?

Log in

2008 Mystery Hunt puzzle: Monopoles - devjoe [entries|archive|friends|userinfo]
devjoe

[ userinfo | livejournal userinfo ]
[ archive | journal archive ]

2008 Mystery Hunt puzzle: Monopoles [Jan. 22nd, 2008|08:32 pm]
devjoe
[Tags|]

The Magnetic Monopoles variant of the standard magnet puzzle I first saw in the 2007 USPC (created by zotmeister  ). The idea to make a Mystery Hunt puzzle based on this variant is due to Mark Halpin, one of our puzzle editors for the Hunt. The mechanism for generating an answer from these puzzles, and the actual construction of the nine monopoles puzzles (and therefore most of the work that went into this puzzle) was mine.

17 teams solved Monopoles. First and fastest was the Evil Midnight Bombers, solving it Friday at 4:40 PM, 3 hours and 23 minutes after they opened it. The last solution came from Left Out, Sunday at 9:51 AM.

The flavor text helps you understand the basic rules here: each 1x2 rectangle in the nine grids is to be filled in with a dipole magnet in either orientation, or left blank. Each 1x1 square is to be filled in with a magnetic monopole, and CANNOT be left blank. The normal magnet puzzle rule also applies: two like charges can never touch along an edge. The numbers around the edge tell you how many positive or negative charges are in a row or column.

First, solve the nine puzzles. Then, as suggested by the flavor text and the repeated field strength values (including upside-down ones), stack the grids following the magnet adjacency rules, rotating and/or flipping the grids, except for the first one provided which goes at either the top or bottom of the stack (flavor text suggests bottom) without being turned or flipped. You will end up with something like:
+-+-+##+-
#+-#-##-+
#-+#+-##-
-+-+-##-+
#####+-+-
#-+-#-###
-+-+###-+
+-+-+#-+-
##-+-##-+

-+-+-+##+
+-+-+##+-
#+-+-#+-+
##+###-#-
+#-#+-+##
-########
+-+-#+-+-
-+-##-+-+
+-+###-+-

+-+-+-##-
-+-+-+#-+
##+-+##+-
+##+-#+-+
-+####-+-
+-+-+#+-+
-+-######
#-##-+-+#
###-+-+-+

-+-+-+-##
#-+-+-+#-
##-+-+-#+
-#+-+##+-
+####-+-+
-+#+-+-+-
###-+-##+
-+-+##+##
+-+##+-+-

##+-+-+-+
+##+-+-+#
-+#-+-+##
+-#+####+
-+##-+#+-
+-+-+-+-#
-+-+-####
+-+###-+-
-+##+-+-+

+-#+-+-+-
-+#-####+
+-#+-+-+-
-+#######
+-+-+-+-+
-+-#-+#+-
+-#######
-+#-+-+-+
##-+-+-+-

-+-#+-+-+
+-+#-+-+-
-+-###+-+
+-+-+#-+-
#+-+-####
#-####-#+
-#+-+-+#-
+#-+-#-+#
##+-+#+-#

+-#+-#-+-
-+#-+#+-+
+##+-+-+-
-+####+-#
+-+##+-##
-+#-+-+#-
+-#+-+-#+
-+##+-+#-
##-+-+-##

-+-#+-+-+
+-+#-+-+-
-+-###+-+
+-+-+#-+-
-+-###+-+
+-+#-+###
-+-#+-#+-
+-##-+#-+
-+#-+-#+-
Now, through each vertical stack of magnet-ends, add up the total field strength, considering the +s and -s and the strength printed on each frame. Use this value to Caesar-shift the letters printed on one grid. This gives the clue MAGNETIZABLE METAL PLACED IN A LATTICE THAT MIGHT BE AN EIGHT LETTER WORD MEANING FOOTBALL FIELD, which clues GRIDIRON.

I had never created magnet puzzles before, but I had a good idea about how to create them. At some point I saw and solved some easy magnet puzzles that gave you strength values for every row and column. These were really easy, and you can do a sort of overall balance calculation on them that tells you the net number of +/- magnets crossing between each two adjacent rows or columns, which simplifies the puzzle almost to the point of triviality.

So I figured that I would just fill in the grid however, compute all the row and column totals, then redraw the grid without those totals or the magnets, and add row/column totals in strategic places where they would actually help the progress of the solution, until I could solve the puzzle (by hand). Occasionally I found that I had an ambiguous pair of magnets, and I restarted with a slightly different solution in mind, but this was in fact how I constructed the puzzles. After the first one, I drew my grids so that each would stack on the previous one, and a program verified that they would stack in no other way, even with rotation and reflection.

Finally, I did the calculations for the final answer and wrote a program to shuffle/rotate/flip the grids appropriately and draw these nice looking diagrams. For my Hunt puzzles I learned how to generate SVG graphics, though since not all browsers have SVG support built in, I used Inkscape to convert them to standard formats for the web site.
LinkReply

Comments:
[User Picture]From: motris
2008-01-23 02:30 am (UTC)
Of the many, many, many puzzles (not counting the metas) I worked a bit on in this hunt, monopoles was my favorite and one where I actually worked on it from start to finish. When it first printed, our team called me over and some others helped with the magnets but I ended up finishing the majority after we passed around our progress. Ideally we would have had transparencies to put the grids on so the overlaying was easier, but we figured how to see it in our heads just fine. Then was the addition of all the poles by the field strength to extract the answer as we expected. Altogether, a really solid construction, considering you are new to making magnets puzzles.

(Reply) (Thread)
[User Picture]From: devjoe
2008-01-23 01:37 pm (UTC)
Thank you for your compliments. I am glad you enjoyed the puzzle. I figured that some teams would split up the magnet puzzles and solve them in parallel, but with the fastest solve in 3 hours, I don't have any evidence that teams were doing this.
(Reply) (Parent) (Thread)
From: (Anonymous)
2008-01-23 06:20 pm (UTC)
Setec worked on the magnet puzzles in parallel -- me, Anomaly, and QED. We handed off the finished grids for someone else to do the stacking/letter extraction.
(Reply) (Parent) (Thread)
[User Picture]From: ztbb
2008-01-23 06:20 pm (UTC)
Setec worked on the magnet puzzles in parallel -- me, Anomaly, and QED. We handed off the finished grids for someone else to do the stacking/letter extraction.
(Reply) (Parent) (Thread)
[User Picture]From: oxeador
2008-01-24 09:04 pm (UTC)
Manic Sages started solving them in parallel (and got the first 3 or so this way), until somebody came up with some coding to solve them and made us humans obsolete.
(Reply) (Parent) (Thread)
From: (Anonymous)
2008-01-23 11:35 am (UTC)

SVG support built-in in all browsers

except for this bad excuse for a browser called Internet Explorer.
Upgrade to anything else, as more and more people do
(Reply) (Thread)
[User Picture]From: devjoe
2008-01-23 01:32 pm (UTC)

Re: SVG support built-in in all browsers

Ugh, yes, internet exploder is the reason for not actually using the SVGs in the puzzle. I am not using IE whenever I can avoid it, but I felt it was good to ensure my puzzles worked in IE for whatever portion of the Mystery Hunt audience is using it. I will have another post on this subject soon.
(Reply) (Parent) (Thread)
[User Picture]From: zotmeister
2008-01-23 07:53 pm (UTC)
You first saw it there because I invented it for the competition! (The monopoles were created as a simple way to report the solution on the web form, and I built the puzzle so as to make them the focus of solving.) I'm honored to see it reused in such a prestigious manner (or, at least, will see it once the Hunt is archived).

For the record, in general I try to avoid making solutions "in advance" in the creation of any of my puzzles. [The only exceptions so far are the Allies with the hidden "21-omino?" (which of course was preplanned) and the queens Sudoku (where I used a program to make sure what I was trying to do was even possible).] For Magnets, I make the grid framework first, usually in an interesting shape, and then just start adding in moves, step by logical step, until the grid is full. As I go along, I look for ways to make the puzzle interesting/clever/tricky. Of course, in making a metapuzzle out of them [clever idea stacking them, by the way - suddenly I want to make one in three dimensions], some things need to be plotted out in advance, but I find it much more difficult to be creative in the construction when I'm tied to a particular solution I need to have at the end.

What does it take to actually participate in the Hunt (player or admin, either way)? I've been curious about it for some time now. - ZM

Edited at 2008-01-23 07:54 pm (UTC)
(Reply) (Thread)
[User Picture]From: devjoe
2008-01-23 09:45 pm (UTC)
To participate in the hunt, you need a team. Theoretically you could sign up by yourself, but you won't get very far. It's designed for teams of many people. So find somebody with a team you can join. You might be able to join my team, for instance. But ask closer to the next hunt, when we are actually getting organized and starting to plan going to the Hunt. Then, ideally, you come to MIT for MLK weekend (though many teams have some remote solvers, it's often a challenge to keep remote solvers updated on the state of what's solved/in progress).

The team that wins the Hunt makes and runs the next one, so if you want to admin the Hunt or write puzzles for it, you need to be on THAT team.
(Reply) (Parent) (Thread)
From: rlangmit
2008-01-24 08:10 am (UTC)
Good puzzle--I didn't do any of the monopoles themselves except stacking some upside-down, but I did think of the field strength idea. Unfortunately, I applied it wrong (first looking for a net strength of zero and then doing the correct thing but only at the monopoles--argh!) and had to wait for a teammate to finally get it in about 5 seconds.

Would you mind posting stuff at (or linking it to)

http://mystery_hunt.livejournal.com

This was created last year, and it would be very cool if people from various teams would post in one common place, since some of us don't really know all of you guys. It's hard to play "follow the comments" and find all the interesting blogs!

Ryan (Up Late)
(Reply) (Thread)