[cworth.org] / src / christmas_code.mdwn

[[!meta title="Carl's Christmas Puzzle Hunt"]]

[[!tag puzzles]]

I've slipped into an accidental tradition of composing a puzzle hunt
for my sons every year at Christmas time. We call it "The Code" and
it's now one of the most anticipated events every December.

If you'd like to see some examples of what the puzzles are like, you
can look at them here (or even try solving them yourself):

  * <a href="/2014">2014 Christmas Puzzle Hunt</a>
  * <a href="/2013">2013 Christmas Puzzle Hunt</a>
  * <a href="/2012">2012 Christmas Puzzle</a>

As one might hope, I believe my puzzle-designing skills are improving
with practice. So hopefully the more recent puzzle above show some of
my better work.

This tradition has become sufficiently involved that the history of
The Code really deserves to be captured. Here's the story of how it
all started, and how it's changed over the years.

## On Keeping Christmas a Surprise: The Invention of The Code

One year, my wife and I were lamenting to some friends that Christmas
gifts were losing their mystery at our house. Our boys were getting
good enough at poking, prodding, and predicting their presents that
there weren't a lot of surprises left for Christmas morning. Our
friends had had similar problems with their children, and we ended up
comparing notes on things we had done to keep gifts secret.

One obvious tactic is to disguise a gift while wrapping it. This
technique has already been developed to an artform by our boys. They
love giving each other a large, wrapped gift which, when opened
reveals a smaller, wrapped gift inside. Then, like a set of matryoshka
dolls, that gift reveals a smaller, and smaller gift. In our family,
as soon as this unwrapping process is started, everyone knows what the
end result will be. The final few layers are so tiny that there's no
room for boxes anymore, just many, many layers of different colored
gift wrap and excessive amounts of tape. These last layers take a
tremendous amount of effort to get through. And the reward for all of
this hard work of unwrapping? Inevitably, it's a single penny. We
don't recall who wrapped the first penny, but it's now a standing
tradition where the boys try to outdo each other each year by
disguising a penny in the most elaborate way possible.

Clearly, we're not willing to go to such heroic efforts to disguise
every gift we wrap. We had experimented with waiting to put the
Christmas gifts out under the tree until just before Christmas, but
that wasn't much fun. It robbed the boys of a lot of the fun
anticipation of seeing the gifts under the tree throughout December.

Our friends shared an idea that had worked well in their family. What
they did was not write names on any of the gifts, but instead wrote a
single number on each gift indicating the recipient. And the method
for choosing the numbers was selected in a new and unpredictable way
each year. For example, one year my friend (who happens to be a
dentist) wrote the number of teeth that each child had lost on their
gifts.

We thought this was a great idea, and we decided to give it a try. And
none of our Christmas gifts have had any of the boys' names on them
sense, (though they've had just about everything else possible). Read
on for a rundown of what we have done for the code for each year.

## The Early Years: Locking the Code up Tight

For the first year, we adopted a very simple strategy. We wrapped each
boys gifts in a unique color of wrapping paper. We didn't explain
anything, and when the boys asked why none of the gifts had names on
them, we were evasive. Then, on Christmas morning, we told them which
color of gift wrap corresponded to each boy.

The second year, the boys correctly predicted that we wouldn't use the
same technique, and they immediately set about trying to crack "the
code" for the Christmas presents. They started convening secret
meetings to discuss theories. We discovered some of the notes from one
of the meetings and found that they had constructed an entire table
mapping out the following variables for each gift under the tree:
Wrapping paper color, Number of bows, Colors of bows, Picture on the
tag, Names on the tag. You see, this year, I did put tags on the
presents, but instead of their names I wrote the names of characters
from nonsense poetry: "To: The cat; From: The fiddle", "To: The cow;
From: The moon", "To: The dish; From: The spoon", etc.

Their table was fairly effective. They were able to eliminate many
variables that couldn't work. If there were more colors of wrapping
paper tan children, they assumed that could ignore that. If there were
only one or two bows per present, they assumed that couldn't identify
one of the four boys. Fortunately for me, they didn't crack the code
that year, but only because their table hadn't accounted for the color
of ribbon on each present.

By the third year, I realized that I had to put more thought into
designing the code. Here was an active group of intelligent agents
determined to find the information I was trying to hide. I was careful
this time to imagine every variable they could track, and ensure that
each variable appeared with four different values, evenly distributed
among the presents. I also ensured there was no correlation between
any of the variables. Grouping the presents by wrapping-paper color,
bows, ribbons, or anything else would always yield four entirely
different sets of presents. (So yes, this meant that now my wife and I
needed to consult our own table before we could know how to correctly
wrap each gift).

Then, for the actual information, I chose two variables I thought they
would never track. We carefully folded the flaps on each gift either
in the same direction on each side or in opposite directions on each
side. Then we either folded all cut edges away to leave clean creases,
or left the raw edges exposed. This gave us four sets of presents:
Matching flaps creased, Matching flaps raw, Opposite flaps creased,
and Opposite flaps raw.

Of course, the boys never even looked at the flaps, and all of their
attempts to find logical groups were foiled. When I revealed the
answer on Christmas morning, I was smug, thinking I had "won" by
creating a code they couldn't crack. Of course, the boys called me out
saying that what I had done was totally unfair. (How could it have
been unfair I thought? This was a game that I had invented myself?)
But the boys were totally right. My problem was thinking that this was
a game, when in fact, this should have been a puzzle.

Years later, I read the book "Puzzle Craft" by Mike Selinker and
Thomas Snyder. In the introduction, Mike Selinker describes the lesson
that my boys were teaching me. He says that a game is a contest with
two equal sides and the outcome is in doubt, (either side has a
roughly equal chance to win). In contrast, a puzzle is a contest with
two wildly-unequal sides where the outcome is never in doubt, (the
weaker side will always win). When I treated The Code as a game, it
wasn't fun for any of us. The odds were stacked too much in my
favor---I could always create an impossible-to-solve puzzle, but who
has fun with a puzzle that's impossible to solve? That just leads to
frustration and giving up. A good puzzle, in contrast, has plenty of
frustration, but enough fun and reward that the solvers stick it
through to the end. So I needed to learn to create a puzzle.

## The Code Today: The Code as a Puzzle

My boys taught me that the code needed to be fair. That is, they
needed to be given enough information to be able to solve the
code. There could still be lots of deception and trickery, but they
needed to know that with perseverance, patience, and creativity they
could actually find the answer. They also gave me a second ground
rule: The solution to the code must be relevant and interesting. A
final answer of "you get the presents with the matching flaps with the
raw edges" doesn't cut it. Instead, the solution to the code should
actually point to the boys themselves. Basically, they were telling me
"Design us a puzzle", and "Make it a good puzzle", and it just took me
some time to figure that out.

<a href="/2012">Christmas 2012</a> was the first year I approached The
Code as a puzzle. For that year, I labeled each present with nothing
more than a small QR code. Scanning the QR code linked to a web page
with a silly animated GIF, a solid-color background, some nonsense
poetry in the title, etc. Somewhere in all of that was a hidden
indication of who the intended recipient of each present was. So this
was simply one puzzle, and a lot of obfuscation. I thought this puzzle
would have been easier than it was, (a common problem for early puzzle
designers from what I understand). But the boys had a lot of fun with
it, and with some hints at the end, they figured things out by
Christmas.

<a href="/2013">Christmas 2013</a> was the first year I stepped up and
instead of designing just one puzzle, I designed an entire puzzle
hunt. A puzzle hunt is a connected series of several puzzles. Many
puzzle hunts also included "meta-puzzles" where the solutions to
several puzzle combine to form a new puzzle. This was my first puzzle
hunt to design and it included 24 puzzles, 5 metapuzzles, and 1 final
metametapuzzle, (where the solutions to 4 previous metapuzzles had to
be combined in another puzzle) That was probably over-ambitious for my
first puzzle hunt, but it worked fairly well. There were a couple of
bugs in the puzzles that I should have caught with better testing in
advance.

One thing I was really pleased with was that I intentionally
included every element from the previous puzzle, (animated GIFs,
random background colors, nonsense poetry, etc.). But where in 2012
many of these elements were meaningless red herrings, in this year's
hunt, every element was used in at least one puzzle. I was also happy
that I included a mechanism for providing additional hints along the
way. (And I did this in a way that I could revise those hints before
the boys encountered them, so I could fine-tune the hints based on
where they were getting stuck.) That was very useful, and I used that
to my advantage again in 2014.

<a href="/2014">Christmas 2014</a> was my second puzzle hunt, so it's
clearly now a new tradition that we won't give up for some
time. (We'll just need to adapt things when some of the boys move out
of the house, etc.). I do feel like I'm getting better at puzzle
design, and the boys are still having a lot of fun solving this. I
wrote up a <a href="/christmas_code/2014">blog post</a> giving some of
my feedback on how the solving experience went this year.