Final Project: PhiPhi's Quantum Tuna Cans

Yuke Ding and I spent a little over a month developing this little project – a friendly demonstration of some fundamental ideas of quantum physics using a (very much alive) cat and some quantum tuna cans.

See some video demonstrations below, as well as notes of our brainstorming throughout our project on Ding's Notion page.

A technical description of the project and the interactions therein follow:

The lid of a tuna can attached to a rotary encoder behaves as the controller for the game. An Arduino Nano IoT 33 sends rotations and button presses from the rotary encoder to Unity3D through serial communication. A digital interface made in Unity3D responds to both button presses and rotations from the rotary encoder. Short and long button presses are used to change between different sections of the project, i.e., instructions, the game, and background information about physics. Encoder rotations influence the probabilities of positive or negative outcomes, which are reflected through animations and graphics of the cat PhiPhi.

Instructions onscreen, presented using both animations and simple text, outline the interaction’s objective — to feed PhiPhi. They introduce the main loop of player actions using the larger tuna can-turned-button in front: a short press to start the game, followed by rotation to increase the probability of reward, and then another short press to calculate the result based on this probability.

Players who choose to participate will rely on PhiPhi’s reactions to the tuna can’s orientation to achieve the most optimal outcome. Visual cues will signal to the player whether they will achieve the best possible outcome. After seeing the result, and returning to the first screen, the player will have the option of repeating the interaction. They may also choose to learn about the physics background of the interaction by long pressing the tuna can.

Players who choose to explore the physics background are able to read and reread a set of informational graphics on quantum mechanics and how it relates to the interaction. Players can interact with most of these graphics by rotating the can. They can choose to return to the interaction with PhiPhi through another long press.

Playtesting

In our last week, we spent most of our time responding to player feedback we received from our classmates and instructors. Below I list their comments and observations we made of their interactions, and changes we made in response.

• To our surprise, a few players went straight for the physics explanations. One even mentioned wanting to learn more than what we'd written, i.e., some real applications. They suggested making these informational slides interactive.
  • We made some slides interactive, and added a few more terms that may pique one's interest to learn more.
  • We separated slides into single sentence chunks, each with an associated image.
• For many players, their "press" was what our code read as "long press", as a result, they began with the physics explanation without meaning to. One player suggested using different words to describe each press, e.g., "click" vs. "press". Another player commented that they normally expect a press to activate the instant they press, rather than when they release.
  • Since a lot of the logic of my code depended on keeping track of when the button was released, I decided to create a visual indicator of when the user has started a long press.
• Generally, people liked the feeling of pressing down on the can and rotating it.
• Some players didn't read the directions. One made a point of pressing every button on the keyboard because he saw the word "press" on screen and ignored the rest. He was being a bit of an annoying shit, but made good points that there were too many words.
  • We replaced the introductory directions with a looping animation of the basic movements - press, rotate, press. It's not clear yet how well people respond, though. It might be better to display all three steps in picture form at once.
• The relationship between probability and rotation was not clear to some players.
  • This is a challenge of coming up with a good metaphor. Perhaps our physics slides can help players understand a little more?
• The paw mark is not centered on the ideal location to click the can.
  • We changed this to a finger, suggesting to press on the can. I personally like the paw mark and would like us to bring it back.
• One user was distracted by what they already knew about Schroedinger's cat, and asked why they could see the cat, even though the cat is not the subject of quantum effects. They raised the point that players with some prior knowledge might make certain assumptions about what the experience should tell them, even though they're wrong.
• The results page had too much information at once, and players weren't sure what to do with the volume of info, or whether they could do anything to change it.
  • We separated the results section into a "reveal" and "report" page. Reveal shows the player their result, and report summarizes information about the probability. I noticed players still want to click through the reveal stage instinctively. Players also try to get 100%, even though it's not that meaningful in terms of the metaphor. Perhaps there can be some kind of special reward for getting a 100% change.
• Players did not notice the hunger meter at all!
  • We got rid of the mechanic. It made sense given our decision to no longer gamify the interaction - that is, players were no longer being timed. That said, there may still be some room for using time to add challenge - for example, randomizing the "ideal" angle every five seconds or something similar.
• It wasn't obvious that the cat was approaching the can.
  • We added a small version of the can onscreen. This can also doubles as a visual signal to rotate the can in the Play phase.
• It is most people's intuition to rotate through the information displayed in the physics slides.
  • We still have players click through, but include interactions that activate by rotation. Maybe everything can work via rotation, and every time a threshold is passed, there's a transition into a new slide.