“Blood on the Clocktower” (BotC) is a game created by Steven Medway. It is a physical tabletop social deduction game. Its target audience is adults and teenagers (14+) who enjoy complex social deduction, storytelling, and group play with lots of player interaction.
BotC has its own take on social deduction by creating a novel “information economy” where players occupy information-asymmetric knowledge roles. While our game “Odd Ones Out” (OOO) also uses information asymmetry through its prompt system, BotC shows how when partial information is carefully distributed, it can create a more layered and dynamic player experience that keeps players engaged even after they’re eliminated. By looking at BotC through the lens of formal elements, especially its MDA framework, we can change our prompt-based system to create a more multidimensional information system without sacrificing the accessibility of OOO.
BotC’s spark lies in its well-crafted “information economy.” Each character role provides uniquely partial information about the game itself, creating a “lens of mystery” – players must piece together fragmented knowledge to pick out threats. During our playtest session, Kimi, an experienced player who introduced me to the game at first, said, “I love feeling like I’m a piece of the puzzle every time I play Blood on the Clocktower. One analogy my friend said what that it’s like everyone’s looking through a pinhole at a different part of the same room, and I really think that’s true.”
Analyzing BoTC through the MDA framework shows how its mechanics (unique character abilities) create compelling dynamics (information sharing, deduction, bluffing) that create and support aesthetics of fellowship, discovery, and dramatic tension. The game was coherent per the MDA paper – every mechanic serves a specific purpose that supports the intended experience.
In contrast, “Odd Ones Out” distributes information along a simpler binary – real vs. imposter prompt. This creates a clear division between the information groups but lacks the varied information patterns that makes BotC so replayable and fun. While our approach is more accessible for newcomers, it sacrifices some of the group information gathering dynamics that sustain interest over many playing sessions.
BotC’s information distribution system creates many overlapping feedback loops where each revelation can shift how knowledge is perceived. Sara, a first-time player, noted in our playtest, “I felt like I had to share what I knew but also hold back some details that revealed my role. The information I had became some expensive resource that I had to look over carefully.”
From the framework in “What Do Prototypes Prototype?”, BotC primarily prototypes “role” (function the game serves in players’ social lives) and “look and feel” (concrete experience of playing), whereas our game concept emphasizes accessibility of implementation for newcomers. The role prototyping in BotC is especially strong – the social experience formed makes information management (both as a group and an individual) the central activity, which our game come benefit from developing further.
I think BotC’s most innovative contribution to social deduction games is how it handles player elimination, or in the specific case of the game, players dying. In traditional games like Werewolf or Mafia, players become spectators once they die. In BotC, “dead” players remain actively engaged through the “ghost” mechanic, which allows them to receive limited information and subtly influence living players. This solves the “player elimination problem,” where early deaths can lead to players becoming disengaged.
Marcus, one player in my playtest, played Mafia the previous day. He noted, “Being eliminated early in Mafia sucks because then I just sit there. Here, I actually got more involved after dying. I had secret info about who killed me, and I could help direct the townsfolk to the devil.”
This solution sounds like what happens to Gin Rummy when you remove the core procedures: the game stops functioning as intended. BotC recognizes that when a player gets eliminated, it can damage the overall experience, so it subverts the traditional elimination mechanic while preserving the dramatic tension that “death” imparts. It’s a perfect adjustment to a formal element that changes the player experience.
OOO currently uses a traditional elimination system where voted-out players cannot speak in future rounds. This creates tension for remaining players but risks disengagement (the problem that BotC solved). BotC’s ghost mechanics directly support the aesthetic of fellowship by ensuring all players can meaningfully engage throughout the whole experience.
BotC, among the aesthetic categories we discussed, emphasizes discovery, fellowship, narrative, and challenge, while OOO currently prioritizes challenge, expression, and submission (as a pastime). If we incorporate post-elimination engagement mechanics similar to BotC’s ghost mechanic, we could strengthen the fellowship and discovery aspects of our game without compromising its accessibility.
BotC requires a dedicated Storyteller (moderator) who runs the game, provides the narrative context, and oversees the rules. This creates an immersive experience but requires one player to remain outside the main game. The Storyteller’s role changes BotC from pure mechanics into something closer to theater. After being the Storyteller, Sara mentioned, “It felt like I was directing a show or writing a script. I had to set the mood, come up with a funny or serious story, and remember so many things. It’s really fun, but it is a lot.”
This moderation structure shows how game mechanics themselves can convey meaning and foster emotions. The Storyteller doesn’t just keep gameplay going; it can change the nature of the game itself by creating an author-like presence with a say on the narrative and flow.
OOO uses a moderator as well, but our design creates a more self-running structure where the moderator mostly distributes materials instead of changing the narrative. This shows different design priorities: BotC emphasizes richness in the narrative and complex information flow, while OOO emphasizes accessibility and quick play cycles.
BotC, like many social deduction games, creates an environment where lying becomes not just allowed but celebrated as a skill. This suspension of social norms creates an interesting space where players can safely play with deception.
“It’s so interesting how BotC makes me think my best friends in college are untrustworthy,” Kimi said in our debrief. “But with time, you can learn to separate evidence and assumption. I even think I’ve gotten better at figuring out when someone in real life is lying to me because of these games.”
In our playtest, I observed some patterns where some players were believed or not believed based on factors unrelated to the gameplay – including speaking style, gender, and assertiveness. Marcus noted that “everyone immediately trusted Sara even when her information seemed contradictory,” but not everyone immediately trusted him. This raises an important series of questions about how social dynamics influence gameplay beyond the formal rules. OOO faces similar considerations, especially in how prompt design could help certain players based on their knowledge domains or how they communicate.
BotC’s role-based system creates a rich system with really interesting and diverse player interactions. The Empath knows how many evil neighbors they have, the Flowergirl learns if a Demon voted, and the Investigator identifies two players, one of whom is the Minion. This creates so many incomplete webs of knowledge.
OOO has a more accessible but less varied information dynamic through its prompt system. All players with the “real” prompt share the same information, while imposters get a different but similar prompt. This division creates a very different gameplay experience that focuses on picking up on subtle incongruities in players’ responses instead of piecing together fragmented information.
Based on what we’ve learned while comparing OOO to BotC, there are some opportunities for how we can continue to develop OOO. One way of doing this is introducing secondary information channels beyond the main prompt system. Maybe some players could receive additional context about their prompts that others don’t, and this could create more varied information states without making the game that much more complex.
Addressing the player elimination problem, we could implement a variant of the ghost mechanic where eliminated players contribute limited information to remaining players. This could be done through things like written notes or one-word hints.
We could also develop a more structured framework for moderators to help build narrative tension while keeping the game accessible to new players. Maybe we could develop story packs in our game that use a series of prompts that all center around a common theme or plot, like Mad Libs.
Rather than strict “real” and “imposter” prompts, we could create prompts in the middle of the binary with varying relationships to each other, like how BotC creates complex relationships between roles.
We could also introduce a more nuanced feedback system that keeps the game balanced even as players are eliminated, maybe by giving eliminated players abilities once they die that keep them engaged while keeping the competitive tension.
BotC shows that a truly good social deduction design is not based in hiding information but rather strategically fragmenting it across many players. This demands collaboration, deduction, and careful communication. The goal of OOO is to create tension and foster collaboration in a format that “non-gamers” or casual players could enjoy without a long explanation or a deep understanding of the rules.
Our key insight is that we do not need to replicate BotC’s complexity, but we can learn from how BotC strategically fragments information to create a more nuanced experience while keeping the game accessible. We want to make OOO into a game that keeps its accessibility but enriches its information dynamics.
