Luke had a lot to say. Friday 4:00pm be damned. He worked me like dough.
Organized, Seurat detail, pant dropping videos - all delivered with a casual smile.
Hoped we would expose and dismantle common game design myths.
Watched Luke blow up the world real good.
Luke detailed stuff learned building destruction into Red Faction 3.
Unlike previous versions, game lets you destroy ANYTHING
... even hope!
Building a game with physics is one thing,
building a game with physics + destruction is another.
Presentation chock-a-block with info. Dense.
Enough bullets per slide to demoralize competition.
Personal games can be described in 1 bullet. Feel flaccid.
Presentation littered with videos showing... destruction.
Show solid, gigantic, glorious structure.
Partially destroy foundation... building lurches.
... destroy more... building lurches further, precarious.
... destroy more... structure tilts, slides, then slowly collapses.
World shakes. Camera shakes (100 different ways).
Debris flies... hits you. You're rag-dolled hundreds of yards.
Actually, there were other variations, but did you read that?!
Limit destruction enemy A.I. can cause, because player can easily fool it.
i.e. Trick enemy tank into shooting bridge above boss.
i.e. Trick enemy tank into sleeping with player.
When something explodes, A.I. must react (turn heads, shout, run away).
Costs little, makes A.I. look smarter and destruction feel more legit.
Projectile debris kills enemies, but never the player.
Feel there is a larger life lesson here.
Large destruction calculated differently than small, otherwise CPU would grind to halt.
Computers - will they ever be fast?
Vehicles can bust through walls...
... so physics demands that wheels damage ground slightly to move vehicle.
... so wheels harm bridges while driving over them, eventually causing collapse.
Still resolving why bridges stand. Somewhere in hell... Erin Catto laughs.
Stress is stressful. Humans can tell if stress is calculated properly.
i.e. Large buildings with few supports must collapse.
Without stress, players felt destruction was lovely, but not realistic.
Stress made instablity and self destruction a constant battle for everyone.
i.e. Stress made me hide in washroom when website failed.
Muchos destruction means overwhelming sound and shaking. Not good.
Muchos time was spent ensuring sounds DON'T play and camera DOESN'T shake
- even if game engine says they should.
Problem: Within 10 minutes, everything is levelled.
Solution: Rounds last 10 minutes.
Problem: Fast vehicles require predictive networking.
(Extra hard since destruction requires authorative server).
Solution: No fast vehicles in multiplayer.
Problem: A.I. programming difficult and CPU intensive due to constant destruction.
Solution: No A.I. in multiplayer.
Single player game has power-ups (mech suit) that allows player to go bananas.
Multiplayer has modes where the only goal is destruction.
Luke knows what side my bread is buttered - the destruction side.
Some kleenex testers were so conditioned to regular FPS limitations,
they never tried to blow anything up.
Needed to teach players to use destruction, game now forces this.
i.e. Don't shoot enemies, drop a ceiling on them.
"rewire brain from run & gun"
Destruction is the ultimate emergent system. Results in memorable moments.
Showed video of player using heavy sledgehammer to bat cars into buildings
(used proper physics, not scripted). Have next award winning IGF idea.
Game had 6 month post production cycle which Luke strongly encourages.
Game was ready at deadline? Preposterous! I deem thee... LIAR!
Destruction affects everyone on team - including artists and audio.
Can't just build destruction, need to build full game.
Unless you're building bad game, or you're Data Design Interactive.
Design elements to interact with physics and destruction.
Give players a reason to use it.
No-one cares if you bring structures back to life after restarts.
Who doesn't want more stuff to destroy?
Destruction + Physics elevate each other.
Like midgets and tossing.
Let go of scripting and let unique stories emerge.
Creates a whole new "axis of interactivity".
Ears bleed. Since affordance not used during lecture, cutting Luke slack.
Ended with "large smokestack tossing" - bug which turned into full level.
Destroying base of smokestack caused it to fly through air wrecking havoc. Hilarious.
Q: How well do your materials map to real life?
A: Not at all.
How can you say No to running through a collapsing building,
knocking out a wall with a sledgehammer,
jumping to another building, busting through the window,
and surprising some poor guards?
HOW CAN YOU!?
Look forward to playing multiplayer with Jenova Chen, Jason Rohrer and Keita Takahashi.