Experienced Points

Experienced Points
The Big Cost of Small Places

Shamus Young | 13 Jan 2012 21:00
Experienced Points - RSS 2.0

Pathing the space

In the old days, the level designer just dropped path markers around the level to show the bad guys where they could go. When a fight started, the bad guys would follow this trail of invisible breadcrumbs to the player. Now it's much more complicated. Bad guys need to understand where cover is, what objects can be vaulted over, and where the destructible objects are. They need to understand vantage points. (Shooting at the player from the upper catwalk is better than running downstairs to play peek-a-boo over the bank of computer consoles.) The AI needs clues as to which cover nodes point in which direction, so that they don't take cover on the wrong side of an object or go into cover where they can't shoot at the player. AI has come a long way in the last ten years, but it hasn't come so far that the enemies can figure this stuff out for themselves. The level designer must set all of this up.

Scripting the space

In the really old days back when Doom and Quake roamed the earth, elevators were nothing more than cubes that went up when stepped on, waited for a second or two, and then came back down. They were wonderfully easy to set up, and if the level designer was feeling really fancy they might throw a button in there.

Now we have elevators that are expected to behave like their real-world counterparts. Call buttons, moving doors, floor selection. There needs to be a bunch of checks and safeguards so that it works right when things go wrong. (What if the player stands in the door as it closes? What if an NPC is trying to reach the player on the elevator? What if the player selects a floor and then jumps off the elevator again?) Plus, it's more or less expected that videogame elevators should have windows with lights outside so the player can experience some sense of motion while in transit.

We need to set up all those hackable keypads and link them to whatever doors they open. We need to set the soda machine to dispense drinks, the toilet to flush, the lights to go on and off, the phones to play messages, the computers to display things or respond to player input, metal detectors to beep, televisions to change stations or turn off, the fountain in the lobby to have flowing water, and all of the other details that make these modern worlds feel so much more alive.

All of this takes programming / scripting to make it work right. Every level becomes a little game world in its own right, with unique code and interactions that usually can't be reused elsewhere.

The cost of space

Every step towards photo-realism brings with it a cost in fidelity. It's fine if the people of Hyrule don't blink, or if Mario doesn't get dirty during his adventure, or if Jade of Beyond Good & Evil doesn't see a shimmering reflection of herself in a small puddle. But in a world shooting for gritty realism, increasingly extreme steps need to be taken to keep the visuals from plunging into the uncanny valley.

Looking over this list, we're talking about things that are a ten or even twenty-fold increase in work. It's more than ten times the effort to create the same "one room" of playable game area that the player will inhabit for about a minute. This should explain why today it costs five times as much to make a game that's one-fourth the size.

This is a big part of why I'm really glad that this console generation is lasting so long. Sure, it's possible right now for Microsoft or Sony to roll out a console that's far more powerful than what we're used to, but I don't think most developers can afford to make games at the next graphical level. Heck, most of them can barely afford the one we're on now.

Shamus Young is the guy behind Twenty Sided, DM of the Rings, Stolen Pixels, Drawn To Knowledge, and Spoiler Warning.

Comments on