Candy Mountain Solution

How It Works

The clue sheet mentioned that the candies came in "5 delicious flavors" and included an ingredient list, which hopefully made it clear that teams needed to actually taste the candy. Teams needed to begin by grouping the candy by flavor. The groupings are below:

Peppermint

Peppermint

Cinnamon

Cinnamon

Grape

Grape

Lemon

Lemon

Banana

Banana

At this point this puzzle should be solvable online.

Once teams have organized the candy by flavor they needed to notice that the number of candies corresponded to the word-length of the flavor (for example, there were 6 banana flavored candies). Once this connection was made, the next step was to notice that the shapes also corresponded with letter frequency (using banana again; there was one shirt [b], three truffles [a], and two hearts [n]).

Banana solved

Using this knowledge teams could decipher the remaining flavors.

Peppermint solved

Grape solved

Cinnamon solved

Lemon solved

Now teams could use this information with the answer sheet and decode the message. The solution is 'PRAGMATIC'.

Notes

The idea for this puzzle came from the Mooncurser's Handbook Game; this is a different implementation of the original idea for the puzzle "Mosaic." Early on in the year I thought it might be possible to implement a taste puzzle using a different mechanic (identifying base flavors such as sweet and salty) instead of relying on actual flavors, but in practice that didn't work well (bitter was hard to do, folks complained that base flavors was an impossible jump, and some smart-alecs argued that umami belonged in there). At that point people actually asked for it to be flavor identification and since we were hard up for puzzles, I went with this idea.

After the original base flavor idea I tried using fondant as the candy. This didn't work out so well, but when buying the fondant I discovered that melting candy worked well enough for our purposes. From then on things were fairly straightforward. Originally I had planned to use color and flavor, but the grape flavoring turned out to be purple, so I instead made all of the candy purple to mask this and used shapes as the cipher instead. The grape candy still turned out a little darker, but I hoped that this would help serve as a clue as well as give GC a easier way to provide hints.

While the implementation worked out nearly as well as it could, survey feedback suggests that the concerns with the original form of Mosaic still held true; teams complained that they had difficulty identifying some flavors. I had hoped that teams would find one or two that they could get the shape->letter mapping down on and then infer the other flavors, but I suspect in practice few teams did so.