You are viewing motris

 
 
07 January 2011 @ 12:30 am
Friday Puzzle #83 - Mystery Puzzle  
One of my favorite challenges in the world of puzzles is to try to solve an unfamiliar puzzle in a foreign magazine (most typically in Dutch or Japanese) where not being able to understand the instructions is half the battle. Looking at the example picture, or more often at one of the solutions, lets you try to divine the rules without any real knowledge of the language.

Instruction-less puzzles have been used a few times at the World Puzzle Championship, and the biggest hurdle is always to have a clear enough example so that the solver can actually figure out what is on your mind. This often means "simple rules", but not always. This week, I offer my own attempt at a mystery puzzle type with no written instructions. Enjoy.

Example:


Puzzles:




ETA: Circle in R2C1 of the example, as noticed by zotmeister.
 
 
 
( 29 comments — Leave a comment )
David MillarDavid Millar on January 7th, 2011 07:59 am (UTC)
I think I've figured out what you're looking for here.

I recently found a Japanese puzzle book I had years ago and decided to give many of the puzzles another go. Sometimes I sneak a peek at the answer keys to make sure I'm on the right track, so your challenge is a bit tougher - on the other hand, knowing that it's intentionally instruction-less kind of gives the solver the knowledge that it can be solved without instructions because it was made that way.

Overall good first puzzle, and I'll be starting the second soon.
Adam R. Woodzotmeister on January 7th, 2011 09:07 am (UTC)
Grrr - I see a pattern that ONE cell of the example violates. Perhaps that was intentional, but I have to ask: is there a little circle missing from R2C1? (I know where in the cell it would be if it were there.) I'm working on the first puzzle now, and I may be able to solve it either way, but it would go MUCH easier if I'm right. - ZM
motrismotris on January 7th, 2011 03:55 pm (UTC)
There should be a circle in R2C1. I've amended the example. Apologies for the error.

Edited at 2011-01-07 04:04 pm (UTC)
Adam R. Wood: butasanzotmeister on January 7th, 2011 04:34 pm (UTC)
Armed with the additional rule gleaned from that, I found a uniquely deducible solution for the first puzzle in short order. It was looking like the puzzle might have had multiple solutions without that rule, but I haven't actually confirmed or denied that. What I did confirm is that the example is still uniquely solvable without that rule, which arguably makes it flawed in a deeper sense for an instructionless puzzle, but given the rule is pretty much needed for the first puzzle to be reasonable (if not uniquely solvable), it can be handwaved. After all, I solved it, so I can't cry foul ;)

Of course, the most important thing was that the puzzle was a lot of fun :) Thank you - it was an excellent cure for my insomnia.

I'm not sure if it should please or disturb me that I can tell something's wrong when I'm not even told what right is... even at 4:30 AM...

This seems to be Instructionless Puzzle Day: I have someone trying to guess the rules to my latest puzzle - something it was never designed for, but it's tempting to wait and see what they come up with... - ZM
motrismotris on January 7th, 2011 04:37 pm (UTC)
I read the first response to the one you put up, and agree with a lot of the statements. Not sure on the meaning of the loops yet but I don't have time to give it a solid look at the moment. (It was my rush to finish this involved design last night that probably led to the small example hiccup. I triple-checked the puzzles themselves but missed one circle on the sample).
Adam R. Woodzotmeister on January 7th, 2011 05:11 pm (UTC)
Not to nitpick, but that circle is still missing from the example solution. I wouldn't have mentioned it, but I realized someone might (erroneously) interpret that as meaning deletion of the circles might be part of the solving process.

Another seems to have joined in on the guessing game, so I'll start throwing them some bones... - ZM
(Anonymous) on January 7th, 2011 12:07 pm (UTC)
Hrm. Figured out an interpretation of colour and circles that makes sense in the sample, but doesn't allow me to solve it, and contradicts the first non-sample puzzle...

Rob
motrismotris on January 7th, 2011 04:22 pm (UTC)
Both puzzles are actually a bit hard, certainly relative to the example, so once you have a set of rules see that it does solve the sample to give a unique solution. If not, you may still be missing some part of the puzzle.
projectyl: puzzlesprojectyl on January 7th, 2011 05:12 pm (UTC)
There's an odd red herring in the sample: Gur ahzore gjb arire nccrnef va n oyhr ertvba, naq ab ahzore bgure guna gjb nccrnef va n juvgr ertvba. Fortunately, that has a very different feel from (what I'm pretty sure are) the actual rules, and it doesn't have a single best way of scaling up to a 1-4 puzzle, so it's not difficult to decide it's an accident.

Greatly enjoyed the first, and looking forward to solving the second! (Of course you, Melon, AND Zot would all release really tempting puzzles the day I absolutely had to get some puzzle construction done, wouldn't you?)
David MillarDavid Millar on January 7th, 2011 05:31 pm (UTC)
Now I'm thinking I've missed something. I was able to solve the example uniquely without the additional dot, so I may be missing some additional part to the dot rule. I need to re-attempt the first puzzle as well - I made a mistake last night due to sleepiness and should just start it over.
motrismotris on January 7th, 2011 05:39 pm (UTC)
If you have doubts, I do have the answers available; the first puzzle is here and the second puzzle here
David MillarDavid Millar on January 7th, 2011 05:58 pm (UTC)
Seeing the first solution, I see that I made a stupid assumption that screwed up my first try. That fixes the error I previously made, but makes the rest of the puzzle much more difficult to solve logically now. Hmm.
David MillarDavid Millar on January 7th, 2011 06:19 pm (UTC)
Figured it out (finally). Great puzzle. A++++ would buy again.
mathgrant.blogspot.com on January 7th, 2011 06:01 pm (UTC)
Eff you, you a-hole.

I don't want the answers -- I want instructions. I have absolutely no clue what the effing ess any effing thing's supposed to mean in this piece-of-ess puzzle -- especially the effing colors.
motrismotris on January 7th, 2011 06:13 pm (UTC)
I'd respond even quicker if you'd used real swear words, but you can Rot-13 these steps.

1. Svyy fbzr pryyf fb gung rnpu qvtvg va gur vaqvpngrq enatr nccrnef bapr va rnpu ebj/pbyhza.
2. Va gur (oyhr/benatr) pbyberq pryyf, gur pvepyrf vaqvpngr gur cbfvgvba bs gur havdhr (znkvzhz/zvavzhz) inyhr va gur fheebhaqvat pryyf, cebivqrq bar rkvfgf. Vs abg gurer vf ab pvepyr va gung pryy.
mathgrant.blogspot.com on January 7th, 2011 11:46 pm (UTC)
Whaddya mean "eff" isn't a real swear word? Next thing I know, you're going to tell me that the Wet Bandits from Home Alone aren't criminal masterminds!
(Anonymous) on January 9th, 2011 04:56 pm (UTC)
Thanks for that. I read (2) word by word until I had an idea. "havdhr" did it for me.
(Anonymous) on January 13th, 2011 03:18 am (UTC)
If those are the rules, shouldn't the dot in R2C6 in the example point... elsewhere?
motrismotris on January 13th, 2011 03:40 am (UTC)
The dot in that orange cell is pointing towards R3C5. That is in accord with the rules. Note that "oyhr" and "benatr" act differently if this is what is causing confusion.
Adam R. Woodzotmeister on January 7th, 2011 07:02 pm (UTC)
I'm disappointed in you, Grant. Whatever happened to your sense of discovery, the desire to see new puzzles with, say, Japanese instructions, and try to figure them out? (Much less your vocabulary?) You are hereby sentenced to community service with Naoki-Projekt. It's for your own good. - ZM
mathgrant.blogspot.com on January 7th, 2011 11:49 pm (UTC)
Giving up easily on puzzles without instructions is why I don't participate much in puzzle extravaganzas!

Truth be told, were it not for http://indi.s58.xrea.com/ and its English explanations in PUZ-PRE, I'd've never figured out all of the rules of Hebi-Ichigo or Shakashaka. I think I just lack a certain creative type of thinking. . . .
(Anonymous) on January 12th, 2011 11:52 am (UTC)
ambiguity in first puzzle?
I've solved the sample and second puzzle successfully now, but the first puzzle appears to have an ambiguous solution. I get the same solution as the one you linked above, except I can switch the locations of the 2s in columns 1,3,6. Probably I'm missing something obvious...

Cheers
Rob
MellowMelonMellowMelon [wordpress.com] on January 12th, 2011 04:40 pm (UTC)
Re: ambiguity in first puzzle?
I think the red cells in the top left give you a problem then.
motrismotris on January 12th, 2011 05:17 pm (UTC)
Re: ambiguity in first puzzle?
While I didn't state the rules formally, in all cases a colored cell with no circle in it also holds useful information. Moving the 2's around would force the addition of some circle clues.
nathan_0 on January 7th, 2011 06:35 pm (UTC)
Last week's puzzle
**spoiler alert**

On last Friday's puzzle, did anyone start out by calculating 220 minus the minimum number of pre-labeled regions? I didn't do it until I was about 5 minutes in and it felt a bit unsporting. The number I came to was 16 - which ended up being correct - and I did this before I had solved for either side's break-in (my eye was drawn to the center for some reason). I felt like having that 16 in my pocket gave me a big advantage as soon as I uncovered the min-11 along the rightmost column and the rest of the puzzle collapsed pretty quickly after that.

Any thoughts on using such brute force mechanics in a puzzle solve? Certainly the puzzle was solvable without doing that, but only because of the clever hooks installed by Thomas. It also felt like more legwork than I'm accustomed to doing on Motris puzzles.

Related: I once had a professor who explained during a Calc III class that we were like kittens and the problems in the text book half dead rats being brought home by our mother cat. He ended up making several elegant points with this analogy, one of which was that there are limits to how far into a brute force solution we should expect to go on a given practice problem as we were still developing our core toolset. I'm sometimes reminded of that professor when I'm working on Tom's puzzles. I'm inclined to call purely computer generated puzzles 'wild' and hand crafted/computer assisted ones 'domesticated', but perhaps the only worthwhile uncrafted puzzle exercise is the creation of puzzles.

Aside: is mathgrant's post the result of a profanity filter or is he naturally that funny?

Nathan
motrismotris on January 7th, 2011 06:48 pm (UTC)
Re: Last week's puzzle
I hadn't thought to check the puzzle for that (I certainly realized afterwards that I didn't have a lot of unmarked regions), but the solution path doesn't need it.

I suppose it's kind of like bringing a sledgehammer to the puzzle, so having it as a technique is valuable (and resolves the connections on the left side).

Aside: mathgrant is naturally that funny.
MellowMelonMellowMelon [wordpress.com] on January 7th, 2011 10:12 pm (UTC)
Re: Last week's puzzle
I find that 'wild' and 'domesticated' dichotomy to be one of the defining differences between Japanese and American interests in general. Japan has nikoli and other hand-crafted sources; many Americans are happy to gobble up computer-generated Kenken and Sudoku. So-called "Eastern RPGs" tend to be more linear but can tell a more controlled, in-depth story as a result, not to mention the whole visual novel genre. "Western RPGs" are farther on the sandbox side of things. There might be more examples (or counterexamples), but I'm not familiar with all types of media or cultures. And of course, this isn't universal - we're all proof of that.

I can understand why someone might be more interested in conquering difficult computer generated puzzles, in the same way many mathematicians prefer doing research to solving contrived problems used in homework or contests. There is something slightly appealing about finding new ways of your own to make progress, rather than following a trail blazed by the author... but it's frankly not my preference, especially when most of those 'wild' puzzles end with frustration so much more often than new discoveries.

I also think calling handmade puzzles 'domesticated' has the wrong connotation. Some of them here online are vicious. I feel like there's a better metaphor under my nose, but it's not coming to me.
nathan_0 on January 7th, 2011 10:18 pm (UTC)
Re: Last week's puzzle
Agreed on the domesticated connotation, it was the only thing that came to me. Though in the defense of hand crafted puzzles, I believe there are whole breeds of dog that could whip the snot out of a full grown wolf.
mathgrant.blogspot.com on January 8th, 2011 12:14 am (UTC)
Re: Last week's puzzle
I can understand why someone might be more interested in conquering difficult computer generated puzzles, in the same way many mathematicians prefer doing research to solving contrived problems used in homework or contests. There is something slightly appealing about finding new ways of your own to make progress, rather than following a trail blazed by the author... but it's frankly not my preference, especially when most of those 'wild' puzzles end with frustration so much more often than new discoveries.

Quite frankly, I agree with you 100%, having encountered many "difficult computer generated puzzles". Nikoli has made some good puzzles utilizing "new discoveries", but there's generally some kind of sensibility unique to Nikoli that makes them less frustrating, somehow. The hardest Fences puzzles on PuzzlePicnic, for example, tend to be more frustrating trial and error than "new discoveries" for me. I enjoy the community on that site, but I find myself more motivated to dive into my Nikoli books than to solve PuzzlePicnic's puzzles. I was recently motivated to submit more puzzles to get to the prerequisite 100 for the title "experienced author". I'm far less motivated to earn the title of "experienced solver", which requires having solved 1000 puzzles (I've done 404).

I also think calling handmade puzzles 'domesticated' has the wrong connotation. Some of them here online are vicious. I feel like there's a better metaphor under my nose, but it's not coming to me.

A German shepherd is domesticated, but that doesn't mean it can't be vicious if trained as a police dog. Perhaps "wild" versus "controlled"?
( 29 comments — Leave a comment )