In the commercial game industry on the inside most big projects are Art-driven.  This is rather surprising – most developers would say they want to be Design-driven.  Small teams like indies seem to be Design-driven, like Braid or World of Goo, but less often Art driven, like Tales of Tales.  But in my experience most big projects seem to get caught up in the art.  I’ll never forget my interview at Blizzard North years ago.  I asked them which department they thought drove development, expecting them to say Design.  Blizzard is known for having some of the best designed games in the industry, so it seemed clear.  But the interviwer said if anything they were Art-driven.  When I expressed my surprise, he couldn’t help laughing at my shock.  Design-driven was almost a foreign concept.  When I asked the same question to Blizzard South, with a more defined Design culture, they still emphasized their Art.  I’ve heard of this time and time again.


If we all want to be directed by Design, if smaller projects are Design-driven, why are big game projects different?

  • Art departments are usually double or triple the size of the other departments.  Environment art in particular is usually the limiting factor in why many games can or can’t ship, justifying larger teams.
  • Art departments usually have some of the best tools.  Maya is state of the art, and additional effort is spent to improve art tools to increase efficiency and reduce level production time.  Conversely, there’s not even a common consensus for what standard system design tools are, and most teams have few or no systems tools, limiting what design can create and explore.
  • Art gets a bigger reaction.  Audiences love high quality art, whereas it’s extremely difficult to convey large-scale game designs in a demo, particularly to executives.  Thus every game has to deliver high quality art early, but design is rarely a requirement.
  • Art is well researched.  Every school has an art department.  It’s an established field with easy reference and intuition.  Game design is still relatively new, the details of dynamics are poorly appreciated, and playtesting is still rare.
  • Because Art has more people, better tools, and is easier to promote, it can create definitive prototypes more easily then design.  This allows them to more quickly shape the game’s vision.

I do strongly believe art is important.  Critical.  In fact, I believe that art plays an under appreciated role in inspiring and pushing system design.   But the processes game teams follow push Art past inspiring Design and into driving it.  Art is “easier” to see, and that naturally drives it to the forefront, creating a feedback loop.  In part, I think this is why we see unproven game designs even at the end of the project, designs that just aren’t fun.  Games that all feel the same but look incredible, and just get sold back within a week.   And Design takes the blame and gets even less respect.  Nobody on the team wants that, including the artists and the designers.  How can we help Design?  How can we all break the loop and bring the balance back?

Maybe we can wait on art even longer, recognizing they can shine when called on, focusing on smaller early art departments who work primarily on pipeline and support until we’ve got a truly good game.  For that we’d also need people approving games based on design, not art, which I’ve requires rethinking the business side.  We could also create the Maya-equivalent for systems and level designers, so that they can actually iterate and prototype through their ideas.  And we could find a common design language and train people in it, so that developers and executives will recognize good design and demand to see it from their Design departments.  This is all happening, slowly.  We’ve heard many of these answers before, and the top development teams are already executing on these ideas.  What’s new is seeing the production loop that drives this to happen, and controlling that system to make sure it balances both approaches, rather then letting it control you.


4 thoughts on “Art-driven

  1. Dan,

    Surprisingly, we have had this issue come up a few times in our discussions on AI. In the lecture that I am doing with Phil Carlisle and Richard Evans at GDC, we are going to point out that the expressiveness that we want to build into characters through personality, mood, and emotion is entirely limited by what the art and animation can convey. For example, it does us no good to generate 100 shades of emotion if the art pipeline can only generate 20.

    In a similar sense, the GOAP-centric FEAR model made sure that each atomic behavior was an action/animation. They then strung those together. Certainly a designer and an AI programmer could have come up with plenty more actions that they wanted the agents to be able to use in their plans. Again, however, why bother if the constraints of the art pipeline prohibit their construction?

    On that note, I would suggest that games are not necessarily art-driven, but art-constrained. Until that pipeline gets fixed (procedural animation, etc.?) they will remain the bottleneck.

  2. In some cases I think this constraint may also be a question of readability.

    Even with all the man-power and talent in the world an art department can only come up with so many emotional variations that will read as different to a player.

    It’s easy enough to read change, but in isolation such subtle differences are lost.

    It would definitely be great though if we could move to a production model where art waits for design to solidify, because re-doing finished art to meet changing requirements is a waste of time and energy…

  3. Great post.

    I fully agree, most big productions are all about the art, and pure design suffers, in comparison.

    I think you hit the nail in the head when you say that art gets a bigger reaction. That’s not only true for the executives or the audience: most designers nowadays are all about the art. A combat designer often gets distracted with animation aesthetics, or fx, and lets it drive the design itself, instead of trying to incorporate interesting mechanics. A level designer might ramble about the beauty of a cutscene that will happen at a certain point, of about the detail of the landmarks in a level, instead of focusing on pacing and gameplay. Sure they will think about these things in the end, but by then a lot of work has been poured on building the art.

    I sure hope you’re right about design playing a bigger role with more recognition and better tools, but I think it’s going to take a long, long time.

  4. Nice post Dan!

    I look at it this way.

    Gameplay and art are both subjective, but it’s easier for the people approving your milestones (and writing the checks) to look at a game and make a business decision based on graphical fidelity.

    I strongly believe that designers are the first and last people to touch a game in the development process. However, the art department usually reaches a point of no return around the half way mark in the production schedule. This observation is based on the fact that the pipeline and schedule for art do not allow for the iterations or often major revisions needed to support design in the crucial last few months of development where gameplay finally gels into something more representative of the final product.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s