StoryDeliveryMethods

Story Delivery Methods Spec

(for Mercs 2.0, Saboteur, War of the Worlds, Lord of the Rings, and Battlezone 3) -- Rob

The purpose of this spec is to ask difficult and annoying questions regarding the means of story delivery in Mercs 2.0 and every other game using the Mercs gameplay engine. The spec covers the actual methods we use to convey the story in our games, since we really don't rely on things like dialog trees and cutscenes.

Hey Designers! We really need everyone's help in creating an authoritative and final list of what I call SDMs - Story Delivery Methods. This is so we can begin to systematize the storytelling process and create a story pipeline (of sorts).

The bottom line is that the true author of the Mercenaries 2 story will be the mission structure and design, which means that the story is really the designers' baby, with the writers acting as wet nurses.

I'm also using this spec to propose a new design rule: ALL MISSION OBJECTIVES MUST HAVE MEANING AND CONTEXT.

Why Mercs 2.0 Will Never Win An Oscar (Besides The Obvious)

Here's the deal. As Cameron notes, many team members seem to think that compelling story, characters, and world comprise a missing crucial "pillar" from the game's design. That if only we had a great story, the game would somehow come alive.

I totally agree that we need to improve the storytelling, but ultimately, Mercs is NOT a story game. And I think for us to make a game with a complex, twisty tale with lots of exposition and character development is a very bad fit for our gameplay engine. That's simply not our game.

Mercs doesn't have long cutscenes, it doesn't have dialog trees, it doesn't have linear set piece missions. And these three techniques are the biggest Story Delivery Methods in gaming. And we don't use them. So we have an uphill job. Because of the way our game is designed, there will always be a inherent "cap" on how much heart-tugging and tear-jerking the game can pull off. No matter what we do, we will never, ever make Planescape: Torment or Grim Fandango. *sniff*

I'm serious. If you sit down and actually think about how much game time in Mercenaries is actually devoted to "story delivery", it's miniscule, especially compared to Planescape, which is a novel disguised as an RPG. We don't even have as many storytelling techniques at our disposal as the GTA series, which apparently has an unlimited cutscene budget. So we literally DO NOT HAVE THE TIME to do intricate plot twists and reversals, long exposition, or anything narratively complex.

We have to keep it simple. As a story experience, Mercs has just enough resources to put down a few dots and connect them. We don't have the luxury of a canvas and a paintbrush, like a LucasArts? graphic adventure or a Bioware RPG.

Matt Colville's Note: Certainly Mercs 2.0 wouldn't win an Oscar, but neither did Three Kings or Blackhawk Down (which did get Ridley Scott a Best Director nod) both of which were critically praised. I haven't yet heard that we couldn't make Mercs 2 more like Three Kings or more like the Losers, which would mean moving into Oscar territory. Though maybe only for editing. :) In fact, from what I've heard, regardless of what we work on next, we are planning on making Mercs 2 or Sab a more "personal" story, meaning plot and action relevant to the main character(s).

Rather than focus on what can't be done, I'd like to focus on what can be done and, for me, the rule of thumb is; story appropriate to the game. A GI Joe game and a Losers game can both have stories, but they'd be widely different. The question before us is; which type of game will Mercs 2 or Saboteur be? How do we 'optimize' the story for the game?

So far, I've not seen any mandates come down from above regarding what can and can't be done, regarding cut-scenes, animations, etc... on Mercs 2. Austin and I had a very profitable discussion about 'briefings' for which I should write up a document, please bug me until I do. I see a wide array of tools, some admittedly complex, at our disposal, but a complex tool does not need to SEEM complex to the end user, yes?

What's a Story Delivery Method? (Warning: Rob Totally Made Up This Jargon.)

A Story Delivery Method (SDM) is any technique the game uses to deliver story content. Most of the time, the story can't be told while you're playing the game (with the notable exception of a certain FPS). So usually the game pauses to give you a story nugget through an SDM, and then you move on.

The traditional SDM is the cutscene: you play a game level, pause for a cutscene, play the game some more, pause for another cutscene, and so on.

But an SDM can be a lot of things. In System Shock 2, the primary story delivery method are text logs dropped by the dead crewmen of the abandoned spaceship. In Planescape, it's dialog trees. In Half-Life? 1, even the tutorial is a story delivery method.

Matt's Note: I think the primary Story Delivery Method should be 'people.' This speaks to your notion of touchable fiction. I think the player should understand what's going on at any given moment because it's important to someone in the world. Note that I did not say "because someone in the world explained it to him."

I'm opposed to emails, load screens, websites, etc, because reading is not playing the game. Talking to people, however, even seeing them act and react, can be playing the game, on one level.

Another thing I'm firmly opposed to is Exposition. I hate it. I hate it so much, I just capatalized it like it's the name of an evil wizard who's trying to corrupt the earth with cancerous lesions. And it seems like 90% of video game dialog is exposition. Saying 'people convey story,' whether in cut-scene or in game, does not mean 'they explain things directly to the player.' Sometimes, it can't be avoided, but there are tricks to make it paletable.

There are many engaging, non-person ways to delivery story, as you've identified. I'm merely proposing 'people' as a rule of thumb.

The #1 Reason Why The Story of Mercenaries 1.0 Didn't Work

Here is the single biggest reason why the story of Mercs 1.0 didn't work.

We just didn't have enough Story Delivery Methods.

And the ones we did have, we didn't use well.

We didn't use long cutscenes, we didn't use dialog trees. What we did end up using was much weaker. Let's begin with the SDMs from our intro:

Mercs 1.0 Intro Story Delivery Methods

• 1) Introductory Cinematic - Most people skip this.
• 2) Slideshow - Ditto. Ask Matt about this. He'll tell you how effective it is.
• 3) C-17 - Not bad, actually. Matt did a really good job of differentiating the characters.

You can tell we're already in trouble here. We just delivered a GINORMOUS amount of exposition through SDMs that most players will totally button through. The C-17 is actually decent as an introductory SDM, but I don't know that we can keep using it as a storytelling technique throughout the game.

That's just the intro. Most of a game's story is told during the game itself, not during the intro or tutorial. So let's examine the bulk of the Mercs 1.0 storytelling experience:

Matt's Note: I agree. I believe the first thing the player experiences, should be some core gameplay. This does many things, one of which is; give us our first chance to dole out some important Story points, using the same method the player's going to experience throughout the rest of the game. You never sit through another slide show, media sequence, or C-17 cutscene throughout the rest of the game, for which players are thankful. The beginning of the game is not a time to break all the rules about story delivery, it's a time to enforce them rigidly, for they are more important here than anywhere else.

We made a mistake on Mercs 1, and that was; we didn't know we needed to explain all this stuff in game. We didn't know the early missions had to be specifically designed so that you dole out, in small portions, all the backstory the player needs. We're not going to make that mistake again. Whatever our ambitions are for the Story of the next game, we're not going to try and frontload it all before the game starts. When I say 'we' here I mean 'the team,' including people who weren't on the project at the time, like both Rob and I.

Here, I'm going to take a second to write down something that should be in its own doc, but which I don't want to forget. Rob may already have this handled in another doc. 'Story,' for our purposes, means Plot: Character; Characerization, Character Development, Dialog, and Setting, and maybe a couple of other things I'm leaving out. It means all these things, even though many people use it, incorrectly, as a synonym for Plot.

Mercs 1.0 In-Game? Story Delivery Methods

• 1) Briefing Room Dialogue - I usually just ran straight to the "Accept Contract" dude, because the briefings never gave you any useful info, game-wise or story-wise.
• 2) E-mails - I only skimmed these for the mission objectives, really.
• 3) Mission Banter - You have to pay attention or you miss these.

You see what I mean, don't you? The writers on a Mercenaries game are HOSED. You want us to depend on these four shitty story delivery methods to tell an entire story? What kind of story would you end up with? Well, we know the answer to that.

I'm not saying it's impossible to tell a story using these three SDMs, but I will say that it's extremely difficult and you'd have to really carefully plan out what information would be delivered WHEN and HOW. Mercs 1.0 doesn't do a very good job of that. And the game also underutilizes its strongest SDM. Here's what I mean:

Matt's Notes: Agreed. Awful. Hated it. Not going to do it this way again. Blech.

The Strongest and Most-Underutilized? SDM of Mercs 1.0

The sad thing about Mercs 1.0 is that we had an extremely powerful and effective story delivery method that we simply didn't use most of the time. This particular SDM should have carried about 90% of our story, and instead, it only carried about 10%. And that's a large reason why the story of Mercs 1.0 is weak.

What was this Story Delivery Method? (Brian Chan likes it when I ask rhetorical questions, and then answer myself.)

Mission objectives!

Mission objectives require the player to carry out a sequence of actions, which happens to be a very convenient way of moving the story forward. If we tell the player to "Kill a VIP", we ensure that 1) he meets the VIP, and 2) the VIP dies. And we can write the VIP character and integrate his death into the story. Everyone's happy.

Unfortunately, we didn't do this in Mercs 1.0. Most mission objectives were of the "Blow Up X Random Targets" variety, which meant that all the missions all felt very samey. And because they were all samey, the player never really cared about what was happening in the mission.

We never took the time to give any of the mission objectives in our game a CONTEXT or MEANING.

Which we will have to do in the next game, if we want our story to really cook.

Matt's Notes: So let it be written, so let it be done.

A Proposed New Design Law: Missions Objectives Must Have Context and Meaning

It's very clear that 90% of the story of Mercs 2.0 will be told through mission objectives and in-game actions performed by the player. Mission design is absolutely the strongest narrative tool we have in our toolbox. Much stronger than e-mails, much stronger than briefings, must stronger than VO.

Unfortunately (or very fortunately, actually), you can't have the writers design all the missions. Mission design is still very much in the hands of the designers. So here's what I ask:

We need to begin crafting missions and mission arcs that tell stories. And they don't have to be GRAND EPIC TALES, either. All I'm saying is, instead of "Kill Random Dude", we make missions that have "Assassinate the Villain You've Known Since The Beginning Of The Game Who Has The Helicopter You've Been Lusting For All This Time".

The example I keep returning to is the bank robbery mission arc in Vice City. You're told you need to rob a bank in order to raise the money to pull off a grand scheme. In order to rob a bank, the game asks you to complete a small mission arc:

Vice City Bank Robbing Mission Arc

• Mission 1: Recruit Wheelman By Demonstrating Your Own Driving Skills
• Mission 2: Recruit Gunman By Demonstrating Your Own Shooting Skills
• Mission 3: Rob Bank By Penetrating Defenses.
• Mission 4: Make Getaway from Police

Now I want to point out that every single last one of these missions is cut from the standard GTA template of "drive to checkpoint and shoot things". But the difference is that each mission has a CONTEXT and a MEANING. There's continuity from one mission to the next. You always know WHY you're doing what you're doing, and you know what you're going to do NEXT. As a writer, I eat this stuff up.

I maintain that the mission design of designing a bank location, and then creating four small missions around it, is much more powerful than creating four unrelated missions with random objectives.

So.

Matt's Note: I just want to add my prediction about what we're up against here.

A designer who says "I don't care about what the justification is, it's just cool to see this building blow up," has inadvertantly delivered the game a serious blow. The objective, and the reason for the objective, need to be inseperable, even within our internal discussions. If we try, as we typically did in this first game, to trick the player by coming up with a hand-wavey justification just to get to the cool bit, we've let the player down. It will look and feel hand-wavey.

If you say "Wow, the Sinuiju-Dandong? bridge is freakin' cool. We need to have a mission where you blow it up," the next question needs to be "Ok, why?" If we can't come up with a compelling narrative reason, we ditch the idea. "OMG you mean story is going to dictate what our missions are?!" I mean "good missions have both cool gameplay, and compelling justifications, not merely one or the other."

We'll have the reverse problem as well. Designers will come up with cool story beats, for which we can find no good gameplay. These must fall by the wayside also.

This symbiosis needent be complex. It doesn't have to be a complex reason. The rule applies equally to our GI Joe game, and our Losers game. 'Cobra's going to blow up this damn, we need to stop them' works just as well as 'the CIA is trying to frame us.'

The Number One Goal of Delta Story

GIVE ALL MISSION OBJECTIVES CONTEXT AND MEANING.

That's 90% of our story. Everything else is just icing on the cake. Everything else is coloring between the lines. Everything else is just flavor.

The Number Two Goal of Delta Story

IDENTIFY EACH AND EVERY SDM WE WILL USE IN MERCS 2.0

We MUST identify all the SDMs we will use. And the earlier we do it, the better.

This is because we need to systematize the way the story will be told in Mercs 2.0. Systematizing story delivery makes Cam and Djordi and Rick and the schedule happy, and allows us to get past all the hand-waving, posturing, ivory tower bullshit that usually comes along with writing.

Here's a partial list I've compiled. Please feel free to propose new ones, or reject the ones I have.

List of Story Delivery Methods In Our Next Game (Please Help!)

1) CUT-SCENES - Weak
I hate 'em. They suck up production resources (Clone Wars had to throw an entire designer at 'em), and most players skip them anyway, especially since they're basically close ups of human models with poor lip-syncing. It's like watching a movie minus the acting, i.e., shitty. But we could probably get away with one or two. System Shock 2 has, like 3 cutscenes, and they're all very effective. Overall, a weak SDM, but very useful for exposition in small amounts.

Matt's Note: I don't hate cutscenes, I hate bad cutscenes, see my earlier comments about how expository (read 'shitty') most game dialog is. Also, in my experience, cutscenes are more offensive to the hard-core gamer than the casual gamer. For the casual gamer, even bad cutscenes are often 'cool.'

I see cutscenes as one possible tool in our toolbox.


2) IN-GAME SCRIPTED EVENTS - Average
Djordi says that you could force the player into a small, enclosed area, and force him to watch a scripted event which he would not be able to interfere with. I agree with this in theory, but I have concerns about the amount of scripting involved. Also, we have to find a way to stop the player from killing all the actors, a la Half-Life? 2's invincible NPCs. Not a bad SDM, if we can pull it off.

Matt Speaks: I'd need to see this more to weigh in on it. I don't like relying on the player to pay attention to something that's easy to miss, nor do I like punishing the player later for not seeing it. This is one of the advantages of cutscenes, from my point of view. Philosophically, I feel that sometimes you need to hold the player hostage. If you are artful, they don't notice you held them hostage. I feel that there's an impulse to get directly to gameplay that's natural, but destructive to the overall experience.

3) BANTER - Weak
GTA does this. You and your bud get in the car, and you're treated to a humorous discussion. This is strictly flavor, though. You can't accomplish anything dramatically important using this SDM. A very cool SDM, but not a very strong one.

Matt's Note: Ah, but you can establish chatacter through banter, and sometimes character can be plot. I'd give this more weight than you have.

4) CONVERSATIONS TREES - Weak
The idea would be that these would take place only in the Green Zone, like when you're talking to a guy who wants to hire you. Of course, we would need the dog wagging its tail that says, "Arf!" when you try to talk to it. Unfortunately, I doubt we'll be able to cram Planescape amounts of text on the screen, so I consider this a fairly weak SDM.

Matt's Note: Text on screen? Shoot me now. People talking, I can get behind. If they're interesting, and saying interesting things, you don't feel like you've been taken out of the game. I never want to read anything on screen with the possible exception of pop-up tutorial text, which never conveys story.

My main objection to conversation trees is that you need some kind of interface to decide how you're going to react. Ick. I'd want to see a solution to this that allows multiple responses, without picking from a menu. I have an idea about this we used at my last job, about which I will exposit further should we talk more about this.

5) GREEN ZONE - ?
I also consider the Green Zone to be an SDM, since the player should see the consequences of his actions reflected in the Green Zone. I'm not sure currently how this would work.

Matt's Note: Setting can convey story. How people react, what they're doing, how the buildings look, which cars are driving down the street, the quality of light, all these things can convey story. It can be done. We can do it.

6) PLAYER'S BASE - Average
This is a really cool SDM, because we know the player will spend a lot of time there, and we can drop all sorts of cool story things here. We can do e-mail, visitors, messages scrawled on the wall from houseguests, the sky's the limit! And of course, we can have the bad guys attack it, a la X-Com! A great plot point, and a great way to deliver it.

Matt's Note: Ya.

7) BRIEFINGS - Weak
First of all, the current briefing paradigm has to go. I totally hate it. But we will need to replace it with something. And that something will definitely be an SDM, albeit one that players will probably skip through. Weak.

Matt's Note: I have a vision. Talk more about later.

8) MISSION OBJECTIVES - SUPER STRONG!
I just went over this this. The true author of Mercs 2.0 will be the mission structure and the mission objectives. Everything else is just flavor. Which is why we need to begin to move towards more cohesive mission design.

Matt's Note: Da.

9) E-MAILS - WEAK

10) QUEST LOG - WEAK

11) PLEASE ADD YOUR OWN IDEAS

COMING SOON!

HOW DO WE DEFINE AND CONSTRUCT A MISSION ARC?
HOW DO THE STORY STRUCTURE AND GAME STRUCTURE OF OUR GAMES FIT TOGETHER?
WHY WON'T THE TRADITIONAL THREE ACT STRUCTURE WORK IN OUR GAMES?