Page 2

A quick warning

We should make clear that the following will be a collection of guesses, speculations and conjectures based on what we saw from the E3 demo. Though it appeared to be running on a live computer, it was a non-interactive demo and many of the events were likely set up to look better than they would normally. For example, when Gabe was highlighting material properties (like wood being made of digital wood) Gordon Freeman turned his gun on three wooden boards and shot them. They appear to shatter at the points where they were shot, but it’s difficult to tell since there is no crosshair. Even if there was one, the demo could have simply been designed to point the camera at the right spots, where the designers knew the boards would shatter. This is just one of those many things that can’t be identified in a pre-scripted demo.

Other features, like the scientist telling Gordon to ‘be careful’ when he knocks over a monitor, are scripted events. The physics of the monitor being knocked over might work, but we somehow doubt that the AI of the scientist is capable of recognizing that the object got knocked over, never mind its value to said scientist and having him subsequently formulate a sentence to express his dismay. It’s a great bluff, and as Tom points out in his E3 overview, one of the pitfalls of non-interactive demos.

Technically, Valve could have scripted every single event in the demo and we’d be none the wiser. However, we’re going to proceed on the assumption that much of what we saw was a collection of recorded demo events in the fashion of Quake demos, not a line-by-line script. In this case we expect that of course the events shown were set up but representative of actual gameplay.

No compromises

The first thing that is evident upon reflection is that Valve took an absolutely no-compromises approach to Half-Life 2. They didn’t make any announcements, never mind wild claims after Half-Life became a smash hit. In short, Valve didn’t back itself into a corner and set up the public for disappointment over a long development period or unfulfilled features. There was no scrutiny, no press and no outside interference from fanboys arguing about which features have to be in and which will suck. Clearly, they learned a lot from the development period of Half-Life and the early hype surrounding Team Fortress 2.

Having thus guaranteed themselves as much time as they need, and clearly having enough money to buffer out publisher interference, they set out on a nearly 5 year long development project. Doubtless, in future interviews after the marketing engine goes into full swing, you’ll hear comments like “we went in with no pre-set ideas of what we could and couldn’t do” and “we weren’t going to make any compromises in the design because something had never been done or was difficult.” If the game really is as good as it looked in the demo, they’ll also be telling the truth.

Thin sheets of metal bend as a spray of bullets hit them. Wood shatters, mattresses and bodies float in water, and blasts and monsters have various levels of strength, mass and momentum. A headcrab might not be able to bash a door down, but a zombie might. He might also know he should and how to do it. Valve claims it won’t be a scripted event, like in the original game. Reinforce the door with some furniture, and it might take an RPG or a strider (giant walking robot) to bash it down.

On top of that, there is the more to consider. An impressively paced game in the vein of the original can be expected, along with advanced group AI, creative weapons and monsters, and an intense vehicular experience. Finally, if Valve delivers on its promise to release the best multiplayer tools they can…