Determinance screenshot

Monday Night Live: Devil In The Details

This week’s guest poster is David Rosen of Wolfire Software, the man behind the incredible Lugaru.

“I am always confused by the trend in big-name PC titles to have load
times lasting several minutes and system requirements that exclude all
but the latest and most expensive computers. Good graphics and high
system requirements do not necessarily go hand in hand. For example,
there are several games on the PS2 such as Final Fantasy 12, Okami,
and Resident Evil 4 that arguably look better than any currently
available next-gen titles. If we design games to run on high-end
computers, then most people will have to turn down the detail to be
able to play the game. A game with detailed graphics looks very bad
when the resolution is turned down, much worse than a game that is
designed from the start to run at lower resolution. Therefore,
targeting the high-end can actually make the game look *worse* for
most players than targeting the middle.

When we spend less time on surface details, we can devote more time to interactivity and interesting behavior. For example, in The Elder
Scrolls: Oblivion, rooms are cluttered with detailed objects simulated
with the Havok physics engine. They react to physical forces, but they
are only interactive in a very superficial sense. That is, if you find
a pile of apples then you can knock it over, but you cannot eat the
apples, squash them, use them as projectiles, or really do anything
interesting with them. This kind of superficiality is becoming more
and more common in next-gen games; it is not unusual to see
hyper-realistic environments that the player cannot affect in any way.
As the number of objects in the environment increases, so does the
amount of work needed to make them behave in an interesting way, so
often interactivity is sacrificed entirely, or just offloaded
indiscriminately onto the physics engine.

In my last game, Lugaru, I had to take the exact opposite approach. I
created most of the artwork myself in my spare time from school, so I
had to make do with very little artistic content. There are really
only nine important models: two character models (rabbit and wolf),
three weapons (sword, staff and knife) and four environmental models
(rock, cube, tree and bush). Because there are so few objects, I could
focus on each of them and add a lot of detail that would otherwise
have been impossible. For example, the environment is not just there
for decoration; you can perform acrobatic tricks off of walls to
confuse opponents, kick an enemy’s head into a rock for an instant
kill, or use snow to clean blood off your knife so the wolves won’t
smell you as easily. There are also many unusual cosmetic details I
could add because of this sharper focus: you can kick an enemy’s tooth
out and blood will run down his mouth, or throw him at a wall so hard
that cracks appear in the rock.

Now that I am working with a team (and have some idea of what I am
doing), I hope to be able to harness the best of both worlds in future
projects. In Lugaru 2 we will have a wider range of environments
including towns and cities, but we will only add objects that are
interactive and interesting. If there is a bucket on the ground then
the player must be able to use it as a container; otherwise it should
not be there at all. This kind of interactivity is not especially
difficult to implement, makes the game world more interesting, and has
negligible effect on system requirements, so I hope that it becomes
more common as the novelty of hyper-realistic superficial detail starts to wear thin.”

6 Responses to “Monday Night Live: Devil In The Details”

  1. malakian:

    Very relevant and interesting points there, Davey! Can I call you Davey? Well it’s too late now.

    Just the other day I finished the new and hugely rated Gears of War. In one level, you had to shoot some platform supports off, and in another you had to goad this big beast into knocking some pillars down. This I found contextually ridiculous, as everything else in the game would not budge. Instantly this cheapened it for me, as I at least hopped on the train of thought, “well this game boasts to be highly tactical, why can’t i down a wall and fashion a shelter out of the rubble if I can fool some stupid blob into knocking pillars to pieces?”. Five years ago this could be relatively easy to forgive, but when taking into account this is the next generation, I think that its either showing a brick wall for creativity or effort. Seriously, without a few of the nicer textures or shaders, i’m sure the xbox could have pulled it off. I’m getting into a rant about something slightly on a tangent here, but what I’m getting at is that nowadays it seems the world is simply there to [i]look[/i] realistic – until the moment you’re playing it, which I guess is when the manufacturers have done what they need to anyway.

    I haven’t heard of your game before, but it certainly looks an interesting concept and I will check it out when I have a free moment. BUNNIES!

  2. malakian:

    THOSE ERRONEUS TAGS WERE DELIBERATE. AS WAS MY POOR SENTENCE STRUCTURE.

  3. Paul:

    Good points all round, I feel. I really think that it’s the place of indie developers to take “en vogue” things in the industry (like physics simulation) and actually explore the gameplay possibilities of them fully, rather than simply incorproating them for the sake of a kind of gimmicky appeal.
    Ian and I talked a little bit last night about wanting to “do stuff” in games – people always look at a game and say, “But what if I could do this?” and good designers use that curiousity as part of making their game fun. But giving the player choices is always difficult, in that you have to retain balance.

  4. malakian:

    I dont quite get what you mean with your closing sentence there, mr T. Giving players choices is difficult? Sounds vague AND political!

  5. Paul:

    Just a fairly bland truism that the more “on-rails” your game is, the easier it is to balance.

  6. dr_pineapple:

    I wish I had the time to read this :O Gotta read later and comment it then.

Leave a Reply