Project 2: Group 7 Final Report

Artist’s Statement 

We created our MVP of The Stanford Files to reimagine one of Stanford’s strangest historical moments: the mysterious death of Jane Stanford. Long passed down as a piece of unofficial campus lore, the story has become something incoming frosh and graduating seniors bond over every year, and a key “if you know, you know” story told on campus. Our goal was to formalize that narrative into a reimagined mystery game that frosh could play during NSO to both familiarize themselves with their new home on campus and uncover key elements of Stanford’s history, layout, and culture that they otherwise may not be familiar with. Players take on the role of an investigative team solving a campus-specific whodunnit using analog physical clues and a companion web portal. The portal serves two purposes: onboarding players into the premise of their role, and offering a deliberately minimal, command-line-style hint system designed to help players push past roadblocks only when it impacts their ability to continue with the game, without disrupting their immersion by having any distracting elements.

Using the MDA framework, we focused on discovery and challenge as our core types of fun. We wanted to offer players the pleasure of piecing together smaller pieces that together form a story if you pay enough attention to the details; from the autopsy report and TikTok timestamp to the suspect interviews having ever-so-slightly conflicting versions of events. Emotionally, the tone we were going for leans into quiet unease over horror, inspired by the narrative tension of games like Unsolved Case Files or Clue, where tension builds with bit-by-bit reveals of information. Given that our game is centered around a single plotline, the game is playable only once (again, taking inspiration from Unsolved Case Files) and we, over time, iterated to make it a multiplayer game since the real momentum only picked up as players collaborated and brainstormed hints and storylines side-by-side. 

The game blends tactile materials (case files, receipts, maps) with a retro command-line interface that dispenses hints. This hybrid format supports nonlinear exploration and rewards players spending as much time as possible immersed collaborating over the actual clues instead of in a digital setting, using the help portal only to get any key pieces of information necessary to keep the game moving along. Ultimately, The Stanford Files isn’t just about solving a murder, it’s about the thrill of collaborating with new peers to uncover how legacy, resentment, and truth blur in the stories we inherit when we call new spaces home. Our hope is that players leave not just with new lore, but with new connections during a critical and formative period for all new Stanford students.

The spatial layout of campus played a central role in our game design. Every artifact in our file is grounded in real Stanford/Bay Area locations. We wanted to transform familiar landmarks like the Chem Building, White Plaza, and Tresidder into narrative spaces where players can walk past and recall a narrative unfolding. In that way, the game uses space as both a storytelling tool and a narrative setting.

System Model + Design Artifacts

System Model:

Concept Map:

Interaction Flow:

Formal Elements & Design Decisions

Core Mechanics, Rules, and Player Roles

The Stanford Files is a collaborative mystery game where 1–5 players take on the role of detectives trying to solve the murder of Jane Stanford. Players begin with a physical case file labelled “CONFIDENTIAL” containing several clues, which include an autopsy report, suspect depositions, newspaper articles, a hand-drawn map, and other evidence like a dry-cleaning receipt and TikTok screenshot. The game has no strict turn order. Players are encouraged to interact with each other, discuss theories, and analyze clues together. The goal is to collectively piece together a timeline, eliminate suspects, and identify the murderer.

We also have a website that sets the initial scene. Players can also use the website to ask for hints associated with specific documents and to and submit their final guess. If their answer is incorrect, they have a one minute lockout period before they can guess again, which discourages them from guessing randomly. If they guess correctly, they are rewarded with the full story about Christine’s (the culprit’s) motive, opportunity, and method of poisoning Jane Stanford.

Win Condition

The game ends when players correctly identify the killer. There’s no “lose” state, since if players submit the wrong answer, they can try again after one minute. This choice aligns with our goal of accessibility and group fun rather than competition.

Designing for Interaction and Experience

This game is designed around collaboration. We intentionally avoided rigid structure to encourage organic conversation. We wanted the physical evidence folder to evoke the feeling of an actual investigation, so we wrinkled papers, clipped on suspect photos with paper clips, and ripped certain clues. We also observed that roleplay emerged naturally during playtests. One player would act as detective, while others performed as suspects, bringing the characters to life and making the experience more memorable.

We also structured the game around a mechanic of elimination. Instead of requiring players to deduce the killer from scratch, we included clues that each rule out suspects one by one. This decision was based on several MDA Framework principles, including the mechanic of evidence driven deduction supports the aesthetic of narrative discovery and the dynamic of group puzzle solving. It also creates a more standardized pacing, as each clue narrows the list of suspects, the game builds momentum toward a clear conclusion.

Audience Goals and Values

Our target audience of Stanford admits and frosh, is reflected in every element. The characters include real and fictional figures from campus culture: Bertha Berner (Jane’s real-life secretary), the Stanford Tree, MTL, and even the Fizz CEO. Every suspect deposition blends historical facts and humor, which introduces students to key elements of Stanford lore. The evidence also reinforces these values. The crossword includes clues about Arillaga, Tresidder, and the Oval, iconic campus landmarks.

Design Framework Justifications

From a formal systems design perspective, we prioritized clarity, immersion, and engagement:

• Mechanics like clue sequencing and suspect elimination support cognitive scaffolding, helping players build mental models as they go
• Aesthetics are built on narrative and surprise, satisfying our audience’s interest in lore and puzzle solving
• Dynamics like discussing theories and collaborative debate emerged from our light rules and open ended structure, intentionally reinforcing social play

Our hint system also follows scaffolded learning principles, providing support while preserving challenge. By limiting hints to specific artifacts like the crossword or map, we ensure that players stay focused rather than overwhelmed. Every design choice, from the wrinkled folder to the suspect list, was made to serve our audience’s curiosity, interest in Stanford-specific lore, and collaborative, engaging experience.

Scope of the Game (MVP vs. Slice)

Stanford Files is an MVP game because of how the central idea of the game is captured through many presentables we have in our mystery game. In our game, players receive a file folder of various items that are used to figure out the murder of Jane Stanford. Alongside this, players are given a companion website that introduces them to the plot of the mystery game and gives them the opportunity to ask for hints based on each item received in the file folder. For a game to be considered MVP, user testing and feedback needs to be possible. Because of how Stanford Files deals with lots of interactive content and has an end goal where the player inputs who they think the murderer is in the website, our game is capable of receiving insights from different players making it ready for feedback. 

Due to the fact that our game prioritized simplicity in regards to Stanford Files being a single solution game mystery, one tradeoff that needed to be made was replayability because the mystery game can’t be replayed once the mystery is solved. While this was a big decision to make, we believe that our game having one solution adds to the focus of it being an MVP product and also makes the player’s goal clear. In addition to this, one big constraint that could come up with our game is how it might be only playable for students that attend and/or are knowledgeable about Stanford because of how the game deals with Stanford niche information. However with that in mind, an ambition for a full release version would be making the clues given in the file folder virtual so that players can virtually “visit” the locations corresponding to certain clues to gain a fuller understanding.

Testing and Iteration History

In the initial version of our game, we included five main components: a crossword, a deposition, an autopsy, a news clipping, and a map. We also provided a digital website where players could get hints for each component. At this stage, we were still exploring how players should deduce the murderer. We debated between two models: having each component eliminate a different suspect independently, like levels of a game, or requiring players to combine clues from all components to narrow down the suspect pool. We were also unsure whether the game should end with a definitive answer or an educated guess. With these questions in mind, ultimately, we chose to include extra suspects and structure the game so that players would need to make an informed guess. We also attempted to design each component to eliminate a suspect in isolation.

In our first playtest, we found that players struggled to piece together the clues. Many described the experience as “disorganized” and “too challenging.” The map was especially confusing because it displayed the entire Stanford campus, which overwhelmed players and left them unsure of where to focus. The crossword puzzle also proved difficult; players said the clues required insider knowledge that the average Stanford student would not know. Furthermore, many players forgot to use the website, instead turning to the moderator when stuck. Based on this feedback, we removed unnecessary locations from the map, simplified the crossword clues, and reaffirmed our decision to make the game end in an educated guess rather than a definitive answer.

Crossword: First iteration

Crossword: Final iteration

Newspaper: First iteration (establishes stronger motive for Christine and contradicts statements in her deposition)

Newspaper: Final iteration (eliminates MTL as a suspect)

In our next iteration, we decided to finalize that the players should have a definitive answer to who the murderer was by the end of the game. We also decided that the clues should be used in conjunction with each other, so players should piece together specific details from each component to deduce the murderer. To aid in this deduction, we also added more evidence to the game. We added a LinkedIn post, Dry Cleaning Receipt, and TikTok screenshot with specific time stamps to eliminate the Fizz Ceo, Bertha, and Stanford Tree. Our goal with adding these pieces was to help the player have definitive proof that those suspects were not the killer. 

New evidence:

In the next playtest, we received feedback that now, there was way too much evidence, and the players felt overwhelmed, as they didn’t know where to start looking. We also found that players still neglected the website, as they would often look to the moderator when stuck instead of the website, as they would “forget that it existed.” On the other hand, we got positive feedback that the crossword and map were easier to navigate. For the crossword, we made the clues significantly easier, and for the map, we only put the necessary locations on it. We still struggled with integrating those components, as people would not pay attention to the map because “they already know what Stanford looks like,” which is a goal we aimed to resolve in the next iteration. 

In our third version of the game, we decided to give players the backstory on the website by having them click through it. This served two purposes in our game. The first one being that it encouraged users to use the website throughout the game to understand how to navigate the puzzles. Additionally, it put less reliance on the moderator, who was previously the one who explained the backstory.

These changes helped us see tremendous improvements in the game overall, as players reported that they had a better understanding of the starting point for the game. This is also the turning point when we saw players consistently be able to solve the mystery and find the game enjoyable. As finishing polishes for the game, we decided to bold key words in each evidence component. We did this to address the feedback that the amount of evidence felt overwhelming at times. We also decided to make a trail on the map that the players had to follow, which helped better integrate the map into the game while also giving the players another clue as to which suspects to eliminate.

In the final iteration of the game, we added the crossword to the newspaper clipping. This helped the game have cohesion, as we received feedback that the crossword felt out of place with the rest of the other evidence. We also decided to add a paragraph of hints for the crossword to the newspaper, as this would still allow players to solve the puzzle even if they don’t have the initial knowledge to solve the clues. For the digital component, we also made the players confirm on the website when they are done with the clues and ready to guess who the murderer is. We did this because we received feedback that the boxes for getting hints and guessing the murderer looked very similar, so players may accidentally guess when they are not ready to. For aesthetics, we added a flashlight, a Stanford hotline card, a highlighter, a pencil, and pictures of each suspect to truly make the player feel like they are a detective solving a true murder case.

Game Delivery 

In our game Stanford Files, we have an analog game with an additional digital prototype to aid the user’s gameplay experience.

Digital Prototype Add-on

Link: https://zionasemota.github.io/Stanford-Files/

Our digital prototype is a website that includes an onboarding of what the game entails, hints that the player can receive based on different clues given in the file folder, and a text field that allows players to input who they think the murderer is after analyzing all the data given. At the end of the website, the player will find out the whole story behind the murder only if they guess the murderer correctly. This website was built using HTML, CSS, and Javascript.

Analog Game

In Stanford Files, the player is given a Stanford tote bag with different items that are usually given to students during NSO(New Student Orientation) at Stanford. In addition to these items, the user is given a file folder with different clues regarding the murder of Jane Stanford. Below is an embedding of all the clues given to the user to allow them to solve this mystery!

Playtest Demonstration Video

In this video, we follow a complete start-to-finish playtest of our final game with three playtesters. It includes multiple moments of fun as they debate the clues and discover hints about the suspects, along with a couple moments of friction where they misinterpreted or missed a few of our props as communicating hints when they were only there for special effects (although, this inspired us to consider using decoy hints!). 

Thank you to all the TAs (especially Amy, our TA!) and Christina for an amazing quarter:)