Undercover — a new game to play with your friends!
Our vision
We built the game Undercover to emphasize the team dynamics component in a social mediation game. Players compete with a teammate — without initially knowing who their teammate is! In our game, players must hide their assigned role from most other players, while revealing enough information such that someone with the same role as them can identify them. To do so, they experience the challenge of coming up with interesting questions and clever answers — a component we love from Spyfall. Additionally, because our game is split into even teams, unlike Spyfall where one player is pitted against the group, each player has a more equitable chance of winning — it is less dependent on the role they’re assigned. The variety of categories of possible roles enables players to continually discover new ways to play the game and offers a wider variety of fantastical situations. In the end, we hope for players to have fun immersing themselves and roleplaying the roles and engaging with their peers with carefully crafted questions and answers!
Concept Map
Undercover came from the realization during our ideation process that games of discovery (e.g. where secret information was discovered) and the fellowship that comes from teamwork were both important to us in games. Initially, we thought up a Mod of Spyfall in which there are two locations, and everyone has to guess who’s on which team; however, when another idea came up in which everyone is on a team of two and must find their teammate in order to win, we realized that we wanted to combine the two. This is how “Undercover” came to be.
Creating multiple teams contributes to fellowship, while making them secret contributes to more challenge and opportunity for strategy. Each of our game’s elements was considered carefully for how it would contribute to the overall fun of the game. Our concept map is shown below in which we highlight key game elements, what dynamics they create, and what kind of fun they lead to. Mechanics such as quirks (aka assigned roles), question-asking, teams of two, and the time limits contribute to game play in which players make strategic choices and try to keep their secrets while uncovering others’. This contributes to many types of fun: fellowship, challenge, discovery, and fantasy.
Formal Elements and Values: Initial Decision
After brainstorming various ideas for our social mediation game, we decided on a few key formal elements we wanted to keep in mind throughout the design process. The form of player interaction was important to us, and we chose to use elements from both multilateral competition, where multiple people are competing against each other, and team competition, where players work together. This takes form in our game as multiple teams of two players competing. We believed a team aspect was important to enable social mediation in a group of players, and having multiple teams, rather than just one side versus another, would create interesting interactions.
Another formal element we considered was the objective. Inspired by similar social games we had played, we wanted the main objective to be “outwit.” By keeping information limited at the start, we wanted players to choose their words carefully about how to gather new information without giving away too much to the opponent teams. With multiple teams of opponents and an unknown teammate, it’s a balance of cooperation and deception. At the start, the procedure through which to facilitate this objective was unclear, but we wanted some rules to limit but also guide the flow of information from/to players.
We wanted the types of fun in our game to involve fantasy and fellowship. Players could immerse themselves into the roles they received as well as befriend or thwart each other. Through our choice of formal elements, we wanted to foster these kinds of fun interactions, and our later iteration process occurs with these goals in mind.
Testing and Iteration
Our initial rule system had a single 8-minute round where players scored as a team. Additionally, at any time, a player could call an early end to the game and skip to the scoring portion. However, if they failed to identify their teammate or their teammate failed to identify them, their team immediately lost the game.
During our first playtest we found that people didn’t engage with their teammates and wanted more reasons and opportunities to do so. One team did call an immediate vote and lost the game since one player failed to identify the other. As a result, we decided to remove the ability to call a vote early and instead split the playtime into two rounds where players were allowed to privately message each other between rounds. We also added bonus points for identifying your teammate to encourage more team interaction.
Our next playtest had only 4 players and ended before the midpoint as players quickly identified their teams and refused to give out any additional information. In order to deal with both problems, we decided to eliminate 4 player games and test a new system for asking questions where players pose questions to the entire group and everyone has to answer. We also removed the ability to private message other players and instead let players know their teammate at the midpoint if both teammates correctly identified each other. To further reinforce the team aspects of the game, we made it so teams could only win if both team members correctly identified each other by the end of the game.
Creating the opportunity for players to confirm their teammate at the midpoint was crucial in order to provide players with feedback and enable a more interactive game-play. Additionally, we thought it would give teammates an ability to strategize together.
We then each held our own playtests with our friends and families and held one more in class where we experimented with question styles (popcorn-style vs. one question to the entire group) before making any large changes. These playtests were also useful because we had larger groups of people (8 instead of 6) in two of them, and had the opportunity to try out the “midpoint teammate confirmation.” We observed that the midpoint team confirmation mechanic was useful because it gave the group feedback on their progress: in the 8-player games, most teams did not guess their teammate at that point of time, but the few teams that did could deliberately mislead other teams or work more towards being vague and discovering others’ identities.
From these playtests, we also decided to continue using the style that poses the question to the whole group because it provides the players with enough information to have a better chance at winning. In our playtests with the popcorn-style questions (Person A asks a question to Person C, then Person C asks Person X, etc.), many teams did not find their partners and lost, most likely due to the little amount of information going around since only one person is answering each question. Before the final playtest, we also decided that during the first round, players would ask questions to the group in a predetermined order until each person had asked one question, and during the second they could do the same questioning in any order but with a 3-minute time limit.
The majority of our playtesting and iteration was focused on encouraging players to engage with the team aspects of the game and ensuring that players revealed information to each other even after identifying their teammates. As a result, the game’s structure and scoring changed significantly during its development.
Final prototype and Rules
The rules for Undercover can be found here. In this game, we utilize a moderator: the moderator should assign roles from the chosen theme, monitors the time limit, and communicates with players. The categories and roles that can be assigned, along with sample questions, can be found here.
A video of our final playtest can be seen here.
For the first game, we’d recommend one of the following themes: Occupations, Animals, or Kingdom. For a real challenge, try playing Colors.
Design and Marketing
These are some screenshots of what we envision our game Undercover to look like if it were actually implemented.
