The Final Hours of Half-Life

 
Part 1 - Introduction
Part 2 - The Microsoft
            Millionaires
Part 3 - The Valve
            Difference
Part 4 - Reassembling
            the Pieces
Part 5 - Squashing the
            Final Bug
The Public Debut
"We did a lot of growing up in public," admits Gabe Newell. When Valve and Sierra jointly announced the game last year, they released two screenshots that bear little resemblance to the images in the final game. And that's a good thing. "The screenshots we put out when we announced the game were terrible," concedes Newell.

Time passed, and the images got better. At the 1997 Electronic Entertainment Expo, Valve had a tremendous showing, demonstrating its high-end AI and Birdwell's skeletal animation system. "We had pretty interesting technology to show what we were doing with the Quake engine," says Newell, "but we didn't have the game yet."


Houndeyes populated the second publicly released screenshot, which Newell admits, was "terrible."
That didn't stop the press or buyers from christening Half-Life as the Next Big Thing. Rumors had the number of preorders for Half-Life in the hundreds of thousands of units after the game was shown at E3. It was even voted as a "Best Game" at the show despite the fact that there wasn't really much of a game on display. The reaction was misguided, but understandable. Awash in a sea of Quake clones, the media was so excited to see original thinking in the genre, it wanted to support the product as much as it could.


"The early showings and screenshots [of Half-Life] never really got us very excited."

- John Carmack of id Software.
Not everyone was so optimistic. John Carmack at id Software remembers that he was still very skeptical about Half-Life: "For whatever reasons, Half-Life was the license we paid the least attention to during development. The early showings and screenshots never really got us very excited."


Scripted sequences such as this one were shown at E3 and won over critics.
Energized by its success at the show, the team pushed forward with development in hopes of getting Half-Life ready for the holiday season. As Scott Lynch at Sierra Studios remembers, "The big competition last year was Quake II, and there was a lot of push to get our game out there to go head to head with it."


Levels such as this one had to be completely scrapped because they didnít live up to expectations.
It wasn't to be. By August of 1997, Valve recognized that finishing the game for the holidays would mean making major compromises on the product. "We realized that to make Christmas, we would have to give up a bunch of stuff we wanted to do," explains Newell. "We had to make a decision about what way we wanted to go, and it was a scary decision. We're a self-funded company, so when we pay people's salaries, I write a check out of my personal checking account."

Money wasn't the only concern. The relationship with Sierra Studios was also conceivably at risk. Half-Life was supposed to be its biggest game of the year, but it wasn't going to be ready. The massive wheels of promotion and PR had already started to turn, and now the company had to put the brakes on. It seemed like a disaster.


The bull-chicken enemy shown here would eventually be re-designed.
Actually, it was the best thing that could have happened. With the holiday pressure off, Valve could spend time evaluating its progress, painfully reviewing every aspect of the game. In essence, the work on Half-Life was weighed in the balance. And it was found wanting. Near the end 1997, privately and behind closed doors, Valve decided that most of its work - including the work on levels and AI - would be completely scrapped. It just wasn't working.

Next: Part 4 - Reassembling the Pieces