Paper Prototype Play
Testing a video game with no code
The Arcadia Report is an immersive murder mystery. Great! Except I’m not a programmer and most of my game design experience comes from the tabletop, not the screen. So how do you tackle designing a reality-invading mystery game? In my case, it’s by applying the same design process you might in any design project. First things first, get the mixed up threads of The Arcadia Report out of my head and into some kind of order.
I had been brainstorming story elements, gameplay ideas, and more on and off since I pitched the project so time to go back to the basics and start again from the ground up. Based on my user research results and the initial personas, plus the emotions and story ideas I have been brewing, I went back to the core in an after hours whiteboarding session:
The Arcadia Report…
- …brings the mystery to life
- …is always on
- …makes you question reality
Users want…
- …to feel like a detective
- …twists
- …suspense
Based on those core ‘truths’, I start forming some ‘How Might We’ questions:
How Might We…
- …make people feel like a detective?
- …make it feel real?
- …play a fake mystery out in the real world?
- …create twists the player is an active participant in?
- …make solving a mystery a process anyone can participate in?
Taking those HMW questions, I started plotting out the story and potential ways this could play out from that perspective, rather than just scribbling out growingly convoluted storylines as I had been in my sketchbook. In the end, I drilled in on one concept which takes the central conceits I had developed around The Arcadia Report as a kind of ‘mystery database’ and spins it on its head to put the players, rather than the mystery’s characters, as the prime investigators.
Whether this new refinement will work out in practice remains to be seen, I need to explore it more. However, it did help me reduce it to base components to develop an initial high level story overview, potential mystery flow, and options for how to break this down and test it.
Focusing in…
Armed with the core elements, I chose three of the main How Might We questions to focus in on and started with a clean board:
How Might We…
- …make people feel like a detective?
- …make the experience feel real?
- …make solving a mystery a process anyone can participate in?
I had an extremely rough story arc sketched out from earlier in the week and began by trying to reexamine that. However, I soon paused to once again drill down to the basics. What is a case and how does it work? How does it make the player feel like a detective? How does it feel real?
Basic Premise:
- The Arcadia Report is a corrupted database being reconstructed from the ashes of a startup company which was trying to democratise justice and build an algorithm that can pinpoint murder suspects.
- The mysterious ‘Reporter’ is reconstructing the database and calling upon you, the public, to help get justice as people with connections to the former Arcadia company begin to die.
Working off those, I mapped out a potential flow for how the mystery could work:
Case flow:
- Case launches on The Arcadia Report website, highlighting victim, the suspect the police missed, and a basic case overview.
- You can search the database, following whatever leads/keywords you want from the case, and/or travel to real world locations using clues from the overview or that you find in the database.
- Once you have a theory for what happened, you can select the three key clues/evidence to outline motive, means, and opportunity that connect the suspect to the victim.
- Submit your report to the authorities.
The idea is that these steps will take on a new significance as you piece together the story and begin to question the reality presented to you but for now, this is a simplified spin on the basics of the detective genre that I feel might just work. Now to test!
Prototype Plotting
I’m not sure there is an aspect of my project that does not present a new challenge for me. Even prototyping. Just when I was thinking I was making some progress on how this whole ‘Arcadia Report’ thing would play out, I realised I had no idea how to test this new concept. Who is to say it is even any fun or evokes the desired emotions at all? Time to brainstorm some more!
Breaking the concept down into stages based on essential components, the first step is the core mystery solving loop. This + the narrative are ‘make or break’ elements along with the location hopping (which is a way an outworking of the core gameplay). It needs to be simple, fun, and make you feel like a detective.I could code a prototype but then I’d have to 1) design a test mystery, 2) figure out how to code it, 3) do said coding, 4) create any necessary placeholder clues/assets. And all that time and effort invested before I even know if the idea is sound and enjoyable for users.
So going back to my user flow, and drawing inspiration from my tabletop gaming background, I thought why can’t I do it be a paper prototype? Two suspects, some clues, then I can act as ‘the machine’ guiding testers through and seeing how they approach the problem. If the gameplay doesn’t work then all I’ve lost a few hours and some post-its to gain some valuable feedback.
Time to break out the paper and pens!
Prototype
Ingredients:
- Sharpies
- 2 A3 sheets
- Pad or two of post-its lying around the office
- Ruler (optional)
- Murder on the mind (fictional, of course)
With that handful of art supplies, I created and wrote a mystery prototype in under an hour that would have taken me days to code. Is it visually the same? Nope. Does it encapsulate all of the essential actions available in the core gameplay? Yep. It’s the simplest proof of (gameplay) concept I could ask for.
How it works:
- You have a suspect, a victim, plus a basic profile and case details.
- Decide which leads to follow and piece together the clues from your search (in this case represented by stacks of post-its serving as a proxy for the answers that would have resulted from your database queries or travelling to actual locations).
- Connect the dots to find the motive, means, and opportunity to prove your case.
Next stop: ambushing some testers!
Paper Playtest
So here we have my paper prototype mystery before and after ambushing a user and an example of how it went down when a gamer from my target audience took it for a trial run. I acted as ‘the machine’, providing search results from the post-it stacks but after following one or two leads they started searching through the stacks on their own.
How it went down:
I gave a short preamble to set the scene for the mystery and establish the goal to find the motive, means, and opportunity to prove your case. After that, it was over to them. The mystery itself was a hastily written thing off the top of my head for the purpose of the prototype, drawing in two of the characters I had been working on in my story planning but without any overarching narrative. Sometimes watching too much Columbo as a kid pays off.
How they played it out:
- Went to the crime scene first.
- “First I’d try and follow the exact steps [of the victim].”
2. Retraced the victim’s steps.
- Wanted to search coffee place/what was bought/details mention (which wasn’t possible).
- “I would search online for report/news to get info about death.”
- Would be satisfied by autopsy report in case overview.
3. Searched database for medical records
4. Searched database for details on the victim.
5. Followed link from victim’s emails to search up details on the suspect.
- Instantly latched on to the suspect’s calendar appointment for coffee, coinciding with the victim’s.
- Wanted to search online for scandals around the company to see what they were arguing about.
6. Guessed the case that the suspect was arguing with the victim about business/money [motive], took advantage of the victim’s peanut allergy [means], to doctor their coffee during their meeting [opportunity].
Feedback and notes from after the test:
- “The three things [means, motive, opportunity] are like the keys I had to find in an escape room”
- Could be cool to have to notice a pattern (ex: clothing person was wearing, identify them based on another photo, etc.) to finda clue.
- Maybe you have to use tools to understand a clue? (Ex: infrared scanner, invisible ink, your wits, etc).
- Could there be particular channels for each clue? Ex: some are found with a particular ‘tool’? Or need to use a different channel to find each of the motive/means/opportunity (ex: one is found through location, one through database, one through online search).
- Mix up puzzle/clues you need to connect dots: visual, numbers, words
- Could know you found a clue but not where it goes (is it motive/means/opportunity? And for which possibility?)
- Puzzles can be good, especially if you’re not a huge mystery fan or struggling to put together clues.
- Corrupt files leading to other avenues as you find certain clues so people don’t get lost?
- “Make it obvious when you are on the right track”
- Are there options for guessing (finite number)?
- Waiting for that ‘thumbs up’ moment to make the accusation.
All in all, about an hour to make the prototype and another 20 minutes or so to complete this first test. Well worth it!
Biggest lessons:
- It doesn’t have to be puzzles but a bit more interacting/searching to understand the clues rather than being spoon fed information as soon as you reach a certain search/location would make the experience more engaging.
- ‘A-ha’ moments don’t have to come from impossible mysteries. It can just be that moment all the pieces come together.
- Make it obvious when you are on the right track.
- Use the three elements of the case you have to fill in as a way to guide the user towards a solution and help if they get stuck.
- Don’t let seeming complexity scare you off: you can always create a basic prototype in some shape or fashion.
Until next time, stay safe.
Thanks for reading! The Arcadia Report is an immersive murder mystery experience, transforming your city into an escape room. To follow along with the development, subscribe to The Redacted Report:
And join the Citizens Club for exclusive updates and help shape the design:
