Monday, December 03, 2012

Agency and Failure

Milady wrote a really interesting post last week about Agency and Powerlessness. She contends that gaming narratives are immature in part because the hero always succeeds.

I agree and disagree with Milady. Gaming narratives are often immature because the hero is always successful, sometimes verging on omnipotence. However, Milady attributes this to the general maturity level of the gaming industry, suggesting that the gaming industry prefers adolescent fantasies. I disagree with this view. I think gaming narratives are the way they are because of the fundamental structure of games.

Failure in games is different than failure in novels and movies. In novels and movies the writer controls both the protagonist and the challenges she faces. Thus the writer can have the protagonist fail, and the audience considers it a valid failure. Similarly, in real life it is person versus the world, and if the person fails, that failure is valid.

But games are innately adversarial, either player versus the rules, or player versus the game writer/designer. For failure to be considered valid, the player must fail because of a choice she made. Failure that is simply imposed by the rules or game designer is not considered valid, not considered fair. Every situation the game puts the player in must have a solution.  If a player fails to find the solution, that is a fair loss. But if no solution exists, the game is flawed.

To put it another way, Kobayashi Maru situations are innately bad game design. The audience instinctively understands this, and when James Kirk reprograms the simulator, they applaud his ingenuity.

It is fundamentally a question of the balance of power in gaming. It is trivially easy for the game designer to make a no-win scenario. Thus it is bad form for one to actually do so. You see this in tabletop RPGs. The Game Master can easily wipe out the player characters whenever she feels like it. But a GM who actually does this is considered a bad GM, and accused of "railroading".  Failure must come from the players' actions, because the GM has all the power.

So from this we see the problem that games have. All situations the game puts the player in must contain a path to success to be considered fair, valid, and good game design. But, especially in an age of saving and reloading, the player can take all successful paths, and end up with a narrative where the hero is effectively omnipotent.

Sometimes games get around this by offering paths which are partial successes and partial failures. For example, in Mass Effect there is a choice on Virmire. The player can either save Ashley or Kaiden. She cannot save both. So that situation is one way of making the player fail in a "fair" manner.

Similarly the game can make the player fail by having the player make the choice to fail. Where failure is actually the right choice. But this is extraordinarily hard to do without making things feel artificial and forced. Of all the games I have ever played, only Planescape: Torment and Bastion have ever come close to this.

A Solution

My solution to this issue is a gamist solution. The problem is caused by the nature of game mechanics, and thus it must be fixed by game mechanics, not narrative ones. We desire a narrative game where the player does not always succeed. But the player's failure must come from the choices the player makes.

My solution is to have narrative success constrained by a resource the player controls.

You always hear political pundits declare that "the president must spend his political capital" as if political capital was an actual resource that is accumulated and then cashed in. So let's borrow that idea.

The narrative game should have a resource called Influence. The player earns Influence in some manner, and can spend Influence to adjust the outcomes of situations. But the player does not have enough Influence to affect the outcomes of all situation. She must chose which situations she must win, and which situations she can afford to lose.

This sets up a game where every given situation has a solution, but not all solutions can be taken. The player will fail sometimes, but she always fails because of her own choices. Either she spent the necessary Influence on other problems, or she chooses not to spend Influence on this problem, that she can live with the default outcome, and saves Influence for a future problem.

Games are not like other works of art. Failure must be handled in a form true to gaming. It cannot just imposed from above in order to create a mature narrative.


  1. I think something like the "Influence" idea would be better by the default, but even then people will still feel frustrated and feel they should be able to get enough influence to solve everything.

    "Sometimes games get around this by offering paths which are partial successes and partial failures. For example, in Mass Effect there is a choice on Virmire. The player can either save Ashley or Kaiden. She cannot save both. So that situation is one way of making the player fail in a "fair" manner."

    It's funny that you mention this, because this is still contentious. For example, what is the Normandy doing while you're talking/fighting Saren?

    Any situation where you tell the player they can't win everything is essentially loss by fiat. You dictate that they can't save both, you dictate how much influence they can have, and players can still easily feel cheated.

    Personally, I thought ME2 had a very interesting system. If you took the time to do everything right, the final mission could go perfectly. But you could also skip stuff and wind up not even surviving. And some people intentionally made bad decisions just to see the effects.

    Sort of a "how much extra work are you willing to do for the best payoff?"

  2. I think her argument depends on two things:

    1) that mature storytelling is the type you find in films and novels, where dramatic things happen to the hero/ine (which will include points of failure so that the hero/ine can rise above them and succeed).

    2) that the type of failure commonly found in games doesn't play a role in storytelling. eg. when you miss a jump or fail a level and have to start again.

    And aside from that, there's the whole issue of some players being really really really failure averse because they define 'winning' as 'a flawless victory' or 'best ending possible' and not 'did I have fun?' So is it the narratives that are immature, or the gamers who have an immature attitude towards failure? (I think the latter, which drives the former.)

    If you ever run tabletop, some players will NOT TOLERATE anything happening to their characters which they did not want. These people are not very fun as you can't even run a cool plot hook where they get kidnapped and wake up on a pirate ship without massive outrage. Even saying 'look, just work with me here and go with it, I have a really really cool adventure planned and you'll love it' will not mollify these control freaks.

    I have enjoyed games with narratives where my character failed at some point. In DAO there's a section where you are set up to fail in a fight and then wake up in prison, for example.

  3. I fear it's more complex than this.

    Books/movies have one big advantage. They can be linear/single outcome and nobody complains about this. Because they are 100% NON-interactive. You are fed the story from the author and you have no choice to make, which makes it possible to do a lot of interesting things (including killing the main character and keeping the story going).

    How do you expect to translate this into a game? There are several solutions, and all have problems:
    - one old RPG which name escapes me right now (the one set in Feist's world), which played like a book. The LotRO epic quest line works in the same way: you may be controlling the characters, but the story is already written and completed. Since you have multiple characters, you can have some get killed. "Choices" are like the one in books: i.e. none.
    - themepark-mode railroading. Some WoW quest zones have interesting stories, but as much as they can be interesting, *you* cannot get irreversibly killed, which necessarily implies that you will win.
    - multiple choices and outcomes: Wing Commander (the 1st one) had this, but there were very minor variations (win war, lose war), dealing with a significant decision tree would be a complete developement nightmare, ending with ages of developement time for 1 hour of play.
    - MMOs have the additional complexity that the world state must be consistent across characters, or you end with massive phasing/layering and all the grouping/interaction problems which follow.
    - tabletop RPG solves this with the gamemaster, who acts as a realtime author adjusting the storyline to the characters, but this can't be done in a mass-market game (be it single-player or MMO).

    The error, IMHO, is trying to apply the rules for books/movies to an interactive medium. You just can't.

  4. Kinda surprised you didn't mention Roguelikes, which have pretty much already solved this problem by baking failure into the genre and catering to a specific audience that's okay with the idea of permanent player death.

    Dwarf Fortress is a great example of this, as it's expected that you will fail, sooner or later, and that failure is persistent in the game world: e.g.: you can fail, restart a new fort or adventurer and can find the ruins of your failed Dwarf Fort (if you played in Fortress mode) or the remains of your adventurer (if you played in Adventure mode) and even loot the ruins/remains.

  5. Personally I would start by disagreeing that the presence of a hero is some sort of adolescent fantasy. In my view of art the proper goal is to present a view of the ideal through the principles attempting to be illustrated by the artist.

    When this is applied to a character you control it very much does come down to game design as you described. Playing a character on God mode however is terrible design. What can go wrong when you character is invincible? Nothing, so the game by default requires some sort of loss delivered through the narrative that you can in no way change it's pointless and annoying.

  6. I do not believe that the idea of a "resource" really makes the choices any more "valid" then before, since the idea that you can use this to directly make things "better" means that people will expect to have enough of it, or else they're forced to "lose" arbitrarily.

    In addition, I believe that you or the original author are conflating two separate things, here. "Failure" in a movie sense, when translated to a video game is more then just an inability to navigate an area or to engage an enemy. It's a disservice to video games to merely look at it that way.

    As an example, look at the Uncharted series, which does include the range of successes and failures you would expect from your typical action-adventure movie (a-la Mission Impossible or Indiana Jones). For drama-style films or books, you could look at various visual novels which basically do what you're suggesting (minus needing a resource). Virtue's Last Reward, while DESIGNED to be a game where you bounce around to make all the choices, is still a game where your choices have (for the most part) direct consequences.

    And then there's Spec Ops: The Line.

    Sure, some games make the scope of their "failures" be an inability to complete some task. And some make them a bit questionable in their legitimacy. But not all books and films are perfect in their logic either. The types of films that would be roughly equivalent to video games have many examples of the writer also just taking the easy way out, even if it leaves the viewer saying that it seems forced.

    If you're going to make an argument about maturity of story, I would make the argument that it's not because developers just don't want to write a "mature" story, but it's because the type of games that make money are the types of stories that just aren't mature in other media. If video games as a media are immature because the bulk of them are roughly equivalent to a summer blockbuster, then sure. But at the same time, I would point out as a counter just how many emotional dramas tend to come out in a year, and note that it's not all that large a number comparatively.

  7. Interesting topic.

    "Failure" is loaded term. What is lost in most games is the concept of setbacks or temporary failures. It seems that the game needs to be a triumphant march of glory to glory. One of the things that made the Wrathgate interesting from a narrative perspective is that the hero's lost. Personally I think that this should have become a major thrust for a patch cycle. The Forsaken had a strong presence in Northrend, outposts in the Outlands, and controlled all of old Lordareon. Why not derail the push against Ice Crown to engage in what becomes a civil war?

    Mostly because, as others have said, a loud minority in the audience would raise hell. The problem with gaming as a commercial enterprise, rather than a social activity for a group of friends, is that it needs to attract the largest mass of players possible and all outcomes need to be predetermined and packaged.

    As an old-school GM I like the idea of setbacks and the unexpected. But I can also just deal with it on the fly and my players are partners in the process rather than a disenfranchised audience.

    As a PM I hate the idea of both because it complicates the story structure and forces me to waste resources on paths that very few players may see.

    I don't think that there's much question where the accounting team or shareholders fall.

  8. I'm not certain this is what you're thinking of, but have you heard of a game called Long Live the Queen?

    Basically in this game specific events and situations occur during a year of the protagonist's life and you have to find a way to survive them. There are usually multiple ways to deal with a situation; however, they usually depend on having enough skill in a certain category to pass a skill check.

    So your resource is time, because there's no way you can raise all of your stats high enough to pass everything, so the game ends up becoming a bit of a juggling act between which skill checks you want to pass and which outcome you want out of a given situation.

    Granted, it's not quite the same as what you were talking about, since it's sometimes possible to choose the "best" solution without passing the skill check once you know the right solution and you can bypass some dangerous situations outright just by choosing not to do it, but in the end I still think it's a good example of choosing where you want to succeed and where you want to fail.

  9. There's a situation when playing a pulp pencil-and-paper RPG campaign wherein it's almost encouraged for the PCs to fail. Triple Ace Games' Daring Tales of Adventure (for the Savage Worlds system) explicitly spells this out; in some of their encounters they expect the PCs to fail, and that players should be briefed about this ahead of time so that they don't think that there is always a way out.

    This sort of situation is mirrored in pulp and pulp inspired movies (such as Indy Jones): the protagonist tries, gets captured, and by doing so advances the plot. Often it takes being captured by the Big Bad for the 'dastardly plan' to finally be revealed. (Cue monologuing, ala The Incredibles.)

    The problem with video games is that there isn't that level of DM/Developer interaction that's needed to pull this sort of "successful failure" off. I can easily imagine that a game wherein the player ultimately can't win in the end would engender the sort of revulsion that the ending of ME3 got.

  10. I was actually just talking about this topic with a friend the other day!

    For me, what it comes down to is this: the nature of games commands me to spend up to hundreds of hours playing a game. If the story ends with an ultimate failure despite my doing everything right, I feel like I wasted my time.

    (Granted, I feel the same way about books and movies that end on a futile note, but it has more impact in a game because of the level of effort required.)

    Now, setbacks and sacrifice are something else. Sticking with Mass Effect, I was fine with sacrificing Ashley or Kaidan, I accepted the failure on Thessia and I even enjoyed Kaidan breaking my heart on Horizon. These all gave depth to the story and to my Shepard.

    But had the game ended with the Reapers just eradicating civilized life no matter what you did, I would have been even more disappointed with the ending. While the current ending is pretty bad, a failure ending for the sake of a failure ending, would have amounted to nothing more than the game saying "congratulations for wasting 100+ hours of your time!"

  11. I think the diagnosis is off. Game stories are not immature because you always succeed. Obviously you can always succeed in any game that isn't impossible, and if games insist on having stories then you will always succeed in getting through the story. You can't then fix games by making you fail, because it would be like fixing a basketball game by making you lose. That's not a fix at all.

    Game stories are immature because they are usually straightforward stories of hero conquers all by doing exactly what it takes to win. Sometimes this is avoided. One WoW example was given above, the Wrathgate. Another minor example is the Klaxxi in MoP. Once you get exalted, you find out that they are working for your enemies all along, albeit not your immediate enemies.

    So a game can't force you to fail in the gameplay, but only as a matter of storytelling. You do what you are supposed to do, you succeed, it just doesn't work. This can, it goes without saying, be done well or poorly.

  12. Hmm, I really should have added an example. I'm not really talking about ultimate failure or losing the game, but more not getting the best possible outcome all the time.

    Let's take an example from ME3: the Quarian and the Geth conflict, and add Influence. Here's how I envision things working out:

    In the "default timeline", the Geth will defeat the Quarians, and Tali will commit suicide. However, the Geth will still join your war against the Reapers. If you spend a little bit of Influence, you can prevent Tali's suicide. If you spend some Influence, you can change things such that the Quarians beat the Geth, and the Quarians join your fleet. If you spend a lot of Influence, you can broker peace between the Quarians and the Geth, and both join you. But spending that much Influence here might stop you from getting the best outcome somewhere else.

    So here are outcomes by Influence:

    0 Inf - Geth win, Tali dies
    5 Inf - Geth win, Tali lives
    10 Inf - Quarians win,
    50 Inf - Geth and Quarians make peace

    The game continues on, it's not absolute failure. But the degree of failure is there. The Influence resource adds a cost to success, and you can chose what successes are important to you.

  13. "The game continues on, it's not absolute failure. But the degree of failure is there. The Influence resource adds a cost to success, and you can chose what successes are important to you."

    But even that feels arbitrary.

    For example, let's say there's 5 major situations to resolve, and you can get the best outcome by spending 50 influence each.

    So you get 150 influence and could resolve 3 perfectly and 2 not at all, or 2 perfectly and 3 reasonably well.

    But 150 is completely arbitrary. Why can't you resolve 4 out of the 5 successfully? Why can't you only resolve 2? Why not 125 points?

    All of this is based on the whim of the game designer, and if you fail it feels like the game designer is causing you to fail intentionally, not that it's any fault of your own.

    Especially on something like the Geth/Quarian thing, when you really want to just punch the Quarians in the face for being such complete idiots. It feels like they've been given the idiot ball, not that they have a valid point of view.

  14. I'm making the assumption that we have two options:

    1. The hero resolves every single narrative problem in the best possible manner.

    2. The hero makes mistakes and resolves some narratives in a less than ideal manner while still being successful on an overall level.

    Milady and I hold that Option 2 is superior to Option 1. If you disagree with this assumption(and it is an assumption), the rest of the post is simply irrelevant.

    Now, if we want Option 2, how do we create a game that results in that?

    Milady argues that the narrative failure should be imposed by the writer, and the player simply cannot avoid it.

    I propose that constraining the amount of success will result in some failures. Tying success to a resource gives the player control over which failures occur.

    You could possibly do it a third way by forbidding saves.

    Yes, it is artificial in that the player fails because of a constraint placed on her by the rules.

    But being ultra-successful at every situation is *also* artificial because it fails at simulating reality, turns the hero into a Mary Sue, makes it harder to suspend disbelief, and makes the overall narrative seem like childish wish fulfillment, rather than a genuinely good story.

  15. The point is that you can still "force" a choice failure without needing an arbitrary resource. Your own example can include that, because if you didn't meet the prerequisites, which were all fairly logical for what you were trying to do, then you have to choose between less-appealing options.

    Adding in the requirement to burn an arbitrary resource makes the choice no more "legit" then having it written one way via writer fiat or you needing to make a sequence of choices. As was pointed out, by making it a resource, you have implicitly limited the number of potential choices you can make, and the effect of each one. While one could argue that being able to navigate your way through to get the best ending to each choice could be seen as "immature", that ends up being a factor of the writing, and not necessarily the choices.

    Basically, there's a difference between successfully "winning" all the challenges in say a Call of Duty game, and "winning" all the challenges in Mass Effect or Persona 4. In CoD, you're just watching the writer fiat push things along like a summer blockbuster, while in ME or P4, you are engaged in negotiations, which can potentially be "won" all the time.

    Basically, you can't take one answer and apply it to all cases. There are games where writer fiat works perfectly fine for imposing failure and setbacks on the player, such as Uncharted or Spec Ops. And there are games where player engagement works better on imposing failure and setbacks, such as Virtue's Last Reward, XCOM, or Persona 4. And even in the games where player engagement works better, there's more then one way to do it, because your engagement in a game like XCOM is not the same as your engagement in Mass Effect is not the same as your engagement in The Walking Dead.

  16. What RJ said.


    "2. The hero makes mistakes and resolves some narratives in a less than ideal manner while still being successful on an overall level."

    Here's my question:

    WHY does the hero make mistakes and resolve some narratives in a less than ideal manner? I mean literally what causes him to do it.

    See, what we would expect in a book or movie is that the protagonist makes bad choices. And these bad choices result in mistakes and less than ideal manners, because the protagonists are human (usually).

    The problem is that video games have guides. If you present a player with 100 choices and each has differing degrees of success, players will find the best choice and go with that. They'll reload saves and calculate the benefits. In short, expecting the players to make bad choices can work, but it also opens the door to the players making all the right choices.

    The only way to avoid them making all the right choices is by fiat. Either you don't give them a choice in some cases (Gordon being captured in Half Life, Ashley/Kaidan, WoW narratives) to ensure they fail or you constrain their resources. Both of these are fiat.

    You can argue the latter might be superior in some cases (like saying, hey, you're a battlefield commander with X troops and Y bases are under attack, go save as many as you can (and it's impossible to save more than Y/2 bases or something)), but the latter is also very susceptible to making the player far angrier than the former if it feels incredibly artificial.

    On the former, it's relatively easy for the player to say "Meh, the story's supposed to go this way, I'll try to deal with it, it's a game," even if it doesn't make a ton of sense. The latter makes the player feel like its their fault and/or that the designer has terrible logic.

    Ultimately, the problem is that the player needs to feel like the game is realistic and that failures are realistic. You can annihilate that idea and make it clear that this is a story by not giving choices. But if you want to preserve it, it puts an incredible burden on the designer to make sure that the failure makes sense. That it doesn't feel cheap and like the designer is taunting you.

  17. Really, a thing I would have to say is that I find most arguments about video games still being "immature" as a media as pretty shortsighted. It feels more like people WANTING video games to remain child toys, and so are attempting to impose a self-fulfilling prophesy on the industry. That they are threatened by the idea that this is finally reaching the same level as film or books, because those damn kids and their devil [s]music[/s] games.

    If you take a look around the industry, you can find many examples of games that have strong stories with or without player engagement. Many of them tend to be indie, which mirrors that most of film media is quick and easy popcorn films from the "main industry" out of Hollywood, and the more abstract or experimental ones coming from the indie directors.

    And since many of these games are being penned by actual film writers, it's a bit insulting to claim that it's somehow worse then the work they do normally, simply because it's a video game.

    Yes, the latest Call of Duty or Battlefield game may be a power trip, but is it really any worse then the latest Bond or superhero film? Even those games have some kind of plot-driven "failure" to keep things moving along. Is it really fair to insult the gaming industry for making the kind of stories the film industry loves, and claim they're somehow not at the same level?

    The last few years have been marked by many strong game stories, and it keeps feeling like people who make the arguments about games being immature simply want to ignore that these titles even exist.

  18. Fascinating discussion, I love reading these kinds of posts.

  19. Did we not have something like the discussed resource system in the "Trial"- quest n Witcher 2, where you simply do not have enough time to piece together all the clues and interrogate everyone, forcing you to work with the clues you managed to get? I mean that system added time as a "resource". I like that better as it promotes multiple playthroughs where you have to prioritise which things you do and there is no "perfect solution" the way I see it is if the "Influence system" gives you a perfect solution, it could affect replayability in a negative way since your run of the mill guy is more likely to want to discover 2 not-so-perfect solutions, just to experience the difference, than to do that if you had a "perfect solution" as it signifies all other solutions are subpar.

  20. Interesting discussion, Rowan. Thanks for adding your two cents. I have been a bit late in reading it and responding, but here it goes: