside (but not both!) is how you will solve these puzzles.  In this respect I'd argue corral has more in common with nurikabe and heyawake than, say, slitherlink or masyu.
    #221 Cave a.k.a. Corral – rated medium
All puzzles © Tom Collyer 2009-12.


  1. You could always hope your solvers are polyglots and point to this:

    Of course Nikoli doesn't call these Corral puzzles. Only the USPC/Nick Baxter really seems to.

    1. Thanks Thomas. I must confess to being unaware of the history of the puzzle beyond it always seeming to feature on the USPC. Are there any other names to be aware of? Bag isn't very lyrical, and Corral has the disadvantage of being confused with another similar loop puzzle with paint by numbers style clues. Cave is better than bag in my book, but perhaps it's just me being picky by saying it doesn't quite have that element of je ne sais pas.

      Corral seems to have been the winner as far as I can see online though, as featured on Palmer's/Grant's/Bram's blogs. I've seen you use both Corral and Cave...

  2. Cave was the first name I remember, from the Eger Hungarian WPC. I probably wrote a little more of the history in my "Better Know a Corral" post, but I've begun to adopt Bag more frequently in recent posts and would go Cave ~ Bag >> Corral.

    And to not miss the main course of your post, the puzzle was quite nice.

    1. Do you have any more information about where the name corral came from? It would seem fairly remarkable to me if it were a Baxter invention. I hardly think it's a coincidence that the puzzle known as Koralle on croco share both similar names and solving features, so maybe there is a European influence somewhere along the line? Or else maybe that puzzle came later...

      I suppose another natural question is when did nikoli got involved publishing these? Given the sparsity of "Bag," and the apparent rich heritage of "Cave" from US sources, I wouldn't be surprised if the history ran in parallel to sudoku...

    2. I don't see any similarities between Korralle and Corral/Cave/Bag/Baggu. But there is another puzzle type called "Höhlenrätsel" in German (hence "Höhle" = "cave") which is identical to Corral with the additional rule that no 2x2-areas are allowed inside of the loop.

      Baggu was invented by "Gesaku" and first published in April 1996 (PCN#60). The name in PCN#60 was BAG (indeed in capitalized latin letters!).

  3. Really enjoyed this, thanks! Quite a pretty solution, too.

    1. it's quite a fun solving theme, but one I suspect that would be better served in a slightly bigger grid. Glad you enjoyed my first attempt!

  4. Great puzzle, thanks!

    > In this respect I'd argue corral has more in common with
    > nurikabe and heyawake than, say, slitherlink or masyu.

    In support of your claim, and as suggested (but not worked out) by Grant Fikes, the rules can be stated in a reasonable manner as a black/white puzzle.
    1. Shade some of the squares in the grid.
    2. Numbered squares cannot be shaded. The clue gives the total number of squares visible (including the clue square itself) up to the next shaded square or edge of the board in each direction.
    3. All unshaded squares must be orthogonally connected.
    4. All shaded squares must be orthogonally connected to the edge of the board. (Though not necessarily to one another. If that seems inelegant, one could assume an implied shaded ring around the border of the puzzle.)

    A bit of thought will show that these rules have the same implications as the loop rules for the "diagonal" clues. (That is to say, there cannot be any cross-cuts where two shaded cells touch diagonally but the other two cells are unshaded.) For myself, I find it helpful to think in terms of both descriptions when solving these.


Contact Form


Email *

Message *