Chris Crawford Has Not Given Up

Though this post has been around since mid-summer it only just recently came to my attention via a student of mine who saw Stephen Totillo’s post at Kotaku. In it, Chris Crawford talks about turning sixty and looking at his self-perceived lack of success with the StoryTron project and interactive storytelling in general. Last year, Chris was a guest speaker at Columbia College Chicago visiting the Interactive Arts and Media department, where he talked about many things, including interactive storytelling. His presentation was dynamic, thought-provoking, and a little controversial, which meant it was typical Chris.

In his blog post Chris talks about hitting sixty and looking backward at the fact that he has not succeeded to the level he’d hoped for with the StoryTron project. He laments that failure but ends with the resolution that he’s going to dig in and take another shot at it.

I heard about this and read Chris’s post while here in Vancouver at the iDMAa  2010 conference, subtitled “The Digital Narrative”. And while conference is about digital art, media, and narrative in all it’s forms there’s been significant conversation about narrative and interactive storytelling. The coincidence of reading about Chris’s introspection while at this conference has not been lost on me.

Personally, I do not think that computer-moderated dynamically generated narrative will succeed to the utopian level that Chris seeks simply because the task is too complex and the output too subtle. I come from a strong table-top role-playing game background and I know the difficulty of teaching a human being (let alone a computer) how to tell good interactive stories. (Everyone who’s had a crappy gamemaster raises your hand…) I also oversee the Castle Marrach which is an online, text-based multiple-user role-playing game (basically a MUD) where a great deal of the storytelling us shared amongst the players. There, some players understand storytelling (in this case cooperative), but many do not on anything beyond a fundamental level. Storytelling and story generation is hard enough with human beings that I find it hard to believe that it is possible to create a computer surrogate that will satisfy our needs as story participant.

That said, I want to be wrong and I hope that Chris succeeds. But even if he does not succeed I am heartened that his endeavor is going to continue. Chris is his post is looking at his failure to bring the StoryTron to a final product (in the loosest sense) or even produce a viable framework that could be built upon. While that may be true, I think Chris fails to recognize the value he’s generated just by taking the journey. The discourse around interactive storytelling has been elevated due to his efforts, his ideas, his energy, his writings, his provocation, and his perhaps Quixotic pursuit of the goals of StoryTron. Thank you, Chris Crawford.

I do not think that StoryTron will ever do what Chris wants it to do, but boy if it ever does I will happily be there to experience it, edible crow in hand.

Read Full Post »

Narrative Chicanery in Heavy Rain

This post is spoilerific for Quantic Dream’s loved and loathed PS3 game, Heavy Rain. I already have two posts about the game here and here, but this one is different. Stop reading now if revealing the main twist of the story is a problem for you.

I mean it.

Really.

Still here?

Awesome – let’s dig in.

First, I want to make it clear that I really really enjoyed Heavy Rain. I think it is more than successful overall  in what it sets out to achieve, and it is fun and engrossing. I do not believe it to be perfect, for a string of relatively little things, and one big piece of narrative fail.  Despite this bit of nonsense that I’m about to dig into, I would still rate it about a 90 on a scale of 100, which means I must have really liked the rest of it given how much of come to really dislike one particular bit.

Heavy Rain places you in the lives of four main characters – Ethan Mars, who’s quest for his kidnapped son Shaun is the backbone of the story, journalist Madison Paige who becomes embroiled in the search, and potentially with Ethan, FBI agent Norman Jayden who is brought in to assist the local police department, and Scott Shelby, a local private eye who is ostensibly acting on behalf of the families of other young boys who have been kidnapped by the same serial killer that has kidnapped Ethan’s son. You are afforded a great deal of control of each of these characters, during multiple sequences or chapters devoted to telling that character’s part of the story, and have significant influence on their actions and decisions. You are also in some scenes, with the press of a control, privy to their private thoughts about the situation they are in. (These private thoughts give you some amount of guidance, if you wish, in which actions or dialogue responses you may want to take with the character.) The key here, however, is the degree of control you have over the characters except for one key moment in the story. And therein lays the rub.

If you’ve played, you know the scene I am talking about. Scott Shelby, accompanied by a woman named Lauren (who is a prostitute and who’s own son was a prior victim of the same serial killer Shelby has told her he’s hunting) go and speak with the owner of a repair shop (clocks and typewriters, apparently) about tracking down a clue in the form of a message typed using an old typewriter. The two of you show up in the store with you in control of Shelby and Lauren tagging along not under your control. You speak with the store owner (an old friend of Shelby’s) about the message in question and he goes off into the back to check his records. Once he does, the player loses control over Shelby as the camera lingers on Lauren engrossed in a small figurine atop a display case. The clocks all chime loudly and oddly and then the player has control over Shelby again who declares that he should go in the back to check on the store owner… and when you guide him back there you discover that the old man is dead, apparently killed literally moments ago by the serial killer.

Except, as the climax of the story reveals, Shelby is the serial killer. So, during a scene/sequence where the player is supposed to be completely in control of Shelby the game removes that control, stages a scene out of sight where Shelby kills the man, and then returns control to the player who doesn’t know what the character he’s been controlling just did.

Um, no. That is all shades of wrong.

Narratively, the game has broken faith with its player. If I am in control of a character then I am in control of that character. You cannot empower the player in that manner and then through slight-of-hand have the character do something that the player is unaware of right under his nose. It means that I as a player cannot believe in the agency I have with that character, since it’s clearly a sham. I cannot begin to explain how much of a cheap trick this is, and one that diminishes the overall experience of the game. My decisions, my choices, become meaningless.

I think the game plays fast and loose with Shelby most of the time you are playing him, since while it’s possible to interpret many of the actions available to him as appropriate for a serial killer pretending to investigate his own crimes so as to cover up any evidence he may have left behind it is again playing narrative Three-Card Monte with the player. I can appreciate the shock of discovering the Shelby is the killer, but too often his actions and thoughts are written so as to be subject to interpretation so carefully that it feels like the character would have to know that someone (the player) was spying on him constantly. And in this case it’s the authors of the story who know the player is “spying” and write things so manipulatively that you don’t think Shelby could be the killer, until you look back on his actions with the knowledge that he is.

If you were a participant in the story, someone other than Shelby, looking at the private eye’s actions one way through most of the story, and then with more insightful eyes once you learn he’s the killer, that’s fine. That makes sense. But while you are playing that character? That’s chicanery.

Additionally, when you go in the back to “check on the old man” (who is already dead), Shelby reacts as if he is surprised and then you are able to control Shelby quickly investigating the scene which all plays out as if Shelby did not know the man was already dead. Why would Shelby behave like this? In case Lauren wandered in (which she does eventually), but that’s an iffy maybe? The bottom line is that there is no reason for the selection of actions you the player have at your discretion for Shelby since he knows he’s the one who did the killing! (Why, for example, stick your head out the open back window to see if the killer escaped that way – and knowingly leave fingerprints doing so – if you just moments ago opened the window yourself to make it look like the killer escaped that way. Who are you trying to fool… yourself? Narratively this is terribly shaky ground.)

Frankly, I can tolerate the narrative tap-dancing that occurs to keep the player from knowing the killer is Shelby until the end. It is him committing murder without me knowing while I am in control of him, during his scene, that breaks it for me. I’d be more accepting if say I was controlling Ethan, or Madison, or Jayden and not Shelby during a scene where Shelby went off to do his dastardly deed… but I am not. I am puppeting Scott Shelby and while the camera distracts me he goes off and commits murder. That’s not cool.

I look forward to seeing what Quantic Dream and Heavy Rain’s primary author/designer David Cage does next… but whatever it is, when I play it, I’m going to carry with me the nagging suspicion that the story is lying to me.

Read Full Post »

More Thoughts on Heavy Rain

Phil Cameron over at Resolution Magazine has a lengthy review of Heavy Rain up that I pretty much agree with. Phil’s a little more effusive in parts than I am, and I am still digesting the narrative shenanigans I mentioned in my post last night, but he said what I would have said. (Phil references the lack of narrative truthiness in his piece as well, and as a writer/designer/author I’m really having to chew on that …)

He also points out something obvious that I completely missed when I was talking about Quick Time Events – the Paragon/Renegade scene interrupt option in Mass Effect 2 are absolutely QTEs using a symbol rather than explicit button graphic as the prompt. Frankly, the entire conversation mechanic in Mass Effect 2 is basically a QTE scheme and it works just fine, thank you.

I also have to give Phil points for invoking the Talking Heads. :::golf clap:::

Read Full Post »

Thinking about Quick Time Events

The debate around Quick Time Events as an action control mechanism has reared up again with the release of Quantic Dream’s Heavy Rain for the Playstation 3 which uses the scheme as it primary interface system. In a nutshell, instead of providing a static mapping of controls to actions, where the [X] button is always Use, the [Right Trigger] is always Run, and so on, the Quick Time Events (QTE) scheme dynamically maps controls to contextual actions and prompts the player to use the required controls right as they’re needed. With most game control schemes the player has access to a discrete set of actions that they can perform at any time (or nearly any time) just by executing the correct button/trigger sequence. In a QTE system a particular action cannot be performed until the player is prompted to perform the action, and then success or failure is judged by the timing of the execution of the controls. (It should be pointed out that Heavy Rain designer David Cage literally called bullshit on the game having QTEs.)

What makes a Quick Time Event a QTE is the timing. One control or multiple controls in sequence or simultaneously, has to be executed in the time allotted. For example, after opening a door the player is shown a bomb about to detonate on the far side. The player receives another QTE control prompt to dive to the side and if that control isn’t executed in a second or two the bomb explodes and the player-character is caught in the blast. If he does execute it, the character dodges aside. The game’s difficulty is managed by changing the timing or the number of commands that must be executed in sequence or simultaneously.

With a traditional game control scheme the player uses the pre-determined control for moving or jumping out of the way. With a QTE scheme that player uses the control indicated right then for getting clear. Similar actions usually have the same control requirement, so the player can often anticipate what control he’s going to be prompted for, but sometimes it’s not what is expected. For example, the action prompt might be to close the door, rather than dive out of the way, and that would require a different control execution than what is normally used for diving out of the way. Player freedom is actually very limited since at any given QTE prompt there is only one, occasionally two, things that the player is allowed to do. For players who are used to traditional control schemes that allow him to chose from the actions available to him (Run, Shoot, Punch, Heal, etc.) to resolve an action scene this feels very limiting. This is an important point in that most game control schemes are tactical control schemes and are built to give the player optimal movement/combat control over a scene in accordance with the game mechanics. These kinds of control schemes are associated with action and shooter games where that kind of tactical move and shoot gameplay is king.

(As a brief aside, some more traditional presentations subtly incorporate the equivalent of QTEs without most people even realizing it. For example, Batman: Arkham Asylum visually prompts the player as to  which opponent is going to attack next, thereby allowing the player the opportunity to react by hitting the right control to block the incoming attack. The game doesn’t flash a controller symbol on screen, but it is effectively the same thing.)

An argument often heard against QTE schemes is that player is basically playing Simon, or Guitar Hero or Dance Dance Revolution just without the musical beats, but I think a better way to look at QTEs is as action scene choreography. QTEs allow action sequences to be much more cinematic and really the entire experience to be much more like playing a movie than playing a traditional action shooter. This may or may not be your thing, but it is just as valid a form of presentation as a tactical combat simulator, except that it places the emphasis on dramatic beats rather than tactical prowess. It also allows a greater, more dramatic, control of the pacing of an action sequence. It also creates a dramatic imperative:  Once the QTE sequence of the fight starts, unless it prompts him with the option to get out of the fight, he’s in it until the end. No matter what the player may want to do he is forced to guide his character through the fight until he either wins or loses. He is dramatically committed to the fight…or whatever sequence he is attempting to complete. And with a well-constructed QTE sequence that experience should be tight, taught, and dramatically compelling since the designers and artists are in absolute control.

A game like Heavy Rain isn’t interested in tactical control, but rather narrative flow, and in this case the flow is like a fast moving roller-coaster – once it gets going, there’s no getting off. . The fight with the thug is important, and momentarily thrilling, but narratively what’s more important isn’t how you defeated the thug – which fist punched the thug where – but why you did it and what happens after you do… or don’t.

Read Full Post »