Thursday, August 15, 2013

Turncoat Dev Diary: Thinking About Characters

(This is part of a series. The first post in the series is here.)

The reason why our original Turncoat concept was going to cost so much to create is that the scope of
the concept was just massive. The squad of soldiers the story follows consists of thirty-four soldiers. They are stationed on an enormous starship with over eight thousand people on board and which is part of a battle group containing about fifty other ships. In addition to the squad, there are many other characters that play into the story, including the Admiral of the fleet and her staff, the ship's captain and bridge crew, the ship's chief medical officer and sick bay staff, the "regular" complement of shipboard marines, and several command officers stationed back at Earth. Then, there's the Seditionists.

There are a lot of fully developed characters required to tell that story and a lot more that would need to be on screen at times in order to make the universe feel as if it was fully populated. On board ships and on an overpopulated planet, space is tight. To convey a proper sense of claustrophobia, we need lots of characters.

Characters are expensive to create.

Characters are also really important to the type of games we want to make. A casual game relies more on the mechanics of the game to generate interest, but a cinematic game puts the story — and thus, the characters — front and center on more equal footing with the gameplay. But, a fully detailed, sculpted, rigged, and facially animatable character like the ones you see in AAA console games might take a single artist a full month to create. We toyed with several ideas for reducing the cost of creating all these characters, such as doing cut scenes with comic-book-like motion graphics instead of fully animated and lip-synched characters and using a lot of close-cropped shots but, in the end, we decided that we couldn't reduce the cost enough without hurting our vision of what the game should be.

When we decided to put the main story on the back burner and start working on the Turncoat Escape game, one of our guiding mandates was to keep the scope of the game down. Way down. That meant having only a single level at first. It also meant keeping the number of characters as small as we could while still making the game work.

That's not an easy mandate to stick to. Stories are about people. If you can get your audience to buy into your characters… to empathize with them and care about what happens to them, you can get those viewers to overlook a lot of other things. But you need that emotional attachment to the characters if your goal is to tell a story.

Or, to put it another way: "It’s the characters, stupid."

There's obviously one character in the Escape game that we just can't do without: the protagonist. The player has to be somebody in the game. There has to be somebody who needs to escape the facility.

In our earliest version of the original Turncoat story, we had one character, nicknamed "Rook", who we envisioned as a customizable character. The player would see events unfolding through this character's eyes, and the player would be able to choose what that character looked like. They could decide what Rook's real name was, whether Rook was male or female, and they could select from a range of skin tones and adjust facial and body proportions and hairstyle.

As the story evolved and grew in complexity, Rook morphed into a specific character with defined traits. I still liked the original idea, but it eventually became clear that the story needed us to know more about Rook.

I like games that let you decide what your character looks like, though. Certainly, there are times when it's important to control the narrative and specify exactly what the main character looks like. Hell, we did exactly that with Rook in order to make our original story work. But, there's also value in making a game welcoming. There's value in telling the player that they can be whoever they want to be in the game, and that whatever they want to be, is okay.

The hero of the story doesn't have to look any certain way.

While we chose to abandon the character customization idea in the larger Turncoat game, the reasons we did that don't apply to the Escape game. This story is less complex. Making the main character customizable won't interfere with our ability to tell the story.

So, that's what we're going to do.

It will add some work for us, which runs contrary to our mandate to keep scope down, but we're still going to do it. It means we need at least two base models for the protagonist - a male and a female - and we need to put some effort into allowing them to be customized. We'll want the player to be able to change the skin tone, body shape, and the facial features in addition to letting them select gender.

So, what other characters do we absolutely need to tell this story? There have to be other people in this facility. Well, I guess there doesn't have to be other people. Through both Portal games, you never see another living person, and that game works extraordinarily well. But we're not making Portal; our story and our game relies on there being more people.

We need guards and inmates. Neither the guards nor the inmates will feature prominently in any cutscenes, so we don't need to make them as detailed as the player model, but we don't want them all to look exactly the same, either. Similar to character customization, we can randomize the features of the guards and inmates to make it feel like there are many different guards and many different inmates in the facility.

We also need a Guard Supervisor. The Guard Supervisor will have a different uniform from the rest of the guards as well as a defined non-random appearance; this character will always look the same every time you play the game. Some of the ways of escaping will require you to either find or avoid the Guard Supervisor, who is more observant and generally more competent than the regular guards.

Finally, we have the "white coats". The white coats are not part of the regular game level. They're never in the cell block or the guard areas. They're only inside "the theater" - that part of the map that the player can see from some of the ventilation ducts, but can't directly interact with. Exactly what it is that the white coats are doing in this facility is one of the things the player will be able to discover if they explore. They won't need to know that information in order to escape, but they'll have a better understanding of why they need to escape and why they're in the facility in the first place if they do.

One thought I had for the white coats was that since the vents are mostly down low at floor level, we might be able to make it so you simply never see their faces. As long as the the characters aren't too far away from the vent, the angle should hide their face from view. This might add to the mystery of these characters, and will have the added bonus that we won't have to fully model or animate the white coats' faces. That should reduce the amount of modeling effort.

So, our tentative list of characters right now is:
  • Configurable female protagonist
  • Configurable male protagonist
  • Randomizable female guard
  • Randomizable male guard
  • Guard Supervisor
  • Randomizable female inmate
  • Randomizable male inmate
  • White Coat Male (no face rigging)
  • White Coat Female (no face rigging)
That seems doable. It's, perhaps, a little longer of a list than would be idea, but as long as we don't go too crazy with the ability to customize / randomize, we should be okay, especially if we try to reüse parts, like faces and hands, between the guards, inmates, and white coats.

While we're going to create some of the characters in-house, I think we've probably got more characters than we can do without some outside help, so if you know any good character modelers with game experience looking for a little freelance work, have them drop me a line at jeff at martiancraft dot com.

Next Up: Prototyping Basic Game Mechanics Part 1
PreviousExperiments in Environment Creation

No comments: