"Bondari Reloads": The Problem of Permutation in Gaming

by osbornep, last updated 20 Jul 2012 05:16


My purpose here is to investigate a simple question: Can video games successfully pull off a 'tragic' or 'downer' ending, given the constraint that games must respect player agency to a certain degree? Let us assume at the outset that in the hypothetical game in question, the tragic or otherwise 'negative' outcome is merely one of many possible outcomes in the game, which comes about depending upon the player's choices up to that point. For a discussion of forced failure in gaming, see this article, which resolves the issue once and for all.

The central question of this paper, then, is this: Does the presence of alternative, non-tragic outcomes within a given game undermine the game's ability to successfully pull off catharsis, tragedy or just overall 'downer-ness?' I give some reasons to answer in the negative, and then consider a more persistent version of this problem, dubbed the 'save game' problem by the good folks here.

I. Must tragedy be inevitable?

According to Roger Ebert (yes, that guy again), the answer is yes. Ebert invites us to imagine what would happen if we could re-do the events of Romeo and Juliet so that our star-crossed lovers could end the story by running off and living happily ever after. Wouldn't this cheapen the story?1 Or suppose that Othello uncovers and defeats the plans of Iago, living happily ever after with Desdemona: Would this not sacrifice the themes of the play for the sake of cheap sentimentalism? Ebert used this little argument to establish that games can't be art; to avoid redundancy, I won't say any more about this question except to refer the reader to this article. Still, if we think that changing the ending of Romeo and Juliet does devalue the play, we may be tempted to draw the following moral: A tragic outcome must be inevitable: It must be the only outcome possible. But for the kinds of games I want to consider here, no particular ending can be inevitable. It follows that doing tragedy in video games is a fool's errand.

In response, I suggest the following analogy: Video games admit of multiple playthroughs, and in virtue of this, we can see video games as similar to other kinds of performance that admit of multiple iterations, particularly musical performances. Suppose I play song X at one concert in a really loud and bombastic manner, and I play an acoustic, 'bummer' version of the song at another.2 Does the presence of the louder, bombastic version of the song make it inappropriate for the audience to feel plaintive when they hear the bummer version? Following the Ancient Alien dude's convention of answering his own rhetorical questions, I say the answer to this question is no.

If you ask me whether song X is a happy or sad song, the right thing to say, it seems to me, is that the loud version is happy and the acoustic version is pretty sad. There's not much point in insisting, "No, I'm not talking about any particular performance of the song, I'm asking about the song itself, dammit!" There's not some 'third' entity here which somehow captures the true emotional meaning of all performances of the song. Similarly with video games, we can ask "Is video game Y happy or tragic?" The answer must be "Playthrough 1 is happy, while Playthrough 2 is not so much, etc." If the analogy with musical performances is correct, then the mere fact that there is some alternative playthrough of X which is relatively happy should not undermine the emotional punch of the more downbeat Playthrough 2.

The error of Ebert's argument, in my estimation, is that it commits an equivocation between two distinct questions: 1) Wouldn't it ruin R & J if you knew that you could change the ending?, and 2) Wouldn't it ruin R & J if you actually went ahead and did change the ending? The answer to (2) seems to be yes, and it is this negative answer which accounts for whatever intuitive revulsion to the idea of multiple endings one might have. But it is not directly relevant to the question of whether or not the mere possibility of alternative endings should suffice to rob any given particular ending of its emotional power.

II. The save game problem

Still, the ease with which outcomes perceived as sub-optimal can be undone in video games should be cause for alarm. Suppose upon, getting the 'bad' outcome, I reload to an earlier point of the game to make different choices at the critical decision nodes which led to this outcome. I then manage to get the 'happy' ending. Given such a mentality, and the game mechanics to implement it, can the downbeat ending be perceived as anything other than a more elaborate 'game over' screen?

I'm reminded of an old and very silly tabletop RPG story from my younger days. The DM, who was also running a player character (bad idea), made a decision to use a certain spell in combat. The spell didn't have the desired effect, so the DM thought for a bit and decided, "You know what, my character wouldn't do that." He proceeded to retcon events to right before he cast the spell. This time, he took a different action, resulting in a much better outcome.

It seems to me that the ease of this sort of retroactive continuity undermines not only tragedy but any kind of storytelling in a way that the mere presence of multiple endings does not. If by "re-doing" Romeo and Juliet, one means re-playing, in the middle of a performance, the events right before the ending and changing them to have the preferred outcome, then Ebert's original point has some force after all. It is just this sort of revisionary action that the save game function frequently makes possible. I see three possible solutions to this problem, which are not mutually exclusive:

1. Incentivize minimal use of the save/reload feature by awarding experience, achievements, etc., or by assessing penalties for a certain number of reloads. Generating a fear of using the reload function, or an incentive to not do so may in turn provide an incentive to live with the consequences of one's in-game choices.

2. Implement a strict checkpoint system, perhaps allowing only one save per playthrough. Such a system would make the save game function effectively equivalent to a bookmark; it allows a playthrough to be distributed through multiple sittings, acknowledging the impracticality of sitting down for 20 hours at a time to complete a game. At the same time, it prevents one from undoing one's actions so easily. The downside to this solution is that many players will not want to play a game with such a strict system.

3. This one is a bit more experimental. Suppose your game has two endings A and B, where A is nominally the 'good' ending and B is nominally the 'bad' ending. What I'm suggesting is structuring the game in such a way that B, while unhappy from a narrative standpoint, isn't perceived as purely a 'losing' state. To achieve this, I suggest making it the case that B, while narratively more downbeat, has more content, perhaps an extra mission, additional cutscenes, etc. This way, when a person who got ending B compares notes with a person who got ending A, she may not simply have to feel that she 'lost'; after all, she got to see a lot of interesting content that the other player was locked out of. Such a structuring of events may encourage players to value the downer ending for its own sake, rather than merely perceive it as an outcome to be avoided on a par with the 'game over' screen.

None of the above systems is absolutely foolproof in the sense that they will prevent all players from trying to abuse the game mechanics to get the outcome they want. For instance, even using option (2), a player might decide to run parallel playthroughs, using the first as a 'test' to see if the outcome of a certain choice is what he wants, and the second as his 'canon.' It is enough, however, that a game discourage the abuse of such functions: It need not make them mechanically impossible, any more than a mystery novel needs to make it mechanically impossible for readers to turn to the last few pages before having read the rest of the novel. At a certain point, the writer or developer must simply trust the audience to not 'break' the story by abusing game mechanics, skipping pages, etc. Those who break that trust will reap what they sow: A weaker, less compelling experience.3


Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License