The Pre-Production Problem - How to Improve the Planning Process in Game Design - Extra Credits

Hello again. As of this week, Allison is officially on her break, so she can recover. And I think you know what that means. Yes, guest artist marathon. Over the next few weeks or months, I don't know, we'll be bringing in a bunch of guest artists to help us out. I think this is going to turn out to be rather fun. On that note, say hello to Molly Maloney. She's a concept artist currently working for her MFA out in San Francisco, and she's here to help us talk about pre-production. Huh. How come we haven't talked about this before? I guess we're usually more about offering possible solutions to problems rather than just raising them, but this is one of those problems that I'm not sure we've got a great answer for. Unfortunately, it's a pretty big issue, so hopefully talking about it will at least provide some perspective on the industry. Most bad video games are the result of bad production methodology, that is to say, bad planning. I know that sounds super obvious, but here's a specific case that's unbelievably common in our industry, and it's responsible

for a plethora of bad games, the complete lack of pre-production. Pre-production is the time when you try to figure everything out before you actually start building it. Basically, the planning phase. Not having enough of this time means things like story, characterization, art style, even mechanics get pushed into the "uh we'll figure it out later" bin. Now, as I'm sure most of you know, planning is everything, especially for big productions. Not having enough time to think about stuff before you just jump in and try to make it is almost guaranteed to cause problems. It's obvious to you, it's obvious to me, and it's obvious to everybody working in the industry today, too. So, the question becomes, why aren't studios making time for pre-production? After all, pre-production should be the cheapest period in the game development cycle. You don't need nearly as much staff as you do in actual production, and investing in that pre-production time is likely to save a studio money in the end. Unfortunately, there's one quirk unique to video games that prevents this, and that's the question of what do you do with all of your programmers? During full production, you're going to need a lot of programmers to put together all the technology that goes into making a game, but in pre-production, there's really not much for them to do. You'll need a few veteran programmers to investigate what

technology you want to use so you can create unique features for your new title, but when you already have an engine to work with from a previous project or you've licensed an engine like Unreal, there's just not much work there for a software engineer before any key decisions have been made about the game they're going to build. Sure, a few of the engine guys can tinker with the engine and optimize it a bit, and the shader guys might experiment with new shaders, which can be an integral part of establishing the art style, but what about the gameplay guys or the networking guys? They're stuck doing busy work or coming in for half days, and programmers are expensive. When you factor in salary, insurance, and paying rent for workspace, as well as paying for the licenses and technology they're going to be using, a mid-level programmer in Seattle runs roughly $110,000 a year or $8,250 a month. Say your standard mid-size developer has about 30 programmers on a AAA project. Now, let's say maybe 10 of them can do anything worthwhile during pre-production. That means you're wasting $165,000 per month you spend in pre-production. Or, put another way, more than 1 and 1/2 man-years of engineering time. Those are man-hours you could have spent actually building your game. This is almost

impossible to justify in many of the corporate hierarchies we have set up, and it's even harder to justify to investors. When you add to this the internal social pressure at studios to get people work to do, you can see how naturally pre-production often gets cut short. So, there it is in a nutshell, the problem of finding enough pre-production time. It's one that plagues our industry and ruins many multi-million dollar projects. Now, like me, many of you are probably thinking, "Well, with a little bit of planning, this could be solved by yanking the pre-production team off the studio's current project a few months early." The cute answer to that is, "Oh, with a little better planning, we could do a little better planning? You don't say." But, joking aside, I actually suspect that is the solution to this problem. There are a few reasons why game studios aren't doing it that way now. First, it would require them to move high-level staff off a product shortly before it ships. Unfortunately, the way this industry runs, games tend to be all hands-on deck affairs near the end, and even when they're not, games have another weird quirk. The more of a game you build, the more you realize how you could have done it better. James has never worked on a game project where they managed to get in every feature they wanted, and I don't know anybody else who has either. This means that even on projects which are coming

together perfectly, people tend to want to see if they can squeeze in just a few more awesome features from the design document that they thought they were going to have to cut, or see if they can get one more iteration in on some of the things they didn't know were wrong till they actually got to play test the game themselves, which in the end leads to no one wanting to jump ship to pre-produce the next project. Well, that and the weird sense of guilt industry folks seem to get about doing pre-production while their studio mates are still slaving away at the last game. For some reason, people tend to think of pre-production as not real work. Ironically, pre-production can also lead to productivity issues, especially when your team's just coming off a long crunch where deadlines were specified down to the hour. Pre-production time can lead to a feeling of having all the time in the world, and things start to run less efficiently as people recover. Okay, so we have a few problems to solve here. But, games aren't the only high-tech multi-disciplinary art form out there. How are those other industries doing it? Maybe we can learn from them. Let's start by looking at the film industry. When a movie's finished, the cast and crew and everybody involved go their separate ways. The people involved in pre-production start developing new projects, and the main production staff join other projects that are just coming out of the pre-production phase. It works just fine

for film, but I'm not really sure this model fits the game industry as it's currently structured. Keeping a development team together tends to be more important in games than in film, or at least that's the conventional industry wisdom. I'm not really sure how true it is. Even if it's not though, there's also the fact that we don't have agents in our industry. Hiring and job seeking are such a nightmare in games no one wants to try to staff up a project every time they start a new one. Besides, people working in the game industry tend not to like that kind of risk. Many developers want to know they're going to have a job after the game ships, and it's tragic how often they won't anyway. So, the film industry may not be the ideal role model, but I think I know another one that might be, animation. Maybe it's just because animation's my industry, or maybe it's because Pixar is like most game studios in that they keep their staff from project to project, but I think the model might work. Pixar generally spends at least two years in the story and pre-production phase for every film, but the vast majority of the crews working on a different project at the time. You've got your main bulk workforce of artists, all the modelers, animators, and lighters, and such. And most of them are probably working on the current main project, trying to get the film finished in time for theatrical release. But, all the while, there's always another

director working with a small team on the next film in line, refining the story reels and doing R&D on the technology they'll need. Meanwhile, yet another director's working with a handful of guys from the story department, putting together early storyboards for the next film in line. And all of those directors probably have ideas in mind for the next film they want to pitch. So, at any given time, you've got your main force working hard on a project, and by the time they finish, there's always another project ready for them to jump to. Now, I am definitely no expert on this sort of thing, but I think this system has some merit. Even if you keep things limited to just two projects, one in production and one being planned by a small team of leads, there's always a project to work on and another one in the chamber ready to go. All the benefits of planning ahead with none of the downtime. Now, despite everything I've been saying, it's probably worth reminding all of you budding developers out there that it's very, very easy to get trapped in the pre-production phase when you have no deadlines. James sees a lot of excited young guys fall into the trap of constantly revising design documents, drawing sketches of their world, and talking about their game, but never actually starting to build anything. It's important to know when to stop planning and start building, and you should never be afraid to prototype your

ideas early on. But, the value of spending this time thinking before we begin to build cannot be overstated. There have been a few games out there that ended up getting an extended pre-production time. BioShock, Fallout 3, Portal. I think the results speak for themselves. It's one of the few things most developers can agree on, but failing to get this time is endemic to the way we build games. I will say this, whichever developer solves this pre-production problem is already halfway to being the next Valve, BioWare, or Blizzard. If any of you guys work in games and have some thoughts or ideas on the matter, get in touch with us because we'd love to hear them. Thank you, Molly, for the pretty pictures. Thanks for watching, and see you next week.

Mhm.