Puzzle Description:

Men's Movado Wristwatch

#1: Web Paint-By-Number Robot (webpbn) on Aug 31, 2009

Found to have a unique solution by Gator.

Found to be logically solvable by Gator.

Much easier than I thought it would be when I first started! Once you see the symmetry in it, there's basically just one way to go.

My only question, once I saw all the clues, was... "is this going to slant right to left, or left to right?" Edge logic helped me to know which!

Good puzzle.

Thanks for another entertaining entry!

Wow, what a great image and fun puzzle. I saw the symmetry, and I saw that it would slant. But my brain could not wrap itself around the edge logic, so I made a guess where to start and it solved easily. My left brain is not working well this morning I guess, but my right brain is in awe of the superb image. :-)

I don't like using symmetry (just a quirk) but, after staring at it for a while wondering "How is this not impossible?" I got started with several-rows-deep edge logic.

Yes, deep row edge logic is how I did it. It is technically not symmetrical...just mostly about the diagonal.

Does "deep row edge logic" actually count?

Deep row edge logic sounds like it's close to trial and error - which actually can be a good method. ;-)

I'll try to explain the first step to see if it makes more sense. If you look at row 1 and imagine trying to place the 5 clue in columns 1-5. The column clues this would cross with are 5 6 7 9 5. If you think about the 5 clues on the columns, they will both produce a dot in row 6; however, the clues in columns 2-4 are larger than 5, so R6C2, R6C3, and R6C4 would all be black. This cannot be though since the first clue in row 6 is a 5. So R1C1 will be a dot.

I think we are used to looking at adjacent rows (columns) most often when using edge logic, perhaps because it is easier to see. When I'm doing this kind of edge logic, I just focus on the two rows, in this case rows 1 and 6.

If you really want a challenge on practicing this deep edge logic, I recommend trying puzzle #6175 (Sports Pikto #4).

The edge logic you use to start is a little more complex than normal edge logic, but not really that bad.

I actually started with a dot in column 8 of row 1. The first clue in that column is a 2, while the first clues on either side of it are 1's. It's pretty obvious that if that cell was black, then row two would either have a 1 in it, or blacks separated by whites, and neither is possible. Basically we looked for a contradiction in row 2 of the puzzle.

I then kind of worked out from there. R1C12 is a similar situation to the one before, except this time there is a 3 with 2's on either side, and its easy to see how that can't be black either. So this time we look for a contradiction in row 3 of the puzzle.

I kept chipping away at row 1 in the same way. Looking a candidate positions for the 5 and observing the contradictions that they cause with rows lower down.

So this requires no more look ahead than normal edge logic, but you have to do it several times, checking against different rows of the puzzle.

I think Gator should write a book on advanced puzzle solving. It would be like normal puzzle book, but the puzzles would be organized by solution technique. The first section would be basic line solving, the next edge logic, then fancy edge logic like this, then smile logic, and fancy smile logic, etc. Each section would start with a single sample puzzle on a page where a person could try to solve it. The next page would contain a step-by-step discussion and description of the new technique being introduced and how it can be used to solve the puzzle. The rest of the section would be a collection of puzzles that are further examples of the same technique. There would be two appendix sections at the back of the book. The first would contains hints like "use edge logic in column 1". The second would have solution techniques. The final section of the book would contain puzzles that require multiple solution techniques.

I think that would make a knock-out book.

Sounds good, but do you think anyone would buy it?

PBN puzzle books are apparently good sellers. If you go to the "books" page here you'll find links to many puzzle books on Amazon, some of which seem to be put together by someone who just sat down and made a lot of puzzles. Since I get a steady (though small) stream of income from the page, so people must be buying them. (Last I looked, there was an out of print PBN puzzle book being offered for $150 by someone who claimed it was "rare".) The guys at the WiiPicross site put together a puzzle book for kids. Dunno how their sales have been.

Whooda thunk it?

The guys at HeroGlyphix are also developing a fantasy game based on paint-by-number (picross).

http://www.heroglyphix.com/forum/viewtopic.php?t=342

great image and puzzle.
clever title. I thought that any time machine would be considered a luxury and then I also thought it'd be the Delorian from "Back to the Future". lol

Jan's method in #12 made sense to me and helped me work this one out. Good puzzle - taught me something new.

Jam, could you make an extra category in the find puzzles for types of logic used? This is my first experience with advanced edge logic.

4:35?! I really should be getting to bed!

Fun solve!
Edge logic made easy by Jan's "Webpbnware" that shows red dots next to invalid rows.

Yup, no easy clues here. This is a roadmap:

From the very start, try filling row 1. First five squares true already a contradiction. Can find contradictions right up to C12. This makes R1C16 and R1C17 true, and use line logic. You will fill the squares in at the far right, with easy completion of a good chunk.

Like, the whole puzzle.

Very easy once I guessed one of the 10s.

Goto next topic

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