Death Pledge



Aesthetic: Tabletop role playing, narrative battle royal.

Platform: Tabletop

Development Time: 2 Weeks

Team Size: 1

Role: Narrative Designer, Puzzle Designer, World Designer

Tasks + Responsibilities:

  • Writing a cohesive, engaging story with believable, fleshed-out characters, motivations, and relationships.
  • Designing an immersive world that supports the narrative across all scales: from micro levels like the mansion layout to the macro levels of an entire island.
  • Designing, prototyping, and playtesting diegetic puzzles that bear symbolic meaning within the narrative context.
  • Designing and producing visual materials: maps, floor plans, puzzles, and character sheets.
  • Analyzing interest curves from playtest sessions and revising narrative structure and emotional arc accordingly.
  • Learning how to be a good Gamemaster! (Easily the hardest part)

Gamemaster story guideline: Here

Characters background history: Here

Characters sheet: Here


Death Pledge is a narrative-focused Tabletop Role-Playing Game developed for the Game Design course instructed by Jesse Schell.

Design Process

From the very beginning, there was a few key elements I wanted to explore in a TRPG setting. These were:

  1. Player dynamics switching between competitive and collaborative
  2. Uncovering a vivid, mysterious world: particularly an island
  3. Taking away player freedom within the game

With this in mind, I brainstormed a set of options that seemed best suited for my objective:

  1. A group of blood cells going through the journey of human anatomy.
  2. Dystopic future where human bodies are hosted and controlled by alternative race.
  3. A paranormal experience in a haunted house setting.
  4. Humans as intergalactic nomads travelling from galaxy to galaxy, planet to planet.
  5. Survival story on a mysterious island at the center of the universe.

Despite their differences in story arc and genre, the options had a single common thread; a focus on exploration and world building. Hence I shifted my focus into creating an immersive, vivid world where the players “feel” as if they were actually there.

Given the limited time constraint, scoping the scale of the world was the first important step. An intergalactic universe seemed to vast, whereas being streamlined through human organs seemed offered little flexibility in its linearity. After hours of brainstorming, I found a solution that balanced the best of both worlds: an island that offered the players a freedom of choice to freely roam various places within its borders.

In terms of the story, I drew inspiration from multiple sources, from poems, mythical novels, movies, and video games. During the research process, it occurred to me that many ancient mythologies depicted the island as a place of the dead, an interstitial connection between the Earth and the Underworld. This idea of Axis Mundi, the center of the universe, seemed to imbue further mystery into the setting of an island. Taking further inspiration from Agatha Christie’s And Then There Were None, William Golding’s Lord of the Flies, and Richard Connell’s The Most Dangerous Game, I started to jot down interesting stories and events that could potentially happen on the island.

First off, I really liked the idea of initially portraying the island as an uncharted realm of paradise, then gradually unveiling its sinister nature. Secondly, I wanted to the character dynamics between individual survival and collective survival (collaboration against a common enemy). Also, I wanted to experiment to what degree can role-switching affect the player’s behaviours mid-game. To top it all off, I wanted to create a detailed environment where the players could feel a sense of place, where they can do whatever they wished.


With this in mind, I arrived at the following story:

One day, three wealthy individuals are invited by a mysterious count named Natas Amo to his private island in the Carribeans. Each one of them is interested in purchasing the island from the Master, and accepts the Open House invitation. At the harbor, they realize, from seeing each other, that they weren’t the only ones invited. Once the guests arrive at the island, they are greeted by the Master, who proposes a two-day competition to determine the ideal candidate to which he would sell the island. All seemed to be normal up until the night before the final announcement, when the guests encounter an old lady confined in the dungeon beneath the mansion. They are informed of the Master’s true identity, the Devil himself, and are warned to leave the island as soon as possible. Battles and deceptions ensue from this point on between the Devil and the guests, eventually leading up to a situation where the guests have no choice but to confront the Devil and kill him. However, the guests soon realize that the one who killed the Devil becomes possessed by him. One by one, they kill each other, until there is only one guest left alive. Through this guest, the Devil continues his lineage.







There are a total of three puzzles integrated into the story: 1) the invitation letter, 2) the dungeon door lock, and 3) the vault combination wheel.

The invitation letter puzzle consists of a series of letters hidden between the words. When added up, they make the phrase “Please see the map ninety west and thirty north.” This map coordinate points to a location on the world map where the players are to come for the Open House and meet the mysterious Landlord. This hybrid approach of combining diegetic role playing and puzzle solving immediately captured the players’ interests from the early stage of the story, which helped them to immediately get into the story. The players would read the letter first, and as they are reading they would start noticing the fainted grey letters between the words. This element of surprise in discovering multiple layers of information kept the players engaged. Also, the players were able to make the mental connection between the written prompt, “You will know where to go once you see it,” and the hidden messages. Then, all three players were able to make the connection between the hidden messages and the world maps, where they understood that they were to map the corresponding coordinates of the correct location. The actual content of the letter helped establishing the setting that the players and the landlord have been already corresponding over the past few weeks. It also helped in solidifying the characters’ interest in purchasing the island.

The dungeon door lock puzzle was initially intended to be a 3-dimensional representation of a chain guard, where the players had to physically slide the end knobs out of the maze-like frame. It was intended that the player had to work with the limited length of the chain and figure out the best route to the hole. Unfortunately, due to time constraints, the final puzzle ended up being a 2-dimensional graphic print instead, where the players had to trace the correct path from the start to the end hole. This caused some initial confusion in the players, as it was not immediately clear that the circles represented the end knobs and the line connecting them the chain link. Regardless, the players were able to eventually solve the puzzle with minimal prompt by the gamemaster. During the actual problem solving process, the players were often misled into dead ends that were intentionally incorporated to allow the players to troubleshoot the correct path without blatantly giving away the answer. In a similar line, intertwining the paths also helped making the correct path not immediately clear to the players.



Finally, the vault combination wheel was designed to be only accessed by the players with more than 10 mental attribute points, which ended up denying the Banker its accessibility during gameplay. This led to an interesting plot development where the Banker had to wait for the CEO to crack the puzzle. Inspired from a cipher disk, the puzzle consists of multiple layers of circles that can rotate around a common pivot point. The fragmented graphics required the players to rotate the different circles and match the patterns. This proved to be rather intuitive and almost too easy for the players; the puzzle taking the shortest time to solve. Encountered near the end of the story, the Island map was specifically chosen to subconsciously remind the players of the great Hole in the lagoon and the shape of the Island itself, which resembles a hand grasp.



What worked well 🙂

  • Providing a detailed character background and history allowed the players to understand the deeper motivation of the characters beyond achieving the goal itself. Also, even short descriptions of their personality and hobbies allowed the players to tailor their behaviours accordingly.
  • Personal portraits on the character sheets helped tremendously in getting the players to familiarize not just with their own characters but also with each other and the NPCs within the story. However, I do think this may be subjective depending on the player, as some may not be necessarily comfortable with self-associating themselves with characters of different gender.
  • The island map and the mansion floor plans allowed the players to form comprehensive mental maps of the story world. For example, having the sense of scale and distance helped the players in determining certain strategic actions, such as using the distant lighthouse to draw the servants far away from the mansion. Likewise, the diversity of rooms labelled on the mansion floor plans provided the players a clear set of choices in exploration as well as certain expectations commonly associated with room types.
  • The deliberate design choices in architecture and spatial arrangement enabled a fluid narrative sequence. The fact that the Guests’ quarters and the Master’s quarters were both located on the second floor and mirrored along the building’s central axis enabled the guests to explore the different rooms and potentially encounter clues. This also indirectly discouraged the players from wandering into the Cellars early on in the story.
  • The players actively used geographical and architectural elements to their advantage, and in a manner most suitable for the given situation. For example, the Banker and the CEO climbed the trees to hide themselves from the bloodthirsty hounds and servants. The fact that the players thought of the action in the first place may be attributed to how well it was communicated to the players beforehand that climbing was even possible to begin with.
  • The shift in Landlord’s role from the player’s common trophy figure to a common enemy have drastically changed the way players interacted with one another during the game. This goes to show that the well the players were able to recognize their shift in goals; from winning an auction to a fight for survival.
  • The puzzles proved to be much more engaging than initially expected (It also took much less time to solve). This may be partially due to how they were themed into the story’s environment and narrative context, and also due to the deliberate design decisions that focused on making the puzzles intuitive and fail-safe. This blending-in effect allowed the players to accept the existence of the puzzles without breaking them out of immersion.
  • An incremental narrative arc kept the players consistently engaged throughout the story. Rather than anchoring the long term goal from the beginning, this way allowed the players to constantly reformulate their goals, both long and short term. As a result, this gave rise to changes in both character behaviours and relationships.


What worked poorly 🙁

  • The players did not interact with one another at the beginning of the story, being more observant of their settings and surroundings. This may be in part due to the time it takes for them to take on role playing, but it also may be that they simply did not know how to react to one another, aka their relationships. Perhaps introducing some knowledge of each other’s character and background history may have given more grounds and topics to discuss.
  • The mysterious ambient soundtrack played from the beginning gave away much of the story’s plot surprises. While it was intended to enhance the atmosphere, choosing an incorrect soundtrack may send the wrong message to the listeners. Starting from a more lighthearted tone would have made more sense in transitioning into the mysterious mood later down the story.
  • The players were not all that intrigued upon arriving at the island. Perhaps providing visual materials would have allowed the players to paint a more vivid picture of the island in their heads. A more significant verbal descriptions would also have helped in this regard. Going even deeper, perhaps their journey to the island was too mundane, with no special events to keep the players engaged. Perhaps it should have been not so smooth sailing; maybe the players had to make conscious effort to get there (ex. fighting against a storm).
  • The players showed little interest during the competitive activities with the Landlord, despite having clearly understood their goals. I think this may be due to the fact that the activities – hunting and poker – were not sequenced in a more dramatic way. Perhaps it should have started with an intellectual game in the Mansion, then slowly expand out into the Island for the hunt. This would have offered the players the opportunity to unravel the mysteries of the Island, incrementally obtaining clues foreshadowing the future events to come.
  • The players were not that competitive against one another during the activities. This may be due to the activities’ format, which were introduced in a winner-takes-all manner. Instead, structuring the activities in a way that allowed the players to directly measure their scores against one another (ex. How many deers they have hunts in a given time) would have kept the players on their toes.
  • The dungeon lock puzzle proved to be confusing and unintuitive to the players. Players could not understand what the puzzle was and how they were supposed to solve it. This may be largely due to its 2-dimensional format and unclear graphics. Translating this into a 3-dimensional lock would definitely have helped explain how the players were to interact with it and how the puzzle can be unlocked.
  • There were certain moments in the story where the players were split up and it proved to be difficult to tell each player’s progress without having to directly give away the same information to the others. This divergence led to a make-believe scenario where the players had to painstakingly pretend they did not know what the other player have been up to. Perhaps the most common way in tackling this problem would be to hand out the story notes individually to each player, or even minimizing the opportunities to diverge in the first place.
  • Other players got impatient and felt left out of the action when the story was told in the progress of a single player. While the story was being told, the others would just sit and wait for the current player’s “turn” to end. For example, the Banker was hanging onto the window ledge all the while the Druglord was fighting the Landlord. In this case, perhaps devising a formal turn-based system would have helped in managing player downtime.
  • The players were under the impression that the Devil is an invincible figure, and that they should rather escape the island than directly confront him. Hence they kept using diversions and distractions as primary strategies. Perhaps this may be due to how the players were encouraged to leave the island by the old lady. If the players were shown, either by the old lady or through an earlier event in the story, any sign of weakness in the Landlord, I think they would have attempted to confront him.
  • Due to the zombie-like body of the Landlord, devising an absolute success / failure combat system between the players and the Landlord proved to be difficult. The players, especially the Druglord, were constantly shooting the Landlord, but were not shown any progressive feedback to their actions. To tackle this problem, I think clearly structuring ways to communicate the incremental defeat of the Landlord to the players, such as tracking his injury on a visual diagram / chart, will help.
  • It had not occurred to any of the players that if they were to completely destroy the Landlord’s body, the person responsible for his death would become possessed by the Devil’s spirit. This was intended to be foreshadows by the old lady’s warning, “Leave the island now… Become it possess you,” however it was not too obvious to the players that it was the Devil himself that would possess their bodies. Perhaps providing an explicit clue that the only way the players can survive from the island is by avoiding crimes would have prevented this confusion at the story’s climax.
  • The players were rather disappointed by the tragic ending. Despite their best efforts in avoiding the Landlord and escaping the island, they felt betrayed by the inevitability of their demise. As the gamemaster, I was too fixated on steering the story into a predefined ending, instead of catering to the current “player story.” Not only did this result in an abrupt resolution, it also left the players feeling like they had no real agency within the story. Perhaps allowing the players the freedom to “throw the Landlord back into the Hole, never to return,” may have been a far better ending.
  • The players were hesitant at times when they were unsure they can perform a certain action. This symptom may also point to the fact that the story felt too handheld and rigid, resulting in a story that felt too narrow and linear. Preemptive consideration of alternative paths may have led to an unpredictable, player-specific story. Hence, the players should have been allowed to detract from the main story to some degree. For example, the players should have been have to take the vault to safety, outside the Mansion.
  • The players were not all that surprised by the twists throughout the story. This may be due to the use of commonly known tropes that made the whole experience predictable to begin with. Or perhaps it may be due to the way the story was executed; by the gamemaster’s mysterious voice or the mysterious soundtracks leading up to the plot twists. This may have been resolved with a subtle balance between surprising the players and providing clues to build up uncertain anticipation.
  • The advertisements on Youtube became a source of major distraction, constantly interrupting the flow of the ambient music and breaking the players’ immersion. In some cases, this resulted in creating comical moments despite the serious and sinister mood of the story. A carefully planned set of soundtracks would have easily prevented this problem.
  • Many times the gamemaster had to provide information to the players that was not immediately obvious to them. For example, the players were told that the gasoline could be found inside the Vault back at the Mansion. Instead, the players could have made this mental connection themselves if they had known previously that there was a Fuel Room inside the Mansion.
  • The players did not make much use of the character sheets, only using their uppermost part. This may be due to the fact that there was little opportunities in the story for the players to progress their character attributes and abilities. Perhaps incorporating more Easter eggs, either in the environment or the challenges, would have given a sense of progress in character development.
  • There was only a handful of moments where the players could perform their actions, resulting in a few dice rolls. This relates back to the linearity of the story; if the players were offered more autonomy in crafting their stories, there would have been more chances for them to perform a diverse set of actions.


What was most surprising?

  • I did not expect the players to be so immersed within the story environments, exploring and discovering things that I had not anticipated. For example, it surprised me that the Druglord went outside the mansion on the first night, whereas the Banker attempts to enter the Landlord’s Study Room while he was asleep. Perhaps this may be due to the fact that the players were given the visual materials, like the island map and the mansion floor plans, that they can spatially refer to. The players created mental maps inside their heads, and based their choices and actions on them.
  • The players made active use of the materials and inventories they possessed.
  • The players embodied their roles to a substantial degree, which led to interesting actions befitting their individual character personality. For example, the Banker constantly requested to talk privately with the Landlord instead of playing games. On the other hand, the Druglord burst into the Armory room and got himself a classic Thompson machine gun in preparation for the danger ahead.
  • Later in the story, the players were more willing to share their personal experience with another, especially at the major turning point in the story. In absence of the Landlord, the CEO reveals her identity and shares what she saw in the Cinema Room and insinuates that the film may be foreshadowing what is about to happen on the island. Following in her footsteps, the Druglord then shares the strangeness in the servants’ behaviours. In light of this new information, the players unanimously decide to unravel the mystery of the mansion and the island.
  • At the height of the conflict with the Landlord, the Druglord tried to create a yet another diversion tactic and ran outside the Mansion. This was an unpredictable behaviour that was hard to mitigate and guide back into the main plot, and it also left the two other players vulnerable to the Landlord. Eventually, when it was declared that the Landlord has foreseen their tactical pattern, this kind of erratic player behaviour decreased significantly.
  • The puzzles, despite being relatively easy in their difficulty, contributed significantly to the players’ interest curves. This was most notable in the first scene where the players received the invitation letter from the Landlord. In contrary to the initial prediction interest curve, the players rated the introduction scene relatively high.



Interest Curve E vs Interest Curve F

  • As previously mentioned, Interest Curve F indicates that the players were fairly engaged when they received the Invitation Letter, as opposed to the initial prediction. This may be due to the fact that the letters were not merely handouts of information for them to read and digest, but also an interactive puzzle which, upon solving, progresses the story narrative.
  • Interest Curve F shows an interest drop during the initial encounter with other players, whereas Interest Curve E projects a positive interest slope. This may be partially due to the fact that the previous scene already set a high interest bar, but also it may be because the players were not willing to interact with one another, in the fear of disclosing information that could hinder their objectives. On top of this, this was the first role playing scene in the story, hence the players may have just been beginning to become familiar with their characters.
  • It was initially expected that the unveiling of the Landlord’s identity would create a spike in player interest, however it turns out that there is no substantial increase in Interest Curve F. This may be due to the fact that the players had already foreseen this outcome earlier in the story, either through the mood of the ambient music or the strange clues that the players observed within the Mansion on their first night.
  • Both the Interest Curve E and Interest Curve F show a similar interest spike at the Battle with the Landlord scene. It was predicted that the players were to devise a collaborative strategy to outsmart and ultimately kill the Landlord, which was very similar to how it unfolded in the game. Giving the players a common objective was instrumental in getting them to collaborate, even after an hour of social competition.
  • Interest Curve F shows an interest drop when the Player fights the Possessed, which is contrary to Interest Curve E. In contrary to prediction, the event happened all too fast for the players to fully comprehend what was happening. The players were also confused about why the Druglord was possessed in the first place.

Interest Curve E vs Interest Curves G1, G2, G3

G1 (Drug Lord)

  • The Interest Curve G1 and Interest Curve E both had a rising interest curve throughout the story. This may be because the set of choices that Druglord made was particularly similar to the story outline. For example, the Druglord successfully gains the Landlord’s favour throughout the hunting activity, whereas he ultimately guns him down later.
  • It was also surprising to see that the “Devil Reveals Himself” part was not particularly marked on the graph. This may be perhaps due to the fact that the Druglord drove down to the Lighthouse to create a diversion, whereas the other players headed towards the Mansion and confronted Landlord even before the Druglord got back.
  • There seems to be a relatively low interest spike in the Lighthouse scene. This points back to issue that the Druglord served a more background role while the Banker and the CEO were pioneering the majority of plot development. Hence the Druglord player had nothing else to do but to sit back and watch (and also pretend not to know).
  • The only interest drop seems to be in the Initial Encounter with Others scene. This may be due to the fact that the previous scene spiked the players’ interest through the puzzles but the current scene offered no basis for interacting with others.

G2 (Tech CEO)

  • There is a flat plateau in the initial competitive activities portion of the story (Arrival at Island, Activities with Landlord). The CEO showed the least degree of participation in the activities, as she found that her character attributes and skills were not all that relevant and applicable to the challenges being presented (hunt, poker).
  • There were interest spikes at the Dungeon and No Other Chance of Escape scenes. In alignment with the prediction curve, it seems that the element of horror and desperate situations were driving factors in keeping the CEO engaged in the story. This can be seen in how the CEO was the one to first encourage the others to join her down the Cellar.

G3 (Banker)

  • Much like the tech CEO, the Banker shows an interest plateau in the first half of the story. This may be partially due to her food coma (she admitted herself), but also it may be because she also felt disengaged with the activities themselves. Given the character traits of the Banker, social activities and actions would have offered more chances to prove herself to the Landlord.
  • The Banker’s interest curve shows the most significant spike in the Battle with the Landlord scene, where the players collaborated to devise the most optimal strategy in obtaining their objective. In this scene, the Banker served as the central figure of strategy, actively suggesting possible options. This was very much in alignment with her social and intellectual characteristics.
  • The graph shows a drop in interest when the players arrived at the island. Again, this may be due to the lack of significant events introduced during the journey.
  • The graph shows no increase in interest when vault has finally been opened. This may be partially due to the fact that the players were not merely given the enough time to use the resource, or that because the Banker had been previously denied accessibility to the Vault, hence detaching her interest from it entirely.
  • This is a perplexing one. The graph shows no increase in interest during the encounter with the servants. During gameplay, it seemed that both the Banker and the CEO were very scared in being caught by the servants, and made strategic decisions to remediate the situation.

G1 + G2 + G3 (All Players)

  • In all players’ graphs, there is no downward slope to account for the resolution (The Devil Lives On) after the climax (Player fights the Possessed). The curves seem to have been chopped off at its highest peak. This may point to that the way the story resolves itself felt incomplete, and perhaps allowing the players to abolish the Devil back into the Hole would have withdrawn a drastically different ending. This goes to show how important it is to respect the players’ investment in the story and their expectations from it.
  • There is no significant increase in interest when the players realize that the Druglord has been possessed by the Devil, which shifts the gameplay to a player-vs-player combat. This was one of the main things I wanted to explore in the story, but it seems that the players felt that they were forced into the situation rather than having created it themselves. Also, this idea was introduced too late into the story, too close to the ending, that the players simply did not have enough time to comprehend the consequences of their previous actions.


Did this experience teach you anything about interactive storytelling?

I have already mentioned the majority of the lessons above. In conclusion, having gone through the processes of designing, iterating, playtesting, and reflecting on interactive storytelling have taught me that role playing is very similar to improvisational acting. In this sense, striking a delicate balance between providing information and letting the players discover / create it themselves is incredibly important for an immersive story. This goes back to the question that Terrence Lee raised in his article: is it about the explicit story? Or the player story? And of course in this case, it is about finding a good rhythm to juggle them back and forth.