16
$\begingroup$

As any member of the PSE community knows, grid-deduction puzzles really aren't my forte.

In fact, I'd like to share one particularly disasterous attempt at designing a Slitherlink puzzle.


Disclaimer: This is a fictional story

So, I was working on a slitherlink late one night, and when I woke up, I realised that my sleepiness had caused me to put some wrong numbers into the puzzle! I had only miscounted these clues by one, but they were wrong nonetheless.
I took a sip of coffee to wake me up, and then I noticed something interesting about the way I messed up the puzzle. I had gotten exactly one clue in every row and column misnumbered by one!
In my excitement in thinking that the puzzle was salvagable, I stood up, and accidentally tipped my coffee onto the puzzle!!!
This rendered some clues unreadable, and I couldn't remember if some of the now unreadable clues were the ones that were wrong or not.

After finishing my coffee, I realised that this was still a solveable puzzle.

enter image description here

There's my slitherlink puzzle, after being misnumbered AND having some clues (possibly correct, possibly misnumbered) destroyed by my morning coffee.

Can you find the intended answer to the slitherlink?

$\endgroup$
  • $\begingroup$ Looks fiendish... $\endgroup$ – Wen1now Feb 28 '17 at 10:40
  • $\begingroup$ Your coffee is remarkably selective in what it destroyed. What brand is that, I might want to try it :) $\endgroup$ – Rubio Feb 28 '17 at 11:04
  • 1
    $\begingroup$ This is a very nice one! I hope you will post a 'making-of' post to see how you constructed it. $\endgroup$ – elias Feb 28 '17 at 11:22
  • $\begingroup$ This is indeed a really great puzzle, I can only wonder what other new variations on grid puzzles you have in store. $\endgroup$ – w l Feb 28 '17 at 14:02
  • $\begingroup$ @wl I'm glad you enjoyed it! I'm very busy (aren't we all, all the time?) but sometimes I have an idea I just really want to see come to life, and seeing a solver have a good time with a puzzle makes the work worth it :) when I next make a new grid puzzle, PSE will be the first to know :) $\endgroup$ – TheGreatEscaper Feb 28 '17 at 14:38
15
$\begingroup$

The solved grid looks like this:

The instructions assume a familiarity with solving normal Sliterlink puzzles. A dot next to a number means the clue is correct, a plus means the correct clue is higher, a minus means the correct clue is lower.

We start in the bottom left corner because a 0 and 3 can not be next to each other on the edge of the grid. This means the 1 and 3 in the same column have to be correct clues.

If the 0 in the top row were correct it would lead to a loop, so it has to be a 1.

This also means the first 3 in the third column has to be correct which means the 2 next to it can not be correct.

The link can only go straight through that 2x2 area. If it would go horizontal, the top link would loop again.

The 2 and 0 next to each other in the third row can not both be correct, so the last 0 in that row has to be.

The 2 above that 0 can not be correct with the 3 around.

Next we take a look at the 2 in the third row again. If that clue were correct the link would have to loop again.

The 1 in the fourth column has to be correct, but the link also has to go through the bottom left corner. This needs 2 horizontal links in the three remaining possibilities leading to a link at the very bottom.

The 2 and 3 in the fifth row have to be either 1 and 3 or 2 and 2. The clues 1 and 3 leave a dead end, so it has to be 2 and 2.

This leaves only one clue to be incorrect in the last row.

Noteworthy: The coffee only destroyed correct clues.

$\endgroup$
  • $\begingroup$ Nice! Also noteworthy: the incorrect clues have rotational symmetry :) $\endgroup$ – TheGreatEscaper Feb 28 '17 at 12:18
  • $\begingroup$ I'm quite pleased that your solving method is exactly the route I intended! $\endgroup$ – TheGreatEscaper Feb 28 '17 at 14:39
  • $\begingroup$ At first I used a slightly different route (7th step before 5th and 6th), but while creating the images I noticed that this route was actually easier. $\endgroup$ – w l Feb 28 '17 at 14:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.