End of Life is a piece of interactive fiction developed by Simon Ferrari, a game critic and theorist currently completing his Masters at Georgia Tech. Though structurally pieces of interactive fiction tend to be less interesting than larger video games their low overhead means that they can take much bigger risks and production is open to a much wider range of people. Centering on a family’s decision of whether or not to remove their comatose grandfather from life support, End of Life continues a tradition in interactive fiction of featuring themes that are neglected in most video games. However, Ferrari’s piece breaks with tradition in that the user is intended to have relatively little agency. The actions taken by the user have next to no immediate feedback and the characters of the narrative form their own positions and forge their own courses of action, which both drift as the story progresses. Instead users are simply a “wisp”, in Ferrari’s own words, pushing each character gently in one direction or another.

What is most interesting about End of Life is that it is not simply a system of branching choices. Each family member’s ‘mood’ is randomly generated at the beginning of the play session, and each character will have random changes in mood as the story proceeds. Because the fate of the grandfather is determined by the combined moods of all family members the path to one conclusion or another is complex and can at first appear highly arbitrary. However, a deeper exploration of the system of the piece done by tracking outcomes over many sessions of play reveals certain biases in the system. These biases can, in turn, lead to different interpretations of the characters presented.

There are five characters in End of Life: the Grandfather, the Father, the Daughter, the Son, and the Mother. The entire story takes place over ten rounds, each round being an ‘hour’ in the piece’s narrative, and users can ‘nudge’ the mood of any one character once per round. A nudge can be in one of two directions: positive or negative. The user can also stay ‘neutral’ for a round and let the system play out. While each character has their own stated position on the situation presented at the beginning of the piece users will quickly realize that influencing their mood is not the same as influencing their position, and that their stated position has little effect on the story’s outcome. For instance, depending on the Father’s mood at the end of the piece he might be bitter over the euthanizing of his father (the Grandfather) or simply resigned to the fact, regardless of whether he opposed it at the beginning of the story.

It might be argued that this disconnection between the stated feelings of the characters and their actual influence over the events of the story shows a sort of fatalism. Whatever happens is going to happen and no amount of hand wringing or happy thoughts is going to change the inevitable. The inevitable, as such, is the result of forces larger than any single person.

However, this would be a shallow reading, probably based on one or two play sessions. The outcome, after all, is still being determined by something, just not the stated desires of the characters. Instead, perhaps the assertion that End of Life is making is that a person’s stated feelings are often simply a rationalization of their overall disposition, and it is that disposition or mood that has the most effect on a situation. This might be the interpretation drawn from a few additional sessions where the system reveals a little more of itself. However, once the user discovers that these rationalizations in no way effect the outcome they are free to ignore the character’s monologues and concentrate on experimenting with their emotions.

Experimenting with the moods of each character reveals a final layer to the narrative. While End of Life may present itself as a parable about the complex decisions and justifications that go into difficult and tragic circumstances, it is actually about the subtle and surprising power dynamics of the family it depicts. This is discovered when users stop spreading their influence over different family members, an act likely to be overridden by the noise in the system, and instead concentrate on pushing one family member’s mood in one direction or another.

By nudging one of the family members in one direction every round, thus keeping them in a consistent emotional state, and then doing so over several play sessions, users can start mapping out how the mood of that character effects the ultimate decision of the family as a whole. Finally, by staying neutral over several sessions, thereby determining the bias of the system, users can draw conclusions about the relative influence of each family member’s mood.

Surprisingly if the user chooses not to interfere and simply lets the family work things out on their own, they are slightly more likely to decide against taking the Grandfather off life support. While at the end each family member is likely to have a different stated reaction to the event, nevertheless as a unit they often decide to let the old man live. In this family though, the mood of certain members has more sway than others.

Take the Daughter, for instance. Over many play sessions it is revealed that if she stays in a positive mood the Grandfather is more likely to live, while if she stays in a bad mood he is more likely to die. Both the Mother and Son, if they are consistently in either mood, make it very likely that the Grandfather will live, but when the Son is Positive or the Mother is negative the Grandfather is virtually guaranteed to survive. Interestingly, the Father has the least influence over the decision of the family. If he is either in a positive or a negative mood over an entire session it becomes a coin flip whether or not the family decides to end the Grandfather’s life. Remembering that the default tendency of the family is to stay with life support, End of Life sports a rather cruel twist in that any activity from the Father actually lowers the chance of his parent’s survival.

Strangely though, it is the Grandfather himself that has the greatest influence over his own fate. Though the piece presents him as being comatose, if he stays in a positive mood he is all but assured to survive, while if he’s in a darker mood it’s overwhelmingly likely that his family will decide to let him go. It may be too much to say that End of Life has some sort of spiritual message, but at least in the case of the Grandfather it seems to argue for the power of positive thinking!

What each of these discoveries reveals is that End of Life is a much different story than it seems at first. A few quick ‘reading’ shows a family of characters, each struggling in vain with the other, suffering from the desires of both themselves and others in a crisis over which they have little control. However, a deep and thorough exploration of the different factors in the system uncovers a set of people that each have a definite effect on their circumstances, if only with the right attitude.

Simon Ferrari’s piece does not simply explore a certain strategy in interactive fiction, it shows that reading any interactive system is a different process than interpreting a traditional text. A more traditional examination of End of Life might look at what the characters are saying about themselves and about each other. It might consider the lessons learned from the juxtapositions of different reactions to different outcomes. A slightly more procedurally-minded analysis, as Ferrari himself suggests, might dwell on the nature of control, both its presence and its absence.

All of these approaches are undoubtedly valuable and any complete discourse on the piece should include them. However, each would also miss the biases of how the system is manipulating those elements of the narrative because each fails to delve deep enough to find the system’s logical extremes. It seems obvious after all that the revelations gained from searching the outer edges of a possibility space must have some impact on its final interpretation. This ‘reading’ of the End of Life shows us that understanding any interactive system is not simply a matter of parsing its content, but a process of iterative experimentation that reveals the larger truths of the system itself.

—

Author’s Note: The conclusions in this post are based themselves on a rather small sample size. I played through each character with each mood around twenty to thirty times. It would be interesting to see if the trends that I observed held up after a hundred or a thousand play sessions. However, this is something that I’ll leave up to Simon to test.