## How It Works

First, collect the data from the spotlight shadow. The expert cutout and his shadow are demonstrating two different semaphore letters. Gather the missing letter mappings. Observe that the mapping is reversible which helps to confirm the correct data has been collected.

Second, realize each paper is a maze. The text at the top when mapped through the data from the first step says "START" and the text at the bottom "END".

Third, figure out how to traverse the maze. This requires some experimentation to figure out the details of how to use all the data that is available:

• The first step provided a substitution cipher.
• The puzzle previously used semaphore.
• The spotlight's shape is an octagon, just like each maze cell.
• The shadow's arms point at sides of the octagon.
• The letters are rotated within each cell.

The correct traversal is to use the substitution cipher to get the shadow letter, then convert the shadow letter to semaphore, and then rotate the semaphore by the letter's rotation.

Fourth, observe that the path through each maze uses every cell except three. Three connecting cells is enough for a semaphore letter. Extract the semaphore letter from each maze and use the substitution cipher one last time to arrive at the final answer word.

AMULET

## Design Notes

Jon knew he wanted to design a technology puzzle that was a single installation because he did not want to configure twenty laptops this year. So he was on the lookout for a puzzle that would either involve no input at all or a single input all, or atleast many, teams could share. Fate brought that to him in the form of a fire alarm. He was testing the 2011 Puzzle Hunt when the fire alarm went off at 1am. Microsoft fire alarms are a great way to re-energize the team's adrenaline! The fire department arrived and started casting shadows everywhere with their lights which solidified in Jon's mind that the puzzle should involve people's shadows.

Almost every playtester wanted more complexity to the puzzle. Some wanted a more video game-like experience where the team had to provide an expected semaphore letter quickly. Others wanted to return to the Kinect and spell out the word from the maze to get a more final answer. Hopefully teams appreciated the unchanged and relatively straightforward puzzle at 1am.

## Construction Notes

This puzzle's complexity was entirely in the software. You may not have noticed but the spotlight was physically correct for how a spotlight would look if the tripod device was a real spotlight—even though the actual projector was thirty feet away and skewed. This was accomplished via Math. There is a standard technique in video game rendering called shadow mapping where the scene is rendered from the perspective of the light in order to determine if a point is in shadow from the perspective of the camera. This puzzle used that techique in reverse. The software knew which points were in shadow from the perspective of the Kinect camera and used that to determine which points needed to be in shadow from the perspective of the light.

A few days before the event, someone suggested we should give teams something to commemorate their acceptance into the Training when they left Leavenworth. How convenient that this puzzle was at the right place, at the right time, and with the right answer word. An amulet design was quickly put together and a day later there were nearly two hundred amulets (but we called them pendants in case a team which had not solved yet overheard).

## GC Notes

Many teams ignored heavy handed staff instruction to face the tripod device and attempted to stand with their back to the Kinect. This was problematic for two reasons: 1) Kinect can't detect if a body is facing forward or backward so the software had to assume you were facing it in order to decode and encode the semaphore correctly, and 2) Kinect works significantly less well when your arms are mostly hidden by your body. Teams attempting to be clever by mirroring their arms to compensate looked frustrated.