Web Paint-by-Number Forum
Comments on Puzzle #10873: Storm Tossed
By Joe (infrapinklizzard)

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

Puzzle Description Suppressed:Click below to view spoilers

#1: Joe (infrapinklizzard) on Oct 30, 2010 [SPOILER]

Comment Suppressed:Click below to view spoilers
#2: Joe (infrapinklizzard) on Oct 30, 2010 [HINT]
Edge logic is necessary for this one.

And again.

Once more.

Home free!


===================
Specifically,


After line logic,
Edge logic on the 4 in r15.


The 1|2|1 in c1-3 means that c1-2 must be white (or there would be a one-long block in r14).

==

Then, immediately, edge logic on the 4 in r14.


The 2|1 in c2-3 means that c2 must be white (or there would be a one-long block in r13). Then a little line logic.

==

Then edge logic on the 4 in r13.


If the 4 in r13 goes all the way to the left (c4-7), the 1 in c3 will force the 4 in r14 into c6. However, c6r13 is white (due to our initial statement), so it would make a maximum two-block in c6 which conflicts with the 4.

So the 4 in r13 cannot go all the way to the left and r13c4 must be white.

The rest solves with line logic.

#3: bugaboo (bugaboo) on Oct 30, 2010
row 18 infrapinklizzard?
better recheck your comments
haha

fun solve
#4: Joe (infrapinklizzard) on Oct 30, 2010
Fixed. Not used to these tiny ones...
#5: Tom O'Connell (sensei69) on Oct 31, 2010
was a toughie, Joe, but i got it :)
#6: ant (agrest272) on Oct 31, 2010 [SPOILER]
Comment Suppressed:Click below to view spoilers
#7: Byrdie (byrdie) on Oct 31, 2010 [HINT]
I used the edge logic Joe describes to get started but where to place the bottom 4, for me, borders on trial and error.

Key words ... "for me." I tend to side with Ant.

I guess the question is, if a puzzle only has one solution does that still mean it is logically solvable?
#8: Paul Zrimsek (pzrimsek) on Oct 31, 2010 [HINT]
Solvable without guessing or excessive look-ahead. Once you've gotten as far as you can using simple line logic, you can place the lowest '1' in C4 using two-way logic. After that you can get the rest of it with a few more uses of two-way logic. A fun one!
#9: Joe (infrapinklizzard) on Oct 31, 2010
Martin, "logic" broadly references any particular method of reasoning. There are many specific definitions of "logic". It means different things to mathematicians, philosophers and computer scientists.

Here, the consensus has fallen to the definition that a given solution is logical if it can be arrived at by someone "working in pen". In other words, any looking ahead must fall within a reasonable amount, say one really tricky move or a few simple ones.

Edge logic really has no more looking ahead than basic line logic; it merely encompasses another line.

Line logic says "if I put the <clue> all the way to the left, it will take up <these spaces>. If I put it all the way to the right, it will take up <those spaces>. Any overlapping spaces are <color>."

Edge logic says, "if I put the <clue> all the way to the left, it will cross <these vertical clues>. These vertical clues will|won't cause a problem in row x."

#10: Byrdie (byrdie) on Oct 31, 2010
You must either think I'm really stupid or really dense.

Line logic is elemental.

Edge logic, as I understand it (or more reasonably, as I practice it), is more elimination than look ahead.

When elimination requires looking 3 or 4 columns or rows in advance *****TO ME***** it seems more to be trial and error, even if the net result is only 1 possible solution.

I accept that there is a consensus of those evidently smarter than I am that it's ok to have to look deeper. It may even be logical. It doesn't meet my definition of being logically solvable and when my opinion is asked, as when I'm offered the drop down, I will continue to mark it as such.
#11: Joe (infrapinklizzard) on Oct 31, 2010
Being someone who does not appreciate being condescended to, I try not to do it to others. I was merely trying to explain (badly, it seems) the way I regard edge logic.

The way I see it is that the entirety of what we do here is trial and error, including line logic. We cannot place anything without saying, "if it goes here, then...".

If placing clues with line logic is a positive way to locate the black spaces, then elimination-types of logic are often ways to positively place white spaces. This can be an effective way to move forward.

The trial-and-error on line logic is trivial. The t&e on edge logic is quite easily kept in one's head. It's when the chain of "if-then"s gets too unwieldy to remember (without marking the puzzle) that it becomes "not logical" *for this site*.
#12: Teresa K (fasstar) on Nov 1, 2010
Martin, I get what you are saying. I think I have above average solving skills and I really enjoy using edge logic. But the logic required for puzzles like this is really beyond my ability. There are a handful of solvers who have frontal lobe connections for working memory that is way WAY above average. Maybe THEY can do it in their head, but I can't, and so "for me" it comes down to guessing and eventually finding out if my guess was right.

Joe, thanks for a clear explanation of what "logical" means here at webpbn. I was trying to explain it to my friend, and was just getting her more confused I think. I'll show her your comment.
#13: Teresa K (fasstar) on Nov 1, 2010
Well, after looking at your hints, Joe, I was able to solve it without guessing. But it really stretched my brain. That's my kind of exercise - mental yoga. :-)

Your hint still says R18. Maybe it need to be fixed in two places?
#14: ant (agrest272) on Nov 1, 2010 [SPOILER]
Comment Suppressed:Click below to view spoilers
#15: Joe (infrapinklizzard) on Nov 1, 2010
Thanks, Teresa. It actually said r18 in many places (like every mention of r13), but when I fixed it the first time I missed that last one. Fixed again.

Thanks, ant. I'm glad you came back and tried again!

Paul, can you explain how you judged where to place that 1 in c4? I can't see it.

#16: Paul Zrimsek (pzrimsek) on Nov 1, 2010 [HINT]
At the point where line logic alone leaves you, you can count and see that the lowest 1 in c4 must go in either r10 or r12. If it's r12, then r13c4 must be a dot and the 4 in that row must go in c5-c8. But that would mean the 6 in c8 has to poke up through r12-- and r12c4 and r12c8 can't both be black. Therefore the lowest 1 in c4 has to be in r10. Now that I look at it again, you don't need anything more than line logic after that.
#17: Joe (infrapinklizzard) on Nov 1, 2010
Ah, yes, very good!
#18: Web Paint-By-Number Robot (webpbn) on Nov 2, 2010
Found to be logically solvable by Gator.
#19: Gator (Gator) on Nov 2, 2010 [HINT]
I solved it a little differently.

After the initial line logic, I dotted R15C1 and R15C2 the same way Joe did.

I then dotted R14C2 by looking at the 2 clue in column 2.

I then looked at the 6 clue in column 8. Using edge logic and seeing how column 4 is affected, R14C8 and R15C8 would have to be dots, otherwise there would be no way to fill in the 2 clue in column 4.

It solved normally from there.
#20: Jota (jota) on Nov 2, 2010
I Like it!
#21: Kurt Kowalczyk (bahabro) on Aug 2, 2014
fun lil puzz, Joe! nice image too :)
#22: Joe (infrapinklizzard) on Aug 2, 2014
Thanks, guys

Show: Spoilers

Goto next topic

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