Web Paint-by-Number Forum
Comments on Puzzle #9066: Trollin', trollin', trollin'...
By Alison Deem (Indigo)

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

Puzzle Description Suppressed:Click below to view spoilers

#1: Jane Doe (telly) on Jun 5, 2010

cute but I think it takes some guessing.
#2: bugaboo (bugaboo) on Jun 5, 2010
image or symmetry logic
maybe double smile logic
either way i agree with telly
#3: Joe (infrapinklizzard) on Jun 6, 2010 [HINT] [SPOILER]
Comment Suppressed:Click below to view spoilers
#4: Tom O'Connell (sensei69) on Jun 6, 2010
Thanks Alison,i agree with bugaboo
#5: Web Paint-By-Number Robot (webpbn) on Jun 7, 2010
Found to require some guessing by Gator.
#6: Teresa K (fasstar) on Jun 8, 2010 [SPOILER]
Comment Suppressed:Click below to view spoilers
#7: Alison Deem (Indigo) on Jun 10, 2010
Yeah, mine actually has yellow hair but, alas,.....
#8: John Helmsen (netheron) on May 7, 2011 [HINT] [SPOILER]
Comment Suppressed:Click below to view spoilers
#9: Joe (infrapinklizzard) on May 8, 2011 [HINT]
John, if the solvability has already been "determined", you need to provide a *detailed* explanation of why you think that determination might be wrong.

Not only is this explanation not sufficient, but also both of those pixels can be filled in with line logic (but one is black, the other white). Perhaps you meant c5r16 and c6r16, but again one is black, the other white.

So, huh?
#10: Gator (Gator) on May 9, 2011 [HINT]
OK, how about this?

Looking at column 5, the 2 clue can be placed in 3 different places without making column 6 invalid. In each of these cases, both R16C6 and R17C6 will be dots. The same logic can be applied on column 16 to make R16C15 and R17C15 dots.

Next look at column 8. The 2 clue can be placed in R15C8-R16C8 (if R16C8 is part of the 2 clue), R16C8-R17C8, R18C8-R19C8, or R19C8-R20C8 (placing it in R17C8-R18C8 makes column 9 invalid). In all of those cases, R20C9 will always be black. The same logic can be done with column 13 to make R20C12 black.

After more line logic, smile logic in rows 15/16 to finish.

I looked at most two moves ahead for each instance of possibility testing, so this works. Agree?
#11: Joe (infrapinklizzard) on May 10, 2011 [HINT]
I both agree and don't.

The logic works and fits within *my* head (not that I think I'd have found it), but...

The amount of lookahead is a very sticky part of our logic definition. Sometimes this sort of thing is easier than others. For instance, following a zig-zag to its end to see if the other end will conflict is often trivial even if it goes for many iterations. Even though there may be many steps, all you really need to keep track of is your position.

In this case, however, you need to keep track of a lot more. While each partial step of the c8 logic is within the logic standard, you need to keep the positions of the pixels in c9 in your head throughout the four individual steps. Thus it is really only two steps each if you were marking r9 in between (which would count as guessing).

While the letter of the law is a two-step lookahead, the spirit is "what a normal person can keep in their head." So, I think this is less "wpbnlogical" than zig-zag logic.



A nitpick: in your first step, edge logic on the 2 in c5 would put it in only two places; r16-17 or r19-20. If it were to overlap r18 from either direction, it'd force the 2 in c6 down to r20 which then would back up into c5 causing a conflict. This is actually fewer steps to keep in your head than the c8 logic (but really doesn't get you any further ahead).
#12: Gator (Gator) on May 10, 2011 [HINT]
Yeah, that's why I didn't go ahead and mark this as solvable (yet). You have to think about how all three 1 clues will fit in column 9 from what you are doing on column 8. It gets to the point where you almost have to write down the possibilities.

Diagonals (zig-zag) are "easy" as you are able to basically throw away information as you proceed down the length of it. You just need the last few moves to tell you where to go from there. You also have to keep your starting position in your head as it will have to be marked as a dot if a contradiction is found.

I think I will leave this one as guessing as I believe it does break the spirit of "what a normal person can keep in their head". These puzzles are still fun to analyze I have to say though. :)
#13: Joe (infrapinklizzard) on Nov 23, 2018
I just ran across this one again. We have a new classification since then, and since a method of solving has been put forward this will be changed to deep-lookahead.
#14: Web Paint-By-Number Robot (webpbn) on Nov 23, 2018
Found to be solvable with deep lookahead by infrapinklizzard.
#15: John Macdonald (perlwolf) on Jan 10, 2019
The title for this puzzle made me think of the theme song for the old TV show "Rawhide".

Rollin', rollin', rollin,
Keep those dogies rollin
Rawhide.
#16: Kristen Vognild (kristen) on Jan 10, 2019
Ever hear the Scottish version?
https://www.youtube.com/watch?v=GXIkL82hJtg
#17: John Macdonald (perlwolf) on Jan 11, 2019
Double post - ignore.
#18: John Macdonald (perlwolf) on Jan 11, 2019
That's great - I'd never heard it before.
#19: Glenn Crider (playamonkey) on Mar 14, 2021
I found it relatively easy using logic to fill in the blue first, then symmetry logic to complete the black.

Show: Spoilers

Goto next topic

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