The Rise And Fall Of The Dreamcast
by Douglass C. Perry [Business/Marketing, Interview]
September 9, 2009
[In this ten-year Dreamcast retrospective, Gamasutra looks back at Sega's last effort in the console market through interviews with former president of Sega of America Bernie Stolar, former Sega of America COO Peter Moore, former SOA Vice President of Communications Charles Bellfield, and former vice president of Electronic Arts, Bing Gordon.]
For a console that broke entertainment retail records, made the Guinness Book of World Records, and laid the blueprints for today's online-centric consoles, it's striking to think the Dreamcast's lifespan was shorter than nearly any console in video game history.
Ten years after 9/9/99, the memorable date of the launch of the Dreamcast in North America, Sega's machine has left a lasting legacy in online gaming, retail history, and the sports genre. But the brief, fiery life of the Dreamcast was fraught with conflict, questionable executive decisions, and ultimately, a shocking and abrupt ending.
A Change in Attitude
The video game world into which Sega launched the Dreamcast was vastly different than today's highly connected wireless experience. The arcade market was still successful, 80% of consumers connected to the internet used a modem, and the PC market was at its peak -- and, more importantly, was the sole domain for online games.
After a successful Japanese launch in late 1998, Sega looked toward the North American market to achieve a head-start over its biggest competitor, Sony Computer Entertainment America, by growing a strong install base and by rebuilding excitement for its products.
In 1995, Sonic the Hedgehog was better known than Disney's Mickey Mouse, but the Sega Saturn, from its disappointing launch to its inevitable cancellation, had soured many gamers on Sega products.
In 1997, Sega hired Bernie Stolar, fresh from his role as president of Sony Computer Entertainment America, as the new president of Sega of America. Stolar was a shrewd, successful businessman who knew the games business from his time working at Sony, the arcades, and at Atari.
In a phone call with Gamasutra, Stolar, currently running Getfugu, Inc, explained how it all started. "Saturn, as you know, was a failure. I was brought in to help restructure and rebuild Sega of America. When I started, there were over 300 people; I trimmed the company down to about 90 people."
Among others, Stolar brought in 17-year Reebok executive Peter Moore, former Sony third-party executive Gretchen Eichinger, head of sales Chris Gilbert, and Charles Bellfield, as well as several other important figures.
Stolar's task was to wage an uphill battle with gamers, many who had bought the short-lived 32X, Sega CD, and the Saturn, and retailers, who were still wincing from Saturn sales and an exclusive launch that cut many retail chains out of the picture.
"We had to change the attitude of retail to believe we were a serious player," said Stolar. "And because of the whole Saturn thing, retailers really hated Sega. It took me a lot of work to change their minds. I went to every retailer and told them this was going to be a great system, it was going to have a modem, it was going to have online play, this was the content it was going to have, and this was what it was going to look like. They all bought into that. They all trusted me. Plus, they really liked the team I put together. They felt this was the right team."
Before Stolar could change anyone's minds, however, Sega had to make up its own mind.
In 1997, Sega of Japan tasked two engineering teams to compete for the design of the Dreamcast. Internally, Sega's President Shoichiro Irimajiri assigned Hideki Sato, who had designed the Saturn, to come up with a chipset design. Externally, Irimajiri created an 11-man "skunkworks" team outside of Sega to create a competing design, led by IBM alumnus Tatsuo Yamamoto; that project was codenamed Blackbelt.
Sato chose the Hitachi SH4 CPU architecture and VideoLogic's PowerVR2 graphics processing unit, manufactured by the Japanese company NEC. "The first version of the Dreamcast was water-cooled; they had a non-moving part, water-based cooling system. It was called Dural," said former Charles Bellfield, former Sega VP of communications (1998-2000) and VP, strategy & corporate affairs (2000-2003).
Yamamoto, based in the U.S. and initially kept secret from Sato's team, chose the IBM/Motorola PowerPC 603e, but was later asked to use the Japanese-made Hitachi SH4, and entered into a contract with the American graphics card maker, 3Dfx, to use a custom version of its Voodoo 3 card as the graphics processor.
"I was at that meeting on July 4, 1997, in Haneda, in the Sega Tokyo offices, where we were sent to present the PowerVR technology," explained Bellfield who, before working at Sega, was a brand manager at NEC Electronics.
"We lined up a series of games running on the PowerVR technology on PCs, which included Tomb Raider, a game called Ultimate Race from Kalisto, and Looking Glass' Flight Unlimited. The presentation basically messaged that the PowerVR technology would deliver high performance at a low cost."
According to Bellfield, part of the selling point of PowerVR was its tile-based rendering solution. Polygons that were not seen on the screen were not rendered, which reduced CPU overhead. The PowerVR solution, unconventional for developers at the time, in theory was a high performance and low cost solution. Bellfield added: "Sega's relationship with NEC, a Japanese company, probably made a difference too."
The Dreamcast also included a modem. However, Sega executives internally argued over whether or not to include it. Sega's decisions on the system architecture and the modem were based on a number of factors ranging from industry ties with chip manufacturers to the ratio/cost from a given manufacturer.
"It turned out they had two different projects going on, but they didn't tell us that," said Bing Gordon, now a partner at the venture capital firm Kleiner Perkins Caufield & Byers; from 1998 to 2008, he was chief creative officer at EA. "They told us they were trying to decide if they would include a modem or not. Our advice was, 'You've got serious competition [referring to Sony and Nintendo], so a modem is a really good idea. If you do a modem, we'll build a whole product line for it.'"
In 1996, 3Dfx began building wide acclaim for its powerful graphics chips, one of which ran in arcade machines, including Atari's San Francisco Rush and Wayne Gretzky's 3D Hockey. In 1997, 3Dfx went public, announcing its IPO. In the process it revealed the details of its contract with Sega, required by U.S. law. The announcement, however, had undesired effects. It publicly revealed Sega's blueprint for a new, unannounced console, and angered executives at Sega Japan.
Numerous reports indicate Yamamoto's Blackbelt chipset using the 3Dfx chips was the more powerful of the two. Sega executives, however, still fuming at 3Dfx, severed their contract with the chip maker. (Soon thereafter, 3Dfx sued Sega and both companies settled out of court.)
In the end, Sega of Japan selected Sato's design, codenamed it "Katana," and announced it publicly on September 7, 1997. To this day, it's unclear whether Sega would have chosen the Blackbelt 3Dfx solution, had 3Dfx not revealed Sega's plans publicly.
"They said they looked at 3Dfx, but decided against it," said Gordon. "They went with some other 3D chip that we had never heard of, and they went with a weird processor. We looked at this and asked ourselves, 'Why did they make these choices? It's gotta be some kind of political thing because these are dumb choices.'"
"I felt the US version, the 3Dfx version, should have been used. Japan wanted the Japanese version, and Japan won," said Stolar. "I lost that argument."
Why Electronic Arts Shunned the Dreamcast
With games like Madden NFL, Road Rash, NHL, and a slew of other popular titles, Electronic Arts successfully teamed with Sega to compete against Nintendo in the 16-bit era. The Redwood City publisher supported all of Sega's short-lived systems after Genesis, from the 32X, Sega CD, Game Gear, to the Saturn.
"EA had a deep love affair with the Sega Genesis/Mega Drive, because it was the platform that brought EA into the big time," said Gordon. "EA leaders shared loyalty and affinity with Sega and it over-invested to help Sega make Sega CD and 32X into winners."
But Electronic Arts didn't support Sega's Dreamcast. At EA, there were three groups that had a stake in voting on whether to support a console: developers, sales, and business. They would get together, collect information on any given console, and discuss its virtues and weaknesses.
"We would get together and say, 'Okay, are we going put 10 games together and be there at launch, and put our corporate will behind this, or not?'"
At the time, EA had invested stock in 3Dfx. Did EA's investment in 3Dfx influence its decision? Gordon says it didn't. "If Sega had picked the direct competitor to 3Dfx at the time, it would have been fine. But they picked someone we had never heard of. It was somebody's friend of somebody's friend at a Japanese country club. It was a head-scratcher, like, 'What are they doing?' That was mostly it."
According to Gordon, Sega had flip-flopped over whether to include a modem, but it also picked the wrong chipset. "I remember our CTO (chief technology officer) talking about the processor and going, 'Oh my God, I don't know anybody who has even heard of this chip. It's non-standard and there are no libraries for it.' It was kind of a slap in the face. But even then, at first blush, EA is thinking, 'It's Sega. We've got to support them.' You know, we went to war with them on Genesis and did great things. So the chipset alone was not enough to stop EA from working on it."
Ultimately, Sega's various hardware and business decisions lined up like a row of red flags for EA. According to Gordon, the indecision over the chipset was one thing, the flip-flopping over whether to include a modem was another, but the final straw was Sega's hardball tactics during negotiations over licensing.
"There was a push from Sega, which was having cash flow problems, and they couldn't afford to give us the same kind of license that EA has had over the last five years. So EA basically said, 'You can't succeed without us.' And Sega said, 'Sure we can. We're Sega.'
"And at that point during negotiations -- when someone is trying to call your bluff -- you have to question whether you want to knuckle under or not. And because of the way they had flip-flopped on the configuration, and because the Dreamcast became the system that EA developers least wanted to work on the in the history of systems at EA, that was pretty much it. In the end, it felt like Sega was not acting like a competent hardware company. I was the person who got quoted in the press as saying, 'Dreamcast can't succeed without EA.' The Dreamcast people hated me for that."
Contrary to Gordon's account, however, Stolar said there was a much simpler reason EA denied the Dreamcast. And licensing negotiations, chipsets, and modems, while somewhat relevant, didn't play a significant role.
"Sega didn't play hardball," Stolar countered. "I was the one who led the negotiations between EA and Sega. The royalty licensing deal was done between me and [EA's then-president] Larry Probst, and nobody else. And there was a real significant reason why it didn't happen. That reason was never really public. You ever hear of a company called Visual Concepts?"
The Sports Situation
To launch a console in North America in 1999, console manufacturers had to lead with their own sports games. EA was used to competing with a host of other sports products made by Nintendo, Konami, Namco, Acclaim, and eventually Sony.
To launch the Dreamcast in North America, Stolar had done his due diligence on a company called Visual Concepts. With Sega of Japan's approval, Stolar bought Visual Concepts for $10 million.
"Visual Concepts was a very, very key part of this because they made all the sports titles," said Stolar. "And if you looked at the sports titles on Dreamcast, they far surpassed EA's sports titles. Our football game at the time, NFL 2K, was far superior to Madden. Everybody agreed to that once they came out."
In 1999, during a lunch meeting on neither Sega nor EA turf, Stolar and Probst discussed terms of a potential EA and Sega deal.
"This is what happened, very clearly," explained Stolar. "Larry came to me and said, 'Bernie, we'll support Dreamcast, but this is what I want.' And I thought, 'Great, I know Larry, I know the company real well, I can negotiate this,' I thought he was going to ask for smaller royalties. I would have given him smaller royalties, no question about that."
In the late 1990s, EA was a keen observer of console launches. It knew that striking a deal with hardware companies while they were building their system -- when, according to Gordon, they were worried about costs and putting the company at risk -- was to EA's advantage. "That's when you strike your five-year licensing deal," Gordon said. Probst saw a field of competitors going against EA's growing sports franchises and wanted something more.
"'We want to be the only sports brand on Dreamcast,'" Stolar recalled Probst saying. "'We want the exclusive rights to be the only sports brand on Dreamcast.'"
This surprised Stolar, whose strategic planning included Visual Concepts as a key element in making Dreamcast a success.
Stolar countered. "I said, 'Larry, I'll tell you what. As a third party, I'll agree to that. But I'm not going to agree to that for first party. I bought Visual Concepts for $10 million. So you can compete with Visual Concepts. We'll have Visual Concepts sports titles and we'll have EA sports titles, and that will be it.'"
Probst didn't budge.
"No, I don't even want to compete with Visual Concepts," Probst told Stolar, who replied, "'Forget it then, end of story.' That's what it was all about, right there."
Weeks after Stolar's and Probst's lunch meeting, Japan tried to change EA's mind, but it didn't work, said Stolar. "Sega tried to lower royalties but EA wouldn't budge."
Considering EA's built-in customer base, did Stolar make the right decision?
"Look what Visual Concepts brought to the table, look what they brought to Sega. If you look at those games today, everybody will tell you those games looked better than EA's games. So I would not have changed my opinion."
Stolar may have been right about Visual Concepts' talents as a sports developer, but even before the Dreamcast launched in the US, Sega was found itself in third-party trouble. The largest independent third-party publisher wasn't going to support Sega? When Gordon unequivocally said, "Dreamcast can't succeed without EA," the press jumped on the story.
The message couldn't have been clearer. Sega would struggle on without its long-term software partner and confront Sony and Nintendo with a giant question mark associated with its new system.
Bellfield put the loss into perspective: "The fact that EA didn't support us was a ding against us. But I'm not sure what breakthrough game EA would have given the platform from September 1999 through January 2001 that we hadn't already got in the same genre," said Bellfield. "Sports, the power...