Half-Life 2's Technical Details, Cost Estimates

Thanks to Computer Graphics Magazine for its feature on the graphical technology being used in Valve's eternally-awaited FPS Half-Life 2. Among the specifics discussed are innovative paths to graphical variety ("Using the same morph targets sculpted for facial animation, the system automatically alters the facial geometry to create, for example, a flatter or broader nose, or a squarer jaw. As a result, all the scientists, soldiers, and other homogeneous characters appear as unique, differentiated models"), and potential game mod options ("To firmly entrench itself in the future of game development, Softimage will package XSI EXP, a lite version of XSI, with every PC copy of Half-Life 2 [and make it available on the Softimage site this week].") Elsewhere, a Maxitmag interview with Valve's Gabe Newell has him musing: "Last time I checked, we were about $40 million into the project. Yikes, that's a scary number."

Comments

How and Where it could fail.

[*weasel]

The 'How' and 'Where' Valve could fail:

The gameplay won't live up to the overwhelming hype. It could be too short, too repetitive, not enough interactivity, too slow, maps too small -- who knows. 'Daredevil' looked like a sure thing on paper too, and was a terrible $40m investment.

The story could be trash. Most people I know hated when the first half-life devolved into 'Doom' at the end -- when the player hops the portal to dimension-XYZ or whatever it was. The game went from good scifi/action to rubber-monster-movie crap. All the interactivity of the environment was gone, all the atmosphere was gone, all the verisimilitude was gone - jump puzzles were in, ammo management was in, mystic healing goo was in. More of the same is not enticing.

The level design could be crap (which may be necessary to cater to horsepower restrictions that 'interactivity' likely creates). A limited diversity of gameplay could easily sink it, less 'scripted' sequences that made the first half-life classic could work against it. Convoluted maps and missions could easily sink it.

The 'interactivity' could be the exception (eg. happens rarely) and not the rule (eg. regularly appearing feature). If all the stuff they talk up at tradeshows happens 4 times in the game, it means nothing to the player. Particularly after all the hype, it'll create animosity amongst the would-be community (eg. short commercial run)

The game could run like absolute shit on all but the highest end rigs. Do 10 million people even have a PC with enough horsepower to run hl2?

The network play could be lame. Solid network play is necessary to build a community, to drive mod makers, to keep the game hot for years after release. If not for counter-strike, Valve would've sold a few million copies of Half-life and been happy. Counterstrike made it a best-seller for 4 years.

The API could be so complex that mod makers don't have the appropraite tools to actually make anything good before the community evaporates. Without a mod community the game will have a short run, and considerably less beneficial word-of-mouth. Without a long commercial run, it won't stay on the shelves until the mass of gamers finally get rigs that can run it.

The engine restrictions could limit the number of enemies on-screen, or the complexity of AI scripts. People generally don't want 1 enemy at a time in action games (doom3 might be an exception due it's 'horror' premise, or it could fail as well). Likewise players have lower toleranace for 'dumb' enemies, particularly after Valve's success with HL1's grunts and assassins.

They can screw it up. The hype could be smoke and mirrors. I doubt it - but it's far from a guarantee.

$40m dollars is assinine though. But if they can get a share of the licensed engine market - who knows. They probably also subsidized Steam and its infrastructure entirely under the HL2 budget.