home
puzzles
genres
studio
help
about
links
forum


Forums
  [Search] Search   [Recent Topics] Recent Topics   [Members]  Member Listing   [Groups] Back to home page 
Crossings Puzzle #2205 by Maarten
Forum Index -> Puzzle Feedback
Author Message
Sesquideus



Joined: 04/04/2014 16:53:34
Messages: 28
Offline

I think the solution is not consistent with puzzle #435 in its interpretation of the rules (number of crossings in the row vs in the nearest section of the row).
HappyDog



Joined: 11/11/2015 18:26:16
Messages: 152
Online

(I think you meant puzzle #436.)


Good spot! In fact, the rules don't specify how this should be handled either way, but you are correct that these two puzzles interpret a gap in the grid in different ways.

Not sure what can be done about it at this stage, though! Also, not sure how widespread the issue is (i.e. how many puzzles use each of the two interpretations).


For quick reference, here are direct links to the puzzles:

* http://puzzlepicnic.com/puzzle?2205
* http://puzzlepicnic.com/puzzle?436

HappyDog



Joined: 11/11/2015 18:26:16
Messages: 152
Online

I've just retried #2205, and moving the '1' to the left, to the right of the first contiguous block (i.e. into the duck's neck), would not affect the solving but would make the interpretation consistent.

If this is the only puzzle with this interpretation, then I would suggest that as a fix. However, it may be more complicated if there are others that also have this issue - they may not all be resolvable so simply.
fl0rrat



Joined: 10/10/2010 20:04:05
Messages: 236
Offline

Well-spotted!

@HappyDog: Do you mean to place the 1 (in #2205) to a place where it is adjacent to three cells of the puzzle? In that case, isn't it ambiguous whether it means the number of crossings to the left, below or above?

Another possibility is to make the cells in the row with the 1 (in #2205) black, and specify in the rules that numbers point to the row/column of contiguous black&white cells.

(That said, as a judge I don't have the ability to change any puzzles, the only thing I can do is to take a puzzle down and ask the author to change it)
HappyDog



Joined: 11/11/2015 18:26:16
Messages: 152
Online

You're right - my suggested location would imply that it potentially operates in an up and down direction as well.

The only other place would be to the far left of the puzzle, on that row, but without anything above or below it.

It would be a bit ugly, but would be technically correct and consistent.

Not sure about your suggestion about black squares - how do they currently work in existing puzzles? Are they equivalent to reaching the edge of the grid or do they function the same as white squares in this regard?

If all existing puzzles treat black squares as being part of the grid then I agree that this would be solution. As you say, it should be added to the rules to make this explicit.

To avoid it being too ugly, it might be worth surrounding the whole duck in black, rather than having a couple of random squares on its back. However, even with that it won't be as pleasing as the original layout.

And, of course, if there are puzzles that use black squares as hard-blocks, then this won't be possible, so that would need checking, first. There are only 27 so it shouldn't take too long...
HappyDog



Joined: 11/11/2015 18:26:16
Messages: 152
Online

I have done a quick check to compile a list of all puzzles that may have an impact on deciding how to handle this situation. I have not checked whether they support one or other interpretation of the rules, just located the puzzles that may be affected by the interpretation.

Puzzles where a number is in the same row/column as a black square:
* http://puzzlepicnic.com/puzzle?593
* http://puzzlepicnic.com/puzzle?2205
* http://puzzlepicnic.com/puzzle?1036
* http://puzzlepicnic.com/puzzle?2600
* http://puzzlepicnic.com/puzzle?2799

Puzzles where a number is in the same row/column as a gap in the grid:
* http://puzzlepicnic.com/puzzle?1439
* http://puzzlepicnic.com/puzzle?446
* http://puzzlepicnic.com/puzzle?436
* http://puzzlepicnic.com/puzzle?2205
* http://puzzlepicnic.com/puzzle?506
* http://puzzlepicnic.com/puzzle?598
* http://puzzlepicnic.com/puzzle?3770

However, my investigation has also thrown up another issue with this genre, relating to whether or not a number applies to a row or column:

* In some puzzles (such as http://puzzlepicnic.com/puzzle?506 or http://puzzlepicnic.com/puzzle?598) the number only applies to the row/column if it is directly adjacent to the grid. It does not apply if there is a gap between the number and the grid.
* In some puzzles (such as the one we are talking about here, and, crucially, http://puzzlepicnic.com/puzzle?1439) the number applies if it is aligned with the row/column, even if it is not directly adjacent to it.
Sesquideus



Joined: 04/04/2014 16:53:34
Messages: 28
Offline

HappyDog wrote:
(I think you meant puzzle #436.) 


Of course, my bad.

I guess a "clean" solution would be to introduce diagonal clues, as in kakuro. But I do not know how difficult it would be to implement, and half empty they do not look very nice.
HappyDog



Joined: 11/11/2015 18:26:16
Messages: 152
Online

One thing I've thought would be a good addition to this site would be a "Author's notes" field for all puzzles, where the author could optionally add any additional notes relating to the puzzle.

This originally came up in the context of competition entries (e.g. "This puzzle was created for Competition X"), linked puzzles (to refer to the others in the series) or special occasions (e.g. "Published on Valentine's day").

However, it could also be used to provide any special variations to the rules, and that might be a simple way to get round this specific issue, i.e. to give notes about rule variants or specific interpretations of the rules that may be necessary for solvers.
 
Forum Index -> Puzzle Feedback
Go to:   
Powered by JForum 2.1.6 © JForum Team