Behind the Scenes

The KOB Letter

My wife surprised me with the KOB letter clue coming home from work one day. I sort of figured out that I needed to fold the letter to get the answer, but I had no memory of how to fold a KOB. Her folding style (being raised in the South) differed drastically from what I remembered the end result being, and was nothing like what I found online, or how I later was reminded by my mother that we used to fold them. This told us that we needed clear folding instructions supplied somehow.

We experimented with various ways of indicating the folding. One test involved drawing "arbitrary" patterns along with the words; folding the letter correctly would cause patterns to line up at each step. Unfortunately this only worked marginally well on the thinnest paper, and was increasingly harder to get right as the paper got to a pleasing thickness.

In the end, we decided to supply the folding instructions explicitly via another folded example instead of implicitly in the clue. It was a little easier on us, and certainly easier for the players.

To ensure that a wrong folding didn't accidentally produce a plausible clue, we took a single piece of paper and repeatedly folded it in a variety of wrong ways. For each resulting folded KOB we wrote A1/A2/A3, B1/B2/B3, C1/C2/C3, etc. on the exposed triangles of paper on each side. Many foldings shared triangles; this allowed us to write words like "NOT" and "WRONG" on the commonly-shared squares. Transcribing the experimental results to the computer allowed us to write down possible "junk" clues and see check that no D1/D2/D3 triplet of words made sense. Here we see the in-progress results from 5 different foldings (where A and B represent the front and back of the correctly-folded KOB).

Realizing how painful it would be to write out and test all 128 folding combinations, we only placed words on half the page, simultaneously halving the possible foldings and also helping to guide players to make the second fold correctly. (Doing it incorrectly would immediately cover up all words; clearly wrong.)

The Map Grid

This challenge was inspired by the Genome Game puzzle Number Grid.

Constructing this challenge required us to:

• Find a period map with explicit gridding and sufficient detail to make it a 'fun' challenge to find many locations.
• Figure out six different destination clues that consisted of the same numbers of letters, using only letters found in the first names of the family members.
• Find that number of locations on the map. In addition to wanting the locations to be appropriately easy/hard to find, we also had the constraints (due to the devious "SWAPGRIDLETTERSANDNUMBERS") that no location was allowed to be on the diagonal, and that once one location was used its mirror across the diagonal could not be used. For example, having chosen a location at M7 on the map we could not use any location from G12, since that was where the first letter of the 'swapped' clue would go.
• Mapping the letters of the alphabet to ordinal name and letter position combinations.
• Create a six unique grids that included "SWAPGRIDLETTERSANDNUMBERS" using the original clues but the appropriate clue in the mirrored spot.

My wife knows of an excellent paper store with maps, and it didn't take her long to find the maps we used.

It took us only a few passes at the clues to decide on the final 25-character target; after that it was not too much work to massage the description of a destination into that many characters. We intentionally chose the large number of characters both to prolong the location-finding part of the challenge, and also to give us more freedom to clearly express a destination.

My wife also handled the choosing of locations herself so that there would be a consistency among difficulties (and, I suspect, from allowing me to choose impossible-to-find locations). I wrote a small Ruby script that processed the locations she found and ensured that there were no duplicates, no diagonals, and no shared mirrored cells.

Possibly the hardest—and obviously most error prone—part of building the challenge was coming up with the correct list of family members. I entered everyone (I thought) into a spreadsheet along with their birthdates and sorted it. I also marked family members who normally go by something other than their given name (such as their middle name) so that I could avoid using them and the confusion they might encourage. I then wrote small Ruby script to parse the output, mapping the letters of each name to the name and letter index. In an attempt to make things easier for the players, I removed family members after the 22nd, stopping early in the 3rd generation.

Finally, I wrote a small HTML page that produced the nicely-formatted grid, with a small amount of JavaScript to randomly fill the targeted locations with any message I wanted (using the letter mapping found by the Ruby script).

I had wanted to write the location names in pencil on the back of the map. My goal was to make them harder to find, and also to reward teams who thought to copy them from the back of the map to their notepad. My wife, who has plausible period handwriting (unlike my chicken scratch) was unwilling to write out 175 locations instead of packing for our trip to the reunion. As a wholly reasonable request, we simply printed them out instead along with the grids.

The Archery Challenge

We did not put too much thought into the rules for this game. The only prep work needed was buying bows, arrows, and targets…and getting them to the reunion.

We bought a couple children's recurve bows, 36" long and 10 pound draw weight, and blunt arrows to match, and a target for our own testing at home before going to the reunion. Due to the size, we decided to buy targets in a town near the reunion.

As we spent too much time scrambling to write the semaphores on the back of the puzzles the morning of our flight, we had to perform some incredibly hasty packing. I packed the arrows, unstrung the bows…and then realized that they would not fit in any suitcase or duffel bag we had. I figured we'd just tape the plastic together and put them in the overhead bins.

Apparently TSA considers putting a handle on a flexible piece of plastic an unreasonable security hazard. BOWS AND ARROWS ARE NOT TO BE TOLERATED. EVEN WHEN IT IS A CHILD'S BOW ROUGHLY INDISTINGUISHABLE FROM A FIBERGLASS TRADESHOW POLE, UNSTRUNG, WITH ARROWS PACKED IN THE SUITCASE. FOR THE LOVE OF GOD, YOU MIGHT HAVE TWINE IN ANOTHER BAG, AND YOU COULD FIRE PENCILS AT THE PASSENGERS.

In case it wasn't obvious by the capital letters, I'm moderately annoyed: both at their silly rules and at myself because it didn't occur to me to check to see if they had these silly rules. Sparing you the play-by-play drama at the airport ("these are now in a secure area, and we need to write up an incident report; 'they' like to hand out fines for this sort of thing, so you should expect a letter"), the result was that I couldn't send them as carry-on, I missed the time window for shipping them as $50 oversize luggage, and in the end I spent $20 at a USPS window in the airport getting them to ship the bows to a family member. As they couldn't arrive in time for the reunion, we had to buy another pair of bows when we bought the targets.

The Puzzle

This challenge was loosely inspired by the Genome Game puzzles Strips, Rectangles and Chasing Pauling.

We originally had wanted to find an old group photo of our ancestors, Photoshop semaphore flags into it, and then get it cut into a custom jigsaw puzzle. I had wanted to get really nasty and create a jigsaw puzzle using only edge pieces (every piece has a straight edge; not all pieces interlock) or with other odd shapes. I wasn't quite so mean as to consider making it a Schmuzzle puzzle or other Escher-esque tiling pattern.

By the time we found an appropriate photo, we realized that we were out of time to get a custom puzzle made. Despite the fact that dinosaurs and clown fish and the like were not very appropriate to the game, we decided to save time and just buy one puzzle for each team and draw the semaphores on the back.

We then had to decide how large a puzzle to buy. We didn't want the overall dimensions too large, given that teams would be constructing them on ad-hoc surfaces. But furthermore, we didn't want so few pieces that it would be trivial to solve, or so many pieces that it would take overly long.

We went out and bought a 48-piece jigsaw puzzle for our children and timed how long it took us to put it together. (Our time included one adult doing the puzzle and two young children 'helping', since the teams would likely be facing a similar situation.) We decided that the 10 minutes it took was not long enough, and (running out of play testing time) guessed at 60-75 as a reasonable target.

To write the semaphores on the back of the puzzle we first had to find pages in Walden the the necessary words or phrases needed to give the five different clues for the next destination. (Only five destinations were needed because this challenge sent three teams to the same destination for the Archery Challenge.)

One of the reasons that we chose Walden originally was because the full text of the book was available for free in electronic form from Project Gutenberg. This allowed me to write a Ruby script that scanned the book for any word or phrase and report page(s) where that occurred, along with surrounding context. The one hitch in this plan was that free versions of the book do not have page breaks, or when they do, they are not in the same locations as the printed copy we had ordered.

I modified my script to first split the book into pages based on our printed copy, based on a list of phrases at the start of each page. Tedious though it was, it took under half an hour for my wife to read to me the first few words on the first hundred pages of the book while I typed them in. At that point we were able to enter lists of words we thought we might use for clues and find out if they existed.

The hardest challenge here was working around missing words. For example, the location known as “The Old Swimming Hole” was particularly difficult as the words “old”, “swim” and “swimming” do not appear anywhere in the book. (We ended up going with “swing into water hole” for this clue, as there is a rope swing here.)

We attempted to place the numbers on the puzzle in obvious pairs. We also offset one row from the next, to avoid the possibility that they could be seen as a grid of numbers. We attempted to place each semaphore almost completely within a single puzzle piece so that it would not be easy to construct the puzzle starting with pieces flipped over. Because some semaphore sigils are rotational mirrors, we ensured that the clue on the back was drawn sharing the same top as the rest of the puzzle.

The Poker Lock

This challenge was inspired by the Genome Game puzzles Cards and K.

Designing this challenge required us to solve several problems:

• What could we buy that was cheap-but-sturdy, that could be secured with a lock?
• What pairs of radio stations resulted in single-digit integers when subtracted?
• What pages of Walden, before page 15, had letters, in order, with radio stations of New York on them?
• What hands of poker could we make using only the the matching pages?
• How could we ensure that non-Gambling players recognized the five cards as poker hands?
• How could we hint to people that the poker hands needed to be arranged in order?
• How should we best convey the intent of the convoluted solution clearly enough, without just giving away the process?

Settling on the hinge as the secured object was inexpensive ($3 per hinge) but presented its own problems. To prevent the clue for the final destination from falling out, or being pushed out, we tightly folded the clue, taped it inside the hinge, clamped the hinge down hard on it and added enough spacers on the lock that the hinge could not be slid open along the lock. Even then, we discovered in play testing that a determined individual with the right tools could still poke the clue out sideways. Further, the night before the game I realized one could simply (with the right tools) pop the pin to the hinge and open it backwards. It might even be plausibly considered an inventive solution.

Before the game, we explicitly mentioned that bypassing or breaking locks was against the rules. This did not stop one of the teams from poking the clue out from the side anyhow. Thankfully, that team didn’t win.

I used a modified version of the script that found words and phrases in Walden to find radio station letters on the same page. I had it prioritize pages with all capital letters, and only when those could not be found did it look for letters at the start of words and (failing that) letters anywhere within a word. Thankfully, most radio stations could be found in the first 14 pages, keeping our options open.

We did not want to use multiple decks of cards for the poker hands for a single team, so we had to come up with four hands of poker using distinct cards. Because each hand needed to include both diamonds and clubs, flushes and straight flushes could not be used. It took a couple hours of experimenting with combinations of usable radio station pairs, the pages they were on, and poker hands to come up with the final hands. During the process, we also ensured that we had some instances where the club was larger than the diamond (e.g. Seven of Diamonds, Queen of Clubs) so that a team mistakenly subtracting the card values directly would get a negative number and hopefully realize that it was not correct.

For the final card hand subtracting to zero, we looked for phrases in Walden that would indicate "Hey, the answer has to be zero!" I was very excited when I found the phrase "it makes but little difference", but unfortunately this appears on page 55 of the printed book, which we couldn't represent. The phrase we eventually found—"It is the same"—was just about as good.

My wife solved the final three problems at once by coming up with phrases for the backs of the cards that mentioned "poker", indicated that you had to use the cards to go look in the book, and (with the rhyme and meter) helped indicate that you would have to arrange the phrases in a particular order.

I am proud of this challenge, and wish that there had been enough time left in the game for teams to really explore it. I do wish that we had the budget to opt for a locking metal box instead of a hinge.

Orienteering

There was not much to design for this challenge. We knew that we wanted the game to finish near the lunch location, so we knew where to start. From there I took a compass in hand and repeatedly sighted for something interesting in a desired direction and made big strides to roughly pace it out. About half way through I saw what I wanted to be the final destination, and made a roundabout route to get there.

We included rough compass points (e.g. WNW) on the clue to help people understand that they should use the compass, but also included the more accurate degrees (shown on the compass) since rounding to the nearest 22.5° was not nearly accurate enough.