San Jose Mercury News columnist Dean Takahashi has written a second book about Microsoft's foray into the console gaming business. Released as an eBook, The Xbox 360 Uncloaked: The Real Story Behind Microsoft's Next-Generation Video Game Console also offers some interesting perspectives on the development of Bungie's Halo franchise.
That the fate of the new console should be tied to Bungie's flagship game is certainly no shock; since 2001 Bungie's series of shooters have been system-sellers for the Xbox, and Halo 2 has had a stranglehold on the top spot in Microsoft' Xbox Live online gaming service since it was released in late 2004.
Takahashi's book claims that Bungie became caught in a struggle between the hardware and software sides of the Microsoft Games Division over how to best promote both the platform as a whole and individual games, and that eventually Hamilton Chu, Pete Parsons, as well as other Bungie staffers, and Ed Fries, who spearheaded the buyout of Bungie by Microsoft, were all casualties of it in one way or another.
The below text is a sort of "executive summary" of the book, including many the points where the stories of Bungie and Halo intersect with the business of Microsoft Games Studio and the division of Microsoft that makes the Xbox and the Xbox 360. Excerpts from Takahashi's book are reproduced here with permission of the author. All quotations and paraphrases of statements by personnel currently or formerly of Microsoft or Bungie, whether identified by name or not, are taken from the book; no one contacted for this article wished to comment for the record at the time of its publication.
Click "read more" from the front page for the entire article.
Anyone with an interest in the Xbox and the games industry beyond just Halo must have wondered what the thinking was behind the scheduling of Halo 2. Released in November 2004, it broke single-day sales records and delivered the division's only profit-generating quarter. But with only a year to go before the launch of the Xbox 360, it put Bungie out of the running for developing a launch title. When Peter Moore unveiled Halo 2's release date tattooed on his forearm, he was essentially telling the market that the Xbox 360, Microsoft's next-generation console, was going to launch without a game by its premier in-house development team, Bungie; without a sequel to the system-selling Halo game.
In fact, the pressure on Bungie to whip up some more Halo magic had started shortly after the first game shipped. Takahashi writes:
[J] Allard wanted Fries to put Bungie to work on another version of Halo, either an expansion pack or a version of the first Halo that worked with Xbox Live. But Bungie had an independent streak and it always tried to make a big leap forward when it undertook a sequel. For Halo 2, the studio wanted to do something spectacular. Fries agreed with them.
With Bungie inside Microsoft, Fries staffed up the division and poured more resources into Halo 2. Getting this game out the door was the top priority for the current generation Xbox, which sorely needed more hits.
Takahashi p. 101
For Bungie to have to choose between doing a full sequel and an expansion pack is not a new thing, although more often than not, Bungie has eventually opted for the former.
However, doing a new engine for Halo 2 would obviously be more work than just an expansion. It would take more time. Assuming that a third game, if there was one, would also need a new engine because the new console would be completely different hardware, this meant that all three games in the trilogy would require a lot of development time and be substantially different engines.
Was Halo 3 planned as far back as that? Takahashi's book seems to indicate it was:
"It was unthinkable to J [Allard] that Halo 3 wouldn't ship at the launch," said one observer. "He had a strong core belief that engineering could be managed. J's position was, this is Bungie's job. Go do your job. This is the single most important thing. When we tried to explain it to him, he couldn't rationalize it."
Takahashi p. 103
Bungie's Joe Staten recently revealed to students at his alma mater Northwestern University that three levels were cut from the end of Halo 2 because they were not finished. Content from these levels, presumably, are to be put in Halo 3. However, if there was a Halo 3 already planned, as is attributed to Allard above, are those levels being added to the beginning of the new game? Or was Halo 2 supposed to be the end, and Halo 3 is going to be those three levels, possibly stretched out with additional new material, to form a third game in a series that was only supposed to be two?
According to the book, it seems not everyone at Bungie was all-fired up to do another Halo game. Jason Jones, the head of the studio and remaining founder after the departure of Alexander Seropian, it says was one of these.
Bungie was split into two teams; one to work on the next Halo project, and one to work on the project codenamed "Phoenix" and referred to in public by Fries as "Fantasy Siege". Perhaps just as the engine that was supposed to power a new Myth-like strategy game with a science fiction theme eventually evolved into a shooter, Halo technology would come full circle and be used in this new game:
Fries decided to farm out the PC version of Halo to Gearbox Software, a game studio in Texas. He wanted the rest of Bungie to focus on exploiting the Halo franchise with a sequel. But Jones wanted to make another original game that involved Minotaurs, a topic that went back to Bungie's roots. He appointed others to run Halo 2, but it wasn't going well at first.
Takahashi p. 197
At about the same time that Nick Baker and Jeff Andrews started dreaming up the Xenon chips, Bungie got started on its next games. It had devoted one team to a game where players laid siege to a medieval castle. That project wasn't working out, and Jason Jones canceled it. Bungie had access to nearly 70 people now. But some of the early magic was dissipating. Against his initial wishes, Jones had to pour almost all of Bungie's available resources into Halo 2.
Takahashi p. 142
Could Bungie have worked effectively that way? Would Halo 2 have shipped any earlier? Any later? What chances did Phoenix have to be a successful game, and what were all the factors considered in its cancellation?
We may never know. At the time it was explained that the game simply wasn't fun. The beneficiary of that cancellation, whether by providence or design, was Halo 2, as all the staff was shifted to that project and Jones took the reins, says Takahashi. The shift of manpower here is very likely the source of the persistent rumors that Bungie had doubled in size. While Bungie hadn't actually grown that large, the cancellation of Phoenix did just about double the amount of manpower that could be devoted to Halo 2.
Pete Parsons came over to Bungie from Microsoft to be the buffer zone between the two very different entities-- the same role played earlier by Seropian, who since the founding of Bungie had been more involved in the business side of the operation.
According to the book, Microsoft was now pressuring Fries to say something about Halo 2 in time for E3 in 2002. For those keeping score at home, that was a mere seven months after Halo 1 had shipped, a far cry from the 18-month silent period that would follow Halo 2.
The Bungie team worked away from the bottom up. "There was a fundamental difference between a top-down organization like EA," Seropian said. "Bungie was an extremely bottom-up company that focused on making a product as good as it could be." Microsoft didn't force directives upon Bungie, with the exception that the original Halo had to be ready for the launch. That had its drawbacks. Without the pressure to ship something by Christmas in order to survive, Bungie could take its time. Yet it lacked a planning process that set timetables for future games such as Halo 3.
Takahashi p. 197
Meanwhile, the scheduling for Halo 3 apparently took a backseat to finishing Halo 2:
For the Xbox executive team, it was important to get Halo 2 out the door during 2003. The platform needed it. Microsoft was losing to Sony. Xbox Live had launched in November, 2002, and it was generating paying subscribers for online games. But Microsoft needed a game that could give Xbox Live a big boost. Halo 2 was such a game, but it was falling further and further behind schedule. Fries pushed back and got more time. Microsoft said later that Fries didn't deserve the sole credit for securing this extra time. But it was clear that Halo 2 was a disaster in cross-divisional communication, said one team member who had to fight the battle alongside Fries. One Bungie veteran said, "I don't care about the platform. I don't care if we're with Microsoft or Sony. I care about the game. And Jason Jones, he's not a sequel guy." It didn't help matters when Hamilton Chu, lead producer at Bungie, left the project and, ultimately, Microsoft altogether. At the beginning of 2003, Jones took back the reins. The work was hard on the Bungie staff, and at one point, Jones decided the missions had to be redone. "It wasn't just about making new levels," Parsons said. "It's about rewriting the game from the ground up."
Takahashi p. 198
Rewriting the game from the ground would certainly cost Bungie precious time. Takahashi writes that since it would now not be ready for the 2003 holiday season, Fries was prompted to go back and bargain with Microsoft again to give Bungie more time:
Now he had to push the launch date for Halo 2 from the fall of 2003 to the spring of 2004. Even then, Bungie was falling further and further behind schedule. To make the new schedule, Jones decided to chop the story in half. The second part would become Halo 3. It would leave the game with a cliffhanger ending, but that was all they could accomplish. Jones asked for even more time. He said his team could hit the April, 2004, ship date, but they would probably all quit once it was done. He even suggested that Microsoft hold the game for the Xenon launch in 2005, but nobody liked that idea.
Takahashi p. 199
It could not have been a decision made lightly. Microsoft's financial year 2003 would end on June 30. Letting Halo 2 slip until the fall meant pushing it into fiscal 2004, surely throwing off revenue projections by a good deal. That would seem the only reason not to follow Jones' suggestion to delay and make Halo 2 a 360 title, a question asked by fans more than once.
It would mean that potentially Bungie would have an engine optimized for the new console at launch that could be used in further sequels for faster turnaround time. It would mean having a sure-fire system seller for the new machine's launch. But it would put a black mark next to the Xbox's place in the financial ledgers forever. The books would close on the original Xbox console without it ever having turned a profit. Those outside Microsoft may never know the true level of the institutional commitment to the Xbox as a platform; the profit generated by Halo 2 might have been the difference between forging ahead and throwing in the towel.
At first, Takahashi writes, the Xbox management team, including Chief Xbox Officer Robbie Bach, wouldn't go for it:
Around July, 2003, Bach put the decision in front of the whole executive team. Fries didn't like that, since it was a product development decision. He had always run his business autonomously, and this was clearly a game decision within his division. But he went along with Bach's move. Bryan Lee, J Allard, Peter Moore, and Mitch Koch all weighed in. They wanted Bungie to finish the game by April, 2004, in spite of Fries' arguments that it would destroy the best franchise. Bach said that the needs of the platform outweighed the needs of any one game. When Fries left that contentious meeting, he thought for the first time that his career at Microsoft could come to an end.
Takahashi p. 199
Is this why Microsoft refers to Halo 2 as an example of doing things right, and not forcing developers into releasing a game before it is done? If they had their way, according to the book, Bungie would have been forced to ship Halo 2 back in 2003, in time to start working on an Xbox 360 launch title. What is funny about that is that some fans still feel to this day that had Bungie been given even more time, Halo 2 might have been more polished; or, at the very least, would not have ended with a cliffhanger, leaving many feeling the story was unfinished.
That would not be the last word, of course, as we all know Halo 2 did not end up shipping in April of 2004; but getting that extra time to finish the game-- even with some of its levels supposedly removed and postponed until Halo 3, was costly:
When the committee voted to solve the problem by forcing the talent to ship an unfinished game, all to meet the needs of the platform, no one was happy. Not Fries, not Bungie, Not Robbie Bach. The committee made the wrong decision. It changed its mind and eventually made the right call. But it was too late to prevent Ed Fries from leaving the company, as well as a number of Bungie employees. Art had clashed with business. The budget and the strategy weren't in alignment with the tactical situation on the ground.
Takahashi p. 389
Until Moore announced it, there was no public release date for Halo 2, so from the perspective of the public, there has never been a "delay" of Halo 2. But Takahashi alleges that there were those inside Microsoft and Bungie who had plans and expectations, at various points in time, that included Halo 2 being finished earlier than late 2004.
"Well, fellas, I got your release date right here," [said Moore] showing off his tattooed bicep with the date on it. "November the 9th, the moment Xbox nation has been waiting for, Halo 2. And yes, that is 2004." No one knew just how much tension was behind that date.
Takahashi p. 244-5
What happened on November 9 any good Halo fan knows. Shipping simultaneously almost worldwide, the game sold $125 million in a single day-- not bad for a game that supposedly cost about $30 to $40 million to make, according to Takahashi. The division of Microsoft that included the Xbox and its games made a profit for the first quarter ever.
Fries wasn't around to see it. He had left Microsoft months earlier. And he was not the last:
The toll on Bungie was high. Hamilton Chu, who had been producer on Halo and Halo 2, left some time before the game shipped. Michael Evans, engineering lead on Halo 2, moved over to work on Perfect Dark Zero and later left Microsoft. Pete Parsons, the general manager of Bungie, also eventually went on a sabbatical. Those who knew him didn't expect him to return. The industry insiders knew that many at Bungie had burned out on Halo 2, but to the public at large, all seemed well. "If you noticed a lot of people left, that was because of burnout," said one veteran Bungie insider.
Takahashi p. 274
Most of those playing Bungie's games today most likely heard of them because of Halo. It's understandable; previous efforts focused primarily on the Macintosh platform prior to Myth because that's where Bungie's interest lay and where there was opportunity. That the Xbox, despite having an installed base far smaller than its main rival, the PlayStation 2, should still end up outstripping the audience for hardcore first person shooters on the Macintosh is not a surprise.
The other fans ardently hope that Bungie is still Bungie; that after countless hours, millions of dollars, and all the pressure that comes with it, that there is still something different and ineffable about Bungie that cannot be extinguished. Takahashi seems to hope so, too; and he believes that if there is, it will be the key to the success of the new console, as it was for the old:
If it is to win the console war, Microsoft will have to enable Bungie to finish Halo 3 on time, on budget, and to the satisfaction of the artists and the business executives alike. Microsoft will have to bring its left brain together with its right brain. It has to make the line toward graphics realism and the line toward creativity intersect. If it can make that happen, then Halo 3 will be a great game. And more great games will follow. That's the only way to win.
Takahashi p. 389
And then, of course, there's the question of what comes next, even at this early stage. At E3 Bungie finally admitted that Halo 3 has been in development since Halo 2 shipped, and now we know that they expect to ship Halo 3 sometime in 2007. And somewhere in Microsoft, it's someone's job to pencil in what the next Bungie game will be, when it will come out, and for what platform.
If Halo 3 ships in November of 2007, the Xbox 360 will be celebrating its second birthday. How many years will be left in it? Certainly Sony's PlayStation 2 is a bit long in the tooth now, but with an installed base of over one hundred million units it is a viable development target even now, more than six years later.
Microsoft is attempting to speed up the console cycle, pulling the plug on the original Xbox after only four years, competing admirably for sales against Nintendo but failing to put much of a dent in the lead built up by the front-runner Sony. How long with the 360 last? Long enough for Bungie's next title to ship for it? Not likely if the development cycle is just as long. If Halo 3 does indeed ship in November of 2007, that means each Halo game has taken about 3 years or so to make. That would give Bungie until November of 2010 to produce another title, at which point the Xbox 360 would be five years old. How close will that be to Microsoft's next console launch? Will Bungie again be depended upon for a launch title, as they were the first time around, and as some within Microsoft expected them to do for the 360? Or will they be allowed to work at their own pace and continue to choose their own projects?
Is there a Next Next Project already? Did Bungie divide into more than one team after shipping Halo 2, as they did after shipping Halo 1? Will Phoenix rise again, or is there another insanely great idea brewing at Bungie?
We probably won't know for awhile; but if Bungie's history is any guide, the finding out will be fun, regardless of when it happens.