For my P1 project, our team is designing a bluffing and deduction game called “Word Clash Undercover,” inspired by the Chinese party game 谁是卧底. In this version, each player receives a secret word—most players receive a common word (e.g., “apple”), while one player (the spy) receives a similar but different word (e.g., “orange”). Rather than taking turns speaking, players participate in a high-pressure “word duel” phase.
During this phase, players have 10 seconds to write down a descriptive word on paper (at least two characters long). They then have 30 seconds to flip and reveal their word, competing to be the first to do so. If multiple players write the same word, only the first to reveal it stays in the game—later duplicates are eliminated. After this, players discuss and vote to eliminate one suspect. The spy wins if they survive until the final two; civilians win if the spy is eliminated.
Below are several design questions I aim to answer through early prototyping:
1. Is the real-time “word reveal” mechanic fun and understandable for players?
Why this matters:
This is the heart of the game. If players don’t understand or enjoy this mechanic, the pacing and appeal fall apart. It tests look and feel: the tactile, emotional layer of fast interaction and bluffing.
Prototype approach:
A paper prototype using word slips, timers, and live play with 5+ players. No digital implementation needed.
Prediction:
The mechanic will generate tension and excitement. Some players may experience light confusion or anxiety on the first round, but likely adapt quickly. It may become a standout feature if well balanced.
2. Do vocabulary overlaps feel strategic or unfair?
Why this matters:
Eliminating players who write the same word (except the fastest) adds risk. This mechanic could feel either clever or arbitrary. It connects to role and look/feel, as it affects both gameplay and how fair the experience feels.
Prototype approach:
Run multiple test rounds using common prompts (e.g., “fruit”) to track overlap frequency and player reactions. Observe both behavior and emotional response.
Prediction:
Players will enjoy the tension but may need clearer rules or visual feedback to accept elimination as fair. If too many eliminations feel unintentional, a refinement (like “grace rounds”) may be needed.
3. How similar should the spy/civilian word pair be to create meaningful deduction?
Why this matters:
The bluffing and reasoning depend on ambiguous overlap between the two words. If the words are too close, the spy blends in easily; too far, the spy is exposed. This directly relates to the role dimension—what the game asks of the player as a social reasoner.
Prototype approach:
Test different word pairs and collect data on whether players felt “safe” or “obvious” in their descriptions. Ask players to rate how hard it was to detect or hide.
Prediction:
Best results will come from pairs in the same category (e.g., “river/lake”) that invite subtle wordplay. Abstract or poetic pairs may reduce clarity and slow deduction.
4. Do players use timing behavior as social deduction clues?
Why this matters:
A subtle part of the design is the behavioral signal from when a player flips their vocab card. This is part of the “tactile interaction” emphasized in the video and relates to look and feel.
Prototype approach:
Observe play sessions and listen to player discussions. Record if players mention hesitation, speed, or mimicry as part of their vote reasoning.
Prediction:
Players will naturally begin to pick up on behavior (“She flipped late,” “He copied that word”) as part of their social logic. This could become an emergent mechanic worth supporting further.
We want to address thos focused, testable question to allow rapid iteration without high-cost development(e.x. digital platform) first. We want to have a test of either looks & feel (timing, vocabulary reveal), or role (how players interact, bluff, deduce) before creating high-fidelity prototype.
