The Force Unleashed: From Concept to Console
March 22, 2007
By W. Haden Blackman
Project Lead on The Force Unleashed
We're not saying it lightly when we describe The Force Unleashed as the next chapter in the Star Wars saga -- we know how much weight that carries to Star Wars fans. Many aspects of this game have been amped up to movie-scale, including the lead-up to its release. For the previous Star Wars chapters -- the prequel trilogy -- this lead-up period was extensively chronicled on the official website and in a number of "Making Of" books.
If you tracked the making of Episodes I-III through webdocs, interviews and profiles online, you've probably got an understanding of the principal stages of filmmaking: pre-production, production and post-production. You also watched how George Lucas and his crew blurred the line between these traditional stages with new digital filmmaking tools.
With articles such as the one you're reading now and Brett Rector's production diaries (listed to the left), we at LucasArts hope to provide similar insights. How does a game move from a mind's eye idea to a finished experience -- and what are the steps involved? How does it compare to the making of a Star Wars movie?
The Concept Stage
There's a stage even before pre-production, called the concept stage. In Star Wars filmmaking terms, it is somewhat analogous to the time George Lucas spent developing story drafts, bouncing ideas off the skeleton crew Art Department before that group of artists increased in number. In the gaming world, it involves generating a lot of ideas for what might make a cool game, and getting those ideas onto paper -- specifically, what we call one-sheets.
One-sheets are summaries of the game that fit onto a single sheet of paper. It usually consists of a single piece of concept art and two paragraphs that sum up the game in a tidy "elevator pitch" -- the first paragraph focuses on the story and character, while the second relates to game features. This is followed by a series of bullet-points that help sell the game's features, and the one-sheet ends with an estimated rating, price point and possible platforms. Imagine this being a retailer's sales-sheet for a game that doesn't yet exist.
The game concepts must pass through a set of criteria we call "filters" in order to move past the idea stage. There's a double gauntlet of Star Wars filters and LucasArts filters that ensure that the game, even in its haziest idea stage, meets the expectations inherent in being a Star Wars game from LucasArts. Does it have relatable characters? It is an epic story? Is there innovation? Is there a sense of freedom, a series of interesting choices and payoffs for everything you do? If so, then the idea can move forward.
These one-sheets become the foundation of some qualitative and quantitative testing. We went to malls and asked a cross-section of people to vote about what they liked about each one. We saw a lot of trends and feedback from that, which let us focus on and prioritize what people most responded to.
These 20 to 25 concepts included a bounty hunter game that focused on weapons customization, a "superhero" type game where you played a Rebel Wookiee warrior, a Darth Maul game, a smuggler game, a game set 500 years into the future of Star Wars where you were "the last Skywalker." To be honest, we weren't planning on making the "next chapter of the Star Wars saga." That was a surprise awaiting us when we pitched the winning game concept to George Lucas in April of 2005.
Lucas said we could set the story between Episodes III and IV, an area in the Star Wars timeline closely guarded for future projects. He encouraged us to develop the Darth Vader's secret apprentice angle. Looking back at our one-sheet results, we saw that "unleashing the full power of the Force" was a particularly popular bullet point, one that Lucas too latched onto.
We took this feedback and crafted a few more one-sheets for testing. Far and away, the story of the Secret Apprentice rose to the surface. With these directives, we had the pre-viz video created to better visually explain the concept of over-the-top Force powers (Brett talks about that video in Diary #2) and made a more well-rounded pitch to Lucas. He had parts of the story that he really liked, and other parts, well, not so much. His fascinating feedback sculpted the story. He really encouraged us to create new characters.
So in my mind, as soon as we knew internally we wanted to go, that's when the concept stage became preproduction.
Pre-Production to Production
In the early concept phase, the Force Unleashed team had a scant five-to-ten people. We were growing a little bit, hiring our first new engineers to work on some of the core technologies of the game. Once we had a lock on the basic game type -- third-person action game with over-the-top Force powers -- we were able to grow the team quickly, up to 20 people. By the time we entered pre-production, we got up to 30 people and continued ramping up.
In a perfect world devoid of moving targets and shifting variables, you'd have clear-cut landmarks that delineate pre-production and post-production. Rarely is it a perfect world that awaits you when you embark on such an ambitious journey, though. We faced hiring challenges, as well as hurdles in that the tools, technology, engine and the pipeline to create the game were being developed at the same time that the game assets were being built. As such, the lines between pre-production and production really blurred, and continue to blur today. But knowing that the Star Wars movies similarly smudged those lines, we're in good company.
As an example, we currently have many levels in The Force Unleashed that are in full production. Artists are building them, action is being blocked out within them; you can inhabit them as playable environments. Others levels, though, like Kashyyyk are in a much earlier state of pre-production, where we're still trying to solve the challenges of rendering the trees, or what exactly happens there. There's still concept art being developed for these levels, a process "traditionally" defined as pre-production.
In the movies, sequences like the droid factory of Episode II and the Order 66 sequence in Episode III came about in a very similar way. Both were designed and executed well after principal photography had wrapped, when the movies were deep in post-production mode. Yet each was built from scratch with concept art and animatics while other parts of the movie were already being layered with finished visual effects.
Part of the challenge that necessitated such blurring is that up until the end of 2006, we were still determining the metrics of our new gaming environment. We were still getting a handle on how many polygons, lights, characters, and AI a next generation environment could support. Even something as simple as jumping from six characters to ten dramatically changes the dynamics of the scene, and may require some new brainstorming to help shape the scene's role in the story.
Such shifting ground makes it necessary for the entire team to remain agile. The concept artists definitely feel it, as they're constantly getting new requests for art and designs as the developers get new ideas based on our new metrics.
The great milestone of this phase of production is the First Playable, a version of the game that really proves out our high risk design and tech items. We've hit First Playable on this project, and that really opens the floodgates an builds confidence that you can build the rest of the game.
We're now in full production on a number of our environments. Our team has finished growing. We're wrapping up the tool work in the next month-and-a-half, and have started our motion capture sessions. We've got all the data that we need for the first round of cinematics, yet we're still doing concept art and writing story in some parts of the game. We're in full production racing towards Alpha, the next great milestone.
The definition of Alpha varies greatly across the industry. We take it to mean "code and content complete with no major bugs." This means everything that's going into the game is in the Alpha version, but it has yet to be rigorously scrubbed for glitches.
The Alpha stage is the time to get rid of all these bugs, while the Beta stage is the first submission candidate that is sent to the platform manufacturer -- Microsoft or Sony, for example. In Alpha, you're internally hunting for bugs. In Beta, you're dealing not only with bugs you've discovered yourself, but also the ones the platform manufacturers may call out. Our definition of Alpha is most companies' definition of Beta -- that is, LucasArts builds in the extra time between Alpha and Beta to refine the game, whereas most companies only spend the time between Beta and a final submission for manufacturing -- which is really just a few weeks -- for polish. Two or three weeks is not enough time, especially if much of that time is spent knocking out bugs.
Our goal is to get a good concrete Alpha, and then use the time between Alpha and Beta to tune and bug-fix, and make the game fun. For example, the Alpha version of the TIE construction facility level may be functional and bug-free, and has all the stormtroopers you're going to fight, but it just ends up being too hard a level. Maybe the stormtroopers' rate-of-fire is too fast, or maybe they have too much health and aren't dropping it quickly enough, or maybe they have some sort of weird AI patterns that are confusing to the player, or maybe the tutorial text messaging on this level just isn't clear. These are factors that can finessed in the time between Alpha and Beta.
At this point, the team does start to scale down in size. There's still the Beta submission period where we need engineers on call in case any surprise bugs are discovered, and there's a stretch of PR and Marketing activity that requires core team members to stay aboard. But that's very much at the tail end of post-production -- the period in filmmaking where there's special screenings, trailers, ads and junkets, just before the curtain rises on the finished product.