Web Paint-by-Number Forum
Comments on Puzzle #5590: Think Pink
By MMD (mmd)

peek at solution       solve puzzle
  quality:   difficulty:   solvability: moderate lookahead  

Puzzle Description Suppressed:Click below to view spoilers

#1: Synthia McBride (synthia) on Apr 11, 2009

Very nice. For a long time I thought it was a lobster. Couldn't figure out what "pink" had to do with a lobster.
#2: Adam Nielson (monkeyboy) on Apr 11, 2009
Lovely image. I hadn't gotten to this one yet.
#3: Teresa K (fasstar) on Apr 13, 2009
Lots of fun. I love the image. It was good even with the guessing.
#4: Gator (Gator) on Sep 9, 2009 [HINT]
Once all of the line solving is done, look at row 2. There are 12 spaces open on either side of the dots. A "3 4 4" set of clues would require 13 spaces. So we will have "3 4" on the left and "4 3" on the right. Let's focus on the left side. The 4 clue can start at column 6. If you try to place the 4 clue in columns 6 - 9, this will cause row 3 to be invalid. So we know the 4 clue can now only go in columns 7-13. This allows us to fill in R2C10 (and likewise R2C20 by working the right side with the same logic). Using the same kind of logic, you can see that the 3 clue can only go in columns 2-6. So R2C4 is a black (and R2C26).

Use line logic to solve more of the puzzle.

When you get stuck again, focus on the left side again and look at the 3 clue in row 1 and the 4 clue in row 2. Notice that no matter where you place the 3 clue in row 1, R2C8 has to be a dot (and R2C22). This allows R2C12 and R2C18 to be black.

Now look at the 7 clue in row 5. It can go in columns 10-20. If you try to place it in columns 10-16, this will make it impossible to fill in row 4. So R5C17 will be a black. Using the same logic on the right side, R5C13 will also be black.

More line logic.

Look at row 1 again on the left side. If the 3 clue goes in columns 8-10, this will cause row 3 to be invalid. So R1C8 is a dot (and R1C22).

Line logic will finish out the puzzle.

So, logically solvable, or does this go too far? I want opinions on this, because I'm not entirely sure.
#5: Arduinna (arduinna) on Sep 9, 2009 [HINT] [SPOILER]
Comment Suppressed:Click below to view spoilers
#6: Byrdie (byrdie) on Sep 10, 2009 [HINT] [SPOILER]
Comment Suppressed:Click below to view spoilers
#7: Robyn Broyles (ginkgo100) on Sep 10, 2009 [HINT] [SPOILER]
Comment Suppressed:Click below to view spoilers
#8: Adam Nielson (monkeyboy) on Sep 10, 2009
"My" method, Robyn???
#9: Jan Wolter (jan) on Sep 11, 2009
I got the first couple of steps that Gator took, but didn't make it the rest of the way. However each of his steps is perfectly clear once it is pointed out.

Hmm...the whole point of the rating system is to help people find puzzles that are of an appropriate difficulty level for them. This puzzle is fine if (1) you are fond of straining your brain over really hard logic puzzles, or (2) you are fond of pretty pictures and don't mind guessing based on "picture logic" when the going gets tough.

The problem is that the difficulty ratings assigned by those two sets of people just get averaged together. So if we tag this as "logically solvable" which it technically is, then it shows up in the ratings as a moderately difficult logically solvable puzzle, which it isn't.

Maybe I should have two difficulty ratings for each puzzle, how hard it is to solve logically, and how hard it is to guess. But I think that making the rating system too complex just makes it more confusing and doesn't end up being that helpful.

I can imagine 50x50 puzzles that can be solved only by performing a sequence of hundreds of subtle moves like the ones Gator describes above. Assuming some madman actually worked out the sequence and posted it, would we then rate that puzzle as logically solvable?

I guess we probably would.

So I guess the only consistent thing to do is to say that it is logically solvable, and I guess it's better to be consistent.

Maybe the best fix is if I add more ratings to the the Solvability scale:

The "line & color" category would always be flagged by the checker. The distinction between the last two would be subjective.
#10: Web Paint-By-Number Robot (webpbn) on Sep 11, 2009
Found to be logically solvable by jan.
#11: Robyn Broyles (ginkgo100) on Sep 11, 2009
Sorry, Adam. Typo. I meant Arduinna's method.
#12: Jota (jota) on Sep 13, 2009
I love it!
#13: Joe (infrapinklizzard) on Feb 26, 2010 [HINT]
I was able to solve it without guessing. It was tricky around the eyes, but it seemed fairly obvious to me that the key to attack was row 2 where the 3,4,4,3 had to be split between the two sides. Edge logic as per Gator above was how I continued.

Show: Spoilers

Goto next topic

You must register and log in to be able to participate in this discussion.