In my last post on tragedy, The Wrong Ending, I presented what I saw as an essential problem of tragedy in videogames: an ending where things go badly is often seen by players as wrong, and therefore in need of fixing. This makes it hard for a tragic ending to seem like a valid choice for players. I also promised that I’d be back to discuss some questions I raised at the end of that post:
So how do you get a player to pick the wrong ending? More importantly, how do you get her to do that and still care?
In this post, I’ll be discussing some games that pull off tragic storylines, with varying degrees of success, and how they fit into strategies for addressing the wrong ending problem. These strategies all boil down to addressing the problem of how to keep a player from trying to fix what they did wrong. The games I discuss are ones I’m familiar with, and by no means an exhaustive set, so please do comment with other examples and strategies that don’t fit these four types. Spoilers for both Mass Effect and both Dragon Age games, Fallout 3, and the indie horror Downfall follow after the jump, although I’ll keep them as ambiguous as possible. Dan.
We Were Always Doomed
The crudest strategy for getting players to pick a tragic option without feeling that they need to fix their choice is simply to present a situation with no good options. This is a favorite of the Dragon Age series so far. The ending of Dragon Age: Origins, for example, gives you several choices, but the best you can say is that one of them appears to be putting off the disaster until a few sequels down the line. Dragon Age 2 is even blunter in forcing you to pick the lesser of two evils and making it clear that the world is boned either way.
The advantage of this strategy is that you might not feel as though you’re missing out on the “best” ending, since the options are just bad in different ways. This is also a disadvantage, however. If there isn’t a better path visible from where you’re standing, the emotional impact of the tragic ending won’t hit as hard. Tragedy isn’t just the feeling that things went badly, it’s the feeling that things went badly and you could have prevented it.
That said, players may be inclined to perceive a “right” ending among any set of varying possibilities. This seems to be the case with Dragon Age: Origins, which I believe makes it less successful as a tragedy for many players. That doesn’t mean that they agree about the right ending, however. Tossing the question out to Twitter revealed a wide range in what players perceived as the best or canonical ending from Dragon Age‘s eight or so possibilities. This wasn’t necessarily based on which was the happiest ending, but rather, factors like how interesting the ending was, dialogue bugs that revealed default conditions, and beliefs about what the morally right thing to do was.
Players have different criteria by which they determine the right ending and whether they’ve deviated from it. Learning more about those criteria and how to exploit them is an intriguing possibility for increasing the emotional impact of a tragic ending.
Bait and Switch
To some extent, the sense that one of Dragon Age‘s endings must be the right one derives from game conventions. We’re used to there being a good ending and some bad endings; in many older games, these endings would be associated with different point values that made this ranking explicit. You still see this in retro adventure games like A Tale of Two Kingdoms, but for the most part it has vanished from AAA games. Still, old habits die hard.
The Tenpenny Towers quest from Fallout 3 is a killer example of a tragic storyline that directly exploits this convention. The quest initially appears to have two bad endings and one good one: a compromise solution that makes everyone happy. When I played the game, I worked towards this solution and traipsed off into the Wasteland, smug in my good karma and diplomatic abilities. When I heard on the radio that things had gone horribly at Tenpenny, I furiously assumed that a story flag had been set incorrectly, and ran back to check on the bug. But all that had happened is that one of the factions had taken my compromise solution as an opening for an attack. Plausible enough, and a tight little tragedy where my tragic flaw was naivete.
This is a trick that may not work forever, should conventions change. For now, though, it’s a clever way to treat the player’s search for the best ending as a solution rather than a problem. Like the Dragon Age strategy, it uses a lack of good endings to push the player into a bad ending. But making one of those endings look better avoids the apathy that can arise from the explicitly “damned if you do, damned if you don’t” presentation of Dragon Age 2, and makes the player feel more responsible for how things turn out. For what it’s worth, it’s possible that the Origins ending is trying to pull this trick at a much longer time scale, which brings me to the next strategy.
The Long Con
If the basic problem with tragedy in games is that players want to fix things they did wrong, there are two obvious ways around that. The first is to make things unfixable, which is what the first two strategies do. The second is to make fixing things really, really unpleasant. That lies at the heart of strategies that put a great deal of game time between the tragic mistake and its consequences. By the time you realize what you’ve done wrong, it’s too late to go back and change things without repeating a great deal of the game. Replaying the game is a harsh punishment for players, since it’s both annoying and severely breaks immersion. Putting enough time between mistake and consequence almost guarantees that the player will give up and live with the consequences.
A straightforward and brutal example of this strategy is found in the indie horror game Downfall, which unfortunately seems to be unavailable for download at this time. (For this reason heavy spoilers will follow, since I can’t expect others to have played it.) Downfall is a stylish point-and-click adventure game in which you mostly play as one character. At one point, the lead is buried alive in a coffin with a shotgun while scrabbling noises are heard outside. This being a horror game, you’re inclined to believe the worst about what’s coming to get you. The scrabbling gets closer, and you’re suddenly given the choice to fire the shotgun or not. As I recall, you have a limited amount of time to make this decision, and while the timer ticks down, nearby ghosts taunt you about the monsters outside. It’s stressful. For what it’s worth, I fired.
Whatever you choose, the game abruptly cuts to an entirely different character and her own problems. You play as this character for quite a while, learn her backstory, and get to know her. She’s being guided by some kind of spirit on a quest of her own. Eventually it becomes clear that the endpoint of this quest is digging up the grave where the protagonist is trapped in order to free him. Dread sets in. Indeed, if you chose to fire the gun earlier, you blow your new friend’s head off right as she saves your life. For good measure, her headless corpse follows you around for the rest of the game.
I felt terrible. I wanted to take my decision back. But I also didn’t want to play through twenty minutes of goddamn point-and-click adventuring over again. So she stayed dead, and I went on feeling guilty about it. The tragedy hit, and all it took was a bit of time.
Other players may have more fortitude than I in this case, and there’s no guarantee that twenty minutes or so will do the trick. But as the decision-to-consequence distance increases, fewer players will be willing to go back. This appears to be one of the operating strategies of the Mass Effect series, where many decisions made in the first two games will seemingly only have their full effect in the final game. Stretching the consequence distance across eighty hours or more leaves only the most devout players willing to go back in time to fix things (what’s up, my Twitter feed). Others will prefer to just live with the consequences of their actions, even if they get a darker ending for it.
A variation on this strategy is also used within the first Mass Effect, in the scene in which your party member Wrex can be killed. It’s easy to avoid this happening if you’ve put enough skill points into one of your persuasion talents. However, many players who focused on combat skills instead found themselves unable to save Wrex. I find this to be a particularly effective strategy, since it follows not from a big explicit decision, but from a series of small choices over time about how to build your character. This makes it more difficult to go back and try again, since the point where things went wrong may be hard to identify. It also feels more earned, since it honestly follows from the type of person you chose to be. Your character’s flaws caused this.
Where Did I Go Wrong?
The scenario surrounding Wrex’s death is unusual among the storylines I’ve listed in that it doesn’t hinge on a single obvious decision point. Nonetheless, the game’s interface does explain where you went wrong: not having enough persuasion skill means that the conversation options that avoid bloodshed are visible, but disabled. You could push this strategy further by removing that explanation. A character would die, and you would feel like you could have avoided it, but you might not be clear on precisely where you screwed up. This is one last strategy to keep the player from fixing the tragic event: prevent them from knowing how to fix it.
It’s the strategy used to a degree in the ending sequence of Mass Effect 2. Prior to this sequence, the game tells you repeatedly that getting more loyal party members means that you increase your chances of survival, but that’s all the information you get. When the ending arrives, you’re thrown into a rather messy mix of loyalty flags, decisions, and random rolls that contribute to whether or not a given party member makes it out alive. By replaying the sequence and comparing notes with other players you can figure out the possibility space and get a perfect run, but on an initial blind playthrough there’s too much going on to easily trace your mistakes.
This is a great strategy in theory, and I found it effective myself. I liked the feeling of being overwhelmed and having events going awry despite best intentions. But that “comparing notes with other players” thing is a bit of a fatal flaw. Every game has a FAQ, and if you give players an ambiguous system, they will always figure it out eventually. And some players will always be tempted by the prospect of getting the right ending, even if (or especially because) it’s hard to get.
* * *
At heart, pulling off a tragedy in a game is about manipulating the player into accepting a situation they don’t want while still making them feel responsible for it. This is no small feat, but it’s not impossible by any means. None of the examples I listed are really immune to the basic “reload and fix it” issue that threatens to rob game tragedy of its impact, but they all suggest methods for making that solution less desirable. Digging into why they do and do not work might lead to better strategies down the line.
Line Hollis 
You had to know I was going to comment. I mean, with a reference like that — “Dan” — I just had to say something. There was no other way. (I had to stop reading at that point earlier. I was laughing too hard to drink water.)
I specifically mentioned “Spoilers” the other day for the same reason I am going to bring it up now: knowledge. I’ve written a great deal about it, but it still applies and definitely in this case. The key to having tragedy is knowledge, the knowing that things might have been better, could have been better if other options were chosen or even that something is inevitable no matter what the person does.
The proliferation of advertisements, game guides, blogs and, yes, Twitter means that I personally have a hard time escaping learning about events, strategies and even endings of games weeks, months or years before I play them. (Luckily, I read more than I play games and very few people talk about books on Twitter anymore.)
It’s this knowledge — for the drinking game, here is my indirect reference to ‘information asymmetry’ — that drives all of this. The Long Con (while also being a good song by Tegan and Sara) is all about this. The player’s ignorance is compounded over and over as she makes choices that she doesn’t know — yet — are bad ones or ones that will lead to ruin.
I like We Were Always Doomed because it echoes something that I learned right after I went back to playing Dragon Age: Origins the other day: Wardens die around 30. The blood that they drink — and the magic that comes with that? — eventually catches to them and they seek out The Deep Roads to either die fighting or, Andraste forbid, become Darkspawn too. Once my character learned she was going to die young, it shadowed my play — added to that was Alistair breaking up with me; I was ready to die at the end of my game.
That would seem to make the case that games would need to be procedural in order to gain this realization of previous unknown knowns. The player has to be in a position of limited knowledge and forced to make decisions. I’m not sold on that. But, on the other hand, I’m not sold on the sprawling story branching that games like Dragon Age: Origins has either.
I had a wild idea about a year ago and ended up using it for a short story in a fiction workshop I was in at the time. What if every time a player reloaded their save file, the game world got more fractured? I have often thought of save files as glimpses into different, parallel worlds. What if, instead, they were all one giant shared and, with each file, increasingly shredded world? With each reload, the game tried to merge the present state and that file leading to an increasingly strange world. (My protagonist, upon his character dying, kept trying to save his character by reloading files. Each time, his game got worse and worse until, not only could he not save his character, the game was now unplayable too.)
Dumb tangent! I’ve always found the “30 years” thing from Dragon Age fascinatingly ambiguous. Alistair says “you get thirty years” which can be taken two ways: you die at thirty, or you die thirty years after the joining. The latter makes more sense in some ways. It seems like Duncan was around during the war with the Orlesians, which was at least long enough ago that the leaders from that time have been able to produce a generation of adult children. Anora strikes me as mid twenties, for example. That should put Duncan way over thirty. Anyway, I read the Wardens of Origins as being about eighteen, which means they’d be getting awfully close as DA2 wraps up. So I guess I’ll find out soon enough.
Slightly more relevant: I dig the short story idea. Sounds very New Weird. And I’ve always wanted to see a game really incorporate a save game mechanic into its world, the way Planescape does with game death. Save games are kind of the elephant in the room as far as issues with impactful stories in games go.
Oh, see, that makes sense to me too. I took it as “You will die at 30” and, since the character is a reflection of me in many ways, was considering what I would do if I only had a couple years left to live. Given the other emotional moments, going out killing a dragon seemed like a pretty good choice.
Ironically, of course, that is exactly what happened. Shale was the only one who survived to the end of the fight and, even then, I was without healing potions, out of stamina and facing three Hurlocks running at me. Suddenly, with the dragon falling dead to a fading area of effect spell, the game cut to video of my character, who was now alive, killing the dragon.
That was after five separate tries to get through that battle too. It was easily one of the most stressful experiences I have had in a game in a long time.
Oh, oh! Yes. I’m quite taken with New Weird now. I read several of China Miéville’s books several months ago and really loved them. (For the readers out there, I really loved Embassytown — be aware it’s a slow burn though. I thought Kraken was too slow in places but loved all the various ideas he causally threw into different parts.)
It’s the save game mechanic that messes everything up in respect to tragedy, right? It’s that time-traveling aspect (“Wibbly Wobbly Timey Wimey”) that allows players to take knowledge and experience into older parts of the story to retry or even re-roll different outcomes and tactics. In fact, if you think about, checkpoints are this way too — if you die, you get to repeat a section with information you did not have the first time.
I really, really liked how Planescape handled it. It was pure genius to link character death to game story. Very few games — maybe Braid is another — have even tried that, let alone gotten it right.
Reply to Dan: An idea that’s always seemed sort of obvious to me, but I don’t think I’ve seen anyone try. Something I notice about RPGs:
– Your player is always, fricking always, the “chosen one”.
– You, uniquely, as the player character, get to keep rewinding every time you die or mess up and try again until you get it right. The villain doesn’t get “save points”.
Why not just unify these two ideas? Maybe you only get to save because you are the chosen one. One could establish this with two lines of exposition. There are statues to the Goddess of Fate everywhere (because for some reason saving always has to be identified with praying), and there is a legend that the goddess will one day bless her champion with the ability to keep rewinding back time to the time they last sought her blessing, but to anyone else they’re inert statues…
(…Or maybe the villain *DOES* have save points– maybe as a variation on your “corrupted save” short story idea, it’s made clear the villain is occasionally saving and reloading their game through the same mechanism way you are. So maybe you encounter and beat the final boss quite early on in the game, but then suddenly find time rewound to a segment 20 minutes before and continuing on a different story branch where you don’t fight the boss at that point– because there and each time after that you kill the boss, they get a GAME OVER screen and a chance to continue from their last save, which they of course take. Meanwhile each time *you* reload from a save the game gets harder, because you gave the villain that much more time to level up before your next encounter. Given, I think if one actually attempted to implement this it would be sort of an unplayable mess…)
An interesting take on this was executed (poorly, many would say — including the author — so I can’t recommend it at as a great game, but it’s an interesting structure) in a text game called “Grief” by Simon Chistiansen. In that game, multiple endings, all negative, were available. However, the game was built in such a way to encourage you to play over and over again (it helped that the entire game could be traversed in under five minutes after you’d played it once before, and even the initial traversal took maybe ten or twenty minutes tops). The idea was that each time you could easily trace back where you went wrong and play up to that point again and try to correct it, but that would just lead you to *another* bad ending. Eventually, the idea was that the player would be stepped through playing each possible ending, with the resulting message being that no matter what you did, you were screwed — nothing you could have done would have changed the outcome into a “good” one. (There were other elements to the final realization, but they’re beside the point of this discussion.) Obviously this wouldn’t be workable for a larger game with a more intricate set of choices, but for a short game with obvious branches, it was an interesting approach.
New Vegas mixes all of these approaches. The revamped skill system means that there is plenty of potential for just being unable to get the results you want. It can be put off and gone back to in many cases, but most important and non-repeatable conversations require some kind of skill check, not always Speech. Some characters are doomed in the end, even if they have a noble exit. The main story can’t ever be a classic win. You’ve got to eliminate or force into submission every other faction except your chosen one, though I will admit that Independent Vegas can be played quite close to a classic win with a lot of planning. There isn’t as much long con, but there is one good example of it in the DLC (though it’s mostly smoke and mirrors, I’d say it still counts and I don’t want to spoil it).
I would like to see some discussion of why one would want to make a video game a tragedy in the first place, because the desire to do so seems so firmly rooted in purely narrative aspirations, with little regard for negative consequences on the player’s engagement.
Take the Downfall example. It sounds like the architecture of choice had a decidedly negative effect on your experience. You wanted to take the choice back, but couldn’t. That could lead some players to just put the game down altogether. Doesn’t sound like a very good choice inasmuch as most designers want their games to be played all the way through.
The interesting thing about the Wrex situation, to me, is that the interface makes it clear why you cannot save him, which would then seem to suggest to the player “See? You should have invested points into these skills. Perhaps you should consider doing so, moving forward, such that you don’t miss out on another opportunity for a better ending in the future.” *Is* there such an opportunity in Mass Effect after that episode?
Figuring out how to keep everyone alive in Mass Effect 2 felt like common sense. Of course you want every ship upgrade you can get. Of course you want everyone loyal. Of course you want every tech upgrade you can find. Commander Shepard, in any incarnation, is a soldier and good soldiers prepare for contingencies. Then it was just a matter of more common sense as to who should take which assignment. Need a vent hacked? Send the machine man. Need a fire team led? Send the man who had experience leading combat teams against heavily-armed, professional soldiers on Omega. Etc.
Perhaps my single-minded determination to defy the suicide mission and lose nary a soul is why the task felt “easy” for me, but a friend recently finished the game and lost quite a few characters. To a point I envied his experience, but then I think about Mass Effect 3 and how pleased I’m going to be that everyone who could still be with me *will* still be with me.
To the contrary! That moment was a large reason that I found the game so memorable. The reason I write about tragedy so much is because it’s something that I personally love when I encounter it in games. I find negative emotions like the ones inspired by Downfall powerfully engaging. Some players don’t, just as some people don’t enjoy tearjerkers or horror movies. But other people are downright addicted to those negative fictional emotions. For this reason, I suspect that exploring how to inspire them may help open up games to player types that aren’t being served right now.
Not counting sidequests that may or may not take place afterwards, the only thing I can think of is the part where you can use persuasion to skip the first half of the Saren boss battle. But that’s a bit lower-stakes, both emotionally and functionally. That is, whether Saren shoots himself probably won’t have any effect on the trilogy storyline, whereas Wrex being dead clearly does.
While it might be closer to “narrative aspirations”, yes, we are also talking about the justification of actions within a system of rules. If you remove the story from a game like… say, Super Mario Bros., it’s about a person going in a straight line across a countryside, occasionally killing, or arranging for the killing, of various creatures. Without the justification of “These are bad guys, they kidnapped the Princess”, Mario is incredibly psychopathic as he kills various creatures which may or may not be native to certain environments in his singular pursuit of either a woman or a creature, depending on your reading of it.
Talking about the player’s story is a whole ‘nother thing. We always construct the experience within our own minds as a narrative, “I went here, I went there.” Even if we change it from first-person to third-person, “The character was directed to this place, this goal,” it’s still a narrative and would follow the same general pattern of other narratives.
The negative emotions that Line mentions is something I seek out too but, in my case, I avoid them if possible — I get swept up in stories too easily. I personally get too wrapped up in the character’s lives and struggles to always maintain the distance to achieve the catharsis that is so necessary for these experiences. That’s the engagement too.
When I tell of my adventures in these virtual worlds, I want to start with “You won’t believe what happened the other day… !” I want emotional journeys in some games. If you are saying that not all games need this, then, yes, I agree with you on that. Not all of them do. But, hopefully, some will have consequences for in-game actions that evoke reactions.
If someone was playing one of my games and their reaction was so strong to something that they stopped playing, I’d want to know about it. Was it a bug? Was it an error? Or was it, in all hope, some emotional response, some connection to the story, be it their own or that of the characters, that made them pause their fun or even give it up.
“I can’t play that, it reminded me of that one time my boyfriend broke up with me.”
“I had to stop playing {game}, it was just too sad.”
If the combination of exposition and mechanics was that strong, something happened there, some transcendence of the media from just code and text into a partial reality happened. I think we need to be talking about such possibilities and present day experiences. It’s a very interesting aspect to more and more games that allow player’s to tell their own stories, roll their own characters.
It sounds like ::SPOILERS at the link:: Metro 2033 did the accumulation-of-small-decisions-affect-the-ending thing too.
I suspect there is room for a few games to use the mechanical conventions and temptations as a source for the tragedy. Like what Bioshock wanted to do, but didn’t have the guts to really implement.
@Line: I was telling Dan my version of the Tenpenny Tower incident in an e-mail just the other day, and the exact same thing happened to me. I trusted the ghouls in their fancy clothing to stick to the truce and live with the humans in peace and harmony. When Roy declared war on the humans from within the tower, I sought him out and shot him. The other ghouls I left in peace. But I never forgot that feeling of betrayal. That was some good tragedy there.
One powerful way to pull off Bait and Switch is to make immediate rewards attend the Bait. Especially when you do something you shouldn’t be doing, morally. This might even be a Fifth category, Turn to the Dark Side or somesuch.
In the cited example, you unknowingly blew the head off a friend because you gave into momentary fear and the taunts of a lot of ghosts with an agenda to deliberately unnerve you. Imagine how much worse it would be if the ghosts were actually offering you Powerz and Benefitz if only you let them help you or possess you. Even after blowing your friend’s head off, you may not want to go back and give up the Phenomenal Cosmic Powerz. And that willingness to rationalize the Evil might have further downsides later.
One old school game did a good gambit on this, even though you play a character who is flat out the Chosen One, saving two dozen species. In Star Control II – The Ur-Quan Masters, while the player explores a vast map of conquered races and tries to amass the resources and lore to throw off the massive fleets that will eventually enslave or genocide everyone … the player meets the Druuge.
In playing the Druuge ships in the quick ship combat mode, a player would already know that the Druuge ship special power was to feed one their crew (ship hit points) into the furnace to instantly recharge the ship’s energy reserve. So when the Druuge offer a great deal of resources and ships in return for crew transfers … the player knows they are being offered the opportunity to run slaves who will eventually be murdered for power boosts, though the Druuge will never come out and admit that’s what is going on.
Doing this is obviously not right (though the Druuge carefully point out that by the contracts your crew have signed, this is all perfectly legal). But doing this is so very rewarding. And if you only do it a little, you won’t necessarily ever pay a game-mechanical price for it (crew eventually becomes more expensive to hire, but that’s just a consequence of hiring so many crew from a limited refugee pool).
But if you really go there, and seriously hand over sizable slave runs, you will eventually be called morally to account by the NPC refugee base commander. While they still need you to defeat the Ur-Quan Armada, the commander plans to see you brought up on charges and executed when the crisis is past. (And crew hiring costs go through the roof as this news breaks, making slave runs no longer economical.)
It’s seriously dark territory, and your character really earns the nasty consequences of their choices. BioWare has definitely gone to similar places with their villain and evil-alignment options in games like KOTOR.
This is really interesting stuff! I have a comment, which I think might have gone better in your previous “wrong ending” post, but: It seems like all four of your proposals here are basically tricks for getting the player to somehow accept the “bad ending” instead of worming out of it.
I’m kind of fascinated by a separate, maybe impossible idea: can one present a tragic, optional ending in such a way the player actually finds it *acceptable*? What I’m thinking of here is Heavy Rain. Speaking vaguely here to avoid spoilers, on my one playthrough of Heavy Rain I turned out to be very bad at it, and got the “bad end”– literally, as I found out later from someone who read a FAQ, the worst possible ending it is possible to get, the ending you get if you fail every story-branching QTE. The game did nothing to sugarcoat this– it passed judgement, specifically making it clear this was the bad, undesired ending in at least three ways. Despite this, when I actually got the ending, it was a *really satisfying ending*. This is a response that no other game with a “bad end” has ever managed to elicit from me, and I’m still trying to figure out exactly how they did it. The way I felt was that the ending I saw there, if I’d been watching a movie the whole time, and it ended that way, that would have been a satisfying, cathartic, believable ending to the movie. I didn’t even really feel specifically compelled to go back and retry some of the story branches, even though I felt curiosity about what some of the characters could have done had [x scene] gone differently, just because I felt like the story had been told and told completely.
Again though I’m still a little puzzled about how they pulled off making the “bad end” seem so satisfying to me– was it something about the ending itself, or something about how the game had prepped me to accept whatever outcome occurs?– so I’m not certain if this is replicable or just something made possible by Heavy Rain’s unique strangeness. (One specific, interesting thing the game does is continuously autosave, such that if you make a mistake, any mistake, the consequences are immediately permanent. You’re allowed to go back and retry sequences, but only *after* you complete the entire game. So maybe by the time I reached the end, the game had simply trained me out of the mindset of make a mistake -> go back and retry that part…)