Sections

Innovation: Doing Things the Hard Way

New columnist Jurie Horneman on why fresh ideas should be difficult

Change is the status quo of the games industry. Platforms and genres are constantly evolving, and players always want something new - perhaps not too new, but definitely not the same game they were playing last week.

We're all trying to make games that no-one has ever played before, even if it is just the first voxel-based Space Invaders on Android. True innovation is harder, but also highly effective to distinguish yourself from your competitors, and to develop a competitive advantage.

I have some approaches that help me think about innovation. They are not easy shortcuts to success, because those do not exist, but they will hopefully help you think outside the box.

Innovating The Hard Way

If there was an easy way to innovate in game development everyone would be doing it already. Rather than complaining about how hard it is, one way to innovate - and to separate yourself from the herd - is to embrace doing things the hard way.

Teller, the silent half of Penn & Teller, once said the following about magic tricks:

"You will be fooled by a trick if it involves more time, money and practice than you (or any other sane onlooker) would be willing to invest"

Teller

"You will be fooled by a trick if it involves more time, money and practice than you (or any other sane onlooker) would be willing to invest. My partner, Penn, and I once produced 500 live cockroaches from a top hat on the desk of talk-show host David Letterman. To prepare this took weeks. We hired an entomologist who provided slow-moving, camera-friendly cockroaches (the kind from under your stove don't hang around for close-ups) and taught us to pick the bugs up without screaming like preadolescent girls. Then we built a secret compartment out of foam-core (one of the few materials cockroaches can't cling to) and worked out a devious routine for sneaking the compartment into the hat. More trouble than the trick was worth? To you, probably. But not to magicians."

The same approach works in game development. Tarn and Zach Adams have been working on Dwarf Fortress since 2002. It has now reached a level of complexity that is deeply scary. The entire world of the game is generated from scratch, including mountains, erosion, and rainfall (plate tectonics are in the works). Then the game simulates 250 years of history. Then you can play with dwarves who have simulated minds, and a damage model that takes into account bones, fat, and internal organs. Dwarves that carry scars and fall prey to strange moods. Dwarf Fortress is insane and nobody will ever make anything like it, but it has influenced a ton of other, more accessible titles such as Minecraft and Gaslamp Games' Clockwork Empires.

What about something slightly less extreme? Stop-motion graphics may seem almost antiquated, but because the form requires a lot of hard work few people are willing to do it. That may be part of the reason why Deirdra "Squinky" Kiai used it to make Dominique Pamplemousse, a stop motion musical detective adventure game, or why Cockroach Inc. in Sweden spent the last six years working on The Dream Machine. By doing things the hard way they made their games look like nothing else.

"How many seven-person development teams have a full-time voice actor? Supergiant Games did, and it allowed them to make games in a different way"

Innovating The Unique Way

Supergiant Games was a small, unheard-of team that was going to make an action RPG. They knew they could not compete with something like Diablo, so they had to find an edge. That edge was Logan Cunningham, the voice actor, who gave his voice to Bastion's narrator. How many seven-people development teams have a full-time voice actor? Supergiant Games did, and it allowed them to make games in a different way.

Kevin Bruner, the CEO of Telltale Games, said a couple of years ago that they had set up highly specialised tools and processes to be able to make episodic games quickly and at low cost, and as a consequence they had unique job descriptions. They have a Choreography team, for example, which you will not find at other game development companies. While that's not the only thing that made them successful, their approach to development was innovative. (Note how very few companies have managed to make episodic gaming work.)

Innovating The New Niche Way

If you don't want to change how you work, think about embracing a new target audience. My first job in the industry was at a long-defunct German company that had the reputation, mostly among European 16-bit computer gamers, of making very technically advanced games. The team consisted of a who's who of the best programmers and graphic artists from the Atari ST demo scene (plus me). We made that limited machine, and its slightly sexier sister the Commodore Amiga, do things they were not meant to do, but which looked extremely impressive at the time.

So imagine my surprise when one weekend in 1991 I found myself hard at work, not in our office programming a game, but in a warehouse at our parent company's distribution centre, putting game manuals and flight maps into boxes. (When I tell people I've done everything in game development, I am specifically referring to that weekend.)

Our company had recently signed a publishing deal with a commercial airline pilot who had written an Airbus A320 simulator for the Commodore Amiga in his spare time. Since flying an A320 was his day job, his simulator was very realistic, and our managing director had seen its potential and set up partnerships with Lufthansa, Deutsche Airbus, and a company making official flight maps, which lead to all that stuff having to be put into all those boxes.

"If a magazine exists there are people interested in that subject, and that may be an untapped gaming market"

And a lot of boxes there were. Despite the fact that the game was written as a side project by someone who was not a hot-shot programmer working in assembly language, the game sold very well. Despite the fact that no one had noticed the niche, there were a lot of people who wanted an Airbus A320 simulator for 16-bit computers. Despite the fact that the game did not use any of the bleeding-edge 3D engines we had developed internally, it filled an unmet need and outsold most of our other titles.

Airbus A320 taught me to pay attention to things that people are fanatical about. Even today I look at racks of magazines - remember those? - for ideas, because if a magazine exists there are people interested in that subject, and that may be an untapped gaming market.

As an another example: consider that long before Twilight became popular, the urban fantasy and paranormal romance genres already had a massive following and dedicated shelf space in book stores. In games, even now, the only place you find that same content is in hidden-object games. That looks like a massively under-served niche to me.

Innovating the Revolutionary Way

What if you want to do something genuinely new?

Research & Development is usually just one division of many at big companies outside of games. All of game development would fall under the 'D' of R&D, even the shared technology used in multiple games that was often called R&D in companies I've worked at. I have never seen real research, in the sense of directed exploration, being done in game development. Perhaps it happens at some of the very big companies, I don't know.

But I think we are bad at long-term research because even if it is approved, it has to be over and done with by the end of preproduction, and then it had better work. Nobody is going to go out on a limb when there's a penalty if the limb breaks. It is these time constraints that are particularly problematic. The period between the official start of a project and the end of pre-production tends be 3 to 12 months. That means that anything that looks like it might not be done and production-ready in that time will be ignored, and that creates a huge blind spot. If your only criteria for starting a revolution is that it has to be a sure thing, you're not going to start a lot of revolutions.

"If you keep making the same kind of game, you will eventually learn a lot. But you will still only learn from small steps, never from big leaps"

Don't we learn a little from each title? Don't we improve a little every time? Isn't that a kind of research? Hopefully we learn from every game we make, and can apply what we learn in the next game. And if you keep making the same kind of game, you will eventually learn a lot. But you will still only learn from small steps, never from big leaps.

One helpful thought experiment I use during production is to consider whether a prospective task is like walking uphill or jumping a ravine. If I work on something for a month, and I end up with 50% of what I was hoping for, and that's OK, that's like walking up a hill. I may not have gotten as far as I wanted, but at least I made progress. If I work on something for a month and I end up with nothing, which can definitely happen, that's like trying and failing to jump a ravine and breaking a leg. Your effort is wasted and you're in deep trouble. Since the worst-case scenario means your effort is wasted, it's not surprising that game developers rarely try to make big leaps, or research things that take a long time.

What about academia? It is filled with very smart people who are incentivised to think about hard problems, but they are not necessarily incentivised to think about problems that lead to better games. And even academics may not stick with problems long enough. I recently spoke to Chris Crawford, who has been researching interactive storytelling for 25 years. When I asked him what he thought of academic research in this area, he expressed disappointment because academics tended to give up after 5 years or so. That's a perspective you don't hear every day, but it's an important one to remember.

What is there to research? Don't pick something easy, like rendering. Sorry, rendering friends! Rendering is hard, but knowing what the problem is that you're trying to solve is easy. Just look in a mirror. See that slight glint on your lower eyelid, that little specular highlight? You could be the first one to put that in a game! Or in a game with a thousand characters! Or in a mobile game! Et cetera.

"If you want to do research very few other people are doing, pick something where even the problem isn't clear"

If you want to do research very few other people are doing, pick something where even the problem isn't clear. Is there a better way to tell stories in games? How do we make games about people, not objects? Who knows?

In pursuing such questions a huge team which needs to do profitable work is more likely to hold you back than help you, but there are still ways to manage the problem. Among the many reasons why big companies have dedicated R&D divisions, one of the most important is that failure at the R&D level won't take the whole company down. By scaling R&D as appropriate you can avoid the same fate without having to give up on it entirely.

Innovating

If that all sounds tiresome and impractical then innovating may not be for you. If you wonder why anyone would do anything the hard way on purpose, or if you're not willing to think about changing the way you work, or to look outside the industry for ideas, then innovating may not be for you.

And that's okay. Spending years researching something that might not lead anywhere, or making something that might not lead to runaway profits, is not for everyone. You can - and I mean this - make a perfectly good living without doing any of this. I have, for years.

If you do want to innovate, however, there are plenty of questions left unanswered. Whether you innovate the hard way, the unique way, the niche way, or the revolutionary way, all you have to remember is to embrace the difficulty. Because the difficulty of innovation is exactly why everyone else will ultimately quit.

Thanks to Mark Barrett, Dave Grossman, Andy Schmoll, and Verena Riedl for their help with this article.

Latest comments (2)

Bonnie Patterson Narrative Designer, Writer 2 years ago
I was thinking about research the other day, mostly because of my great unspoken crush on Peter Molyneux's brain.

When I see his statements on what a game in development is going to be like, I see his plan for the future of the game, I see what he's going to tell his team to try to do. To me it then looks like they then fall short by trying to produce neural network effects with ordinary code, but maybe it's not that. Maybe they tried NN, GA, symbolic AI and couldn't get the effects they wanted with the computing power they had. And one of the problems of having a company is that you have to get something out in order to make some money to stay afloat.

In my dream world, someone then showers Mr Molyneux (or another developer like him trying to do new things with intelligence in gaming - imagine Jane McGonigal with infinite funding for research) with cash and he sets up the Peter Molyneux foundation for Excellence in Artificial Intelligence in games. Programmers, teams of programmers or academics go there and spend 5-10 years researching new technology. It takes the same approach as scientific research in Germany, or SkunkWorks in the US: "Find out something new - just one new thing. Let us worry about working out how to use it."

The output goes to the Game Development team to make into new games or the technology patents get licenced to other companies who are willing to pay nasally for this new paradigm in gaming (imagine a single player game where the enemies in hard mode don't just have more HP or harder attacks, but fight like a player would.... or better, using the kind of off the wall strategies that only an AI can come up with, like this super-strong truss: http://www.talkorigins.org/faqs/genalg/3dbeam.gif - no human came up with that wonky design, yet it is better).

Real, innovative research is something that most companies don't have the time or money for, and it's a terrible shame. When they try, the banks get on their banks saying "Produce something saleable NOW, it's only a bloody game." But sooner or later, someone is going to have to, lest the gaming experience fade just the way enjoyment of television has. We've seen it all already, we get bored, and when we get bored enough, our money will move from this industry to another one.

Hoops and sticks are about due a comeback.

Edited 1 times. Last edit by Bonnie Patterson on 8th April 2015 11:03am

1Sign inorRegisterto rate and reply
Keldon Alleyne Developer, leader, writer, Avasopht Ltd2 years ago
This is a fantastic article!!!

It's all about balancing creative freedom with pragmatism. It takes equal amounts of experience and shrewdness to get it right but when it's done wrong you either end up with formulaic games that fail to innovate or what we have with Molyneux, which is extraordinary enthusiasm for what could be possible but failing to know how realistic the idea is, particularly before announcing it to the world.

Everyone knows that human level AI is always just another ten years away and has been for decades.

Research is very difficult to fund and monetise because you can never know if and when it will pay off. And worst of all you never know whether you will be beaten to the punch and thus be unable to use the findings of your own research!

Calculus was independently discovered by two mathematicians in different countries at roughly the same time. The same coincidental discoveries are a major risk for the development of new technologies that involved several years or decades of development.

And we must remember that flight was discovered not by the heavily funded operation consisting of the world's best engineers, but by two bicycle repair men who had never finished school!

Edited 1 times. Last edit by Keldon Alleyne on 9th April 2015 7:17pm

1Sign inorRegisterto rate and reply

Sign in to contribute

Need an account? Register now.