The Final Hours of Prince of Persia

By Geoff Keighley Designed by James Cheung

As the sun streams through the windows of a five-story brick building in Montreal, Canada, Yannis Mallat, 29, briskly walks across the hardwood floors. His destination, it soon becomes apparent, is a 6-foot-high black cabinet up against a wall. As he approaches the cabinet, he turns to a programmer, snaps his fingers, and points to the lock with his index finger.

"Get me the keys," he says with a smirk.

Moments later, a young programmer with a shaved head walks up to the cabinet, rummages through a key chain, and slides the key into the lock. He steps back, and Mallat, sporting a heavy five o'clock shadow, flings the doors open. Inside, 151 CD-ROM cases are neatly lined up in two rows, organized numerically by a red number on the spine of each CD.

"This," Mallat says, as he runs his finger along the top row of CDs, "represents the last two and a half years of our lives. It's every version of the game we have made." He continues to run his finger along the line of CDs until he reaches the last one, build 151, which was completed just hours ago. "We've got a bet going right now," he says while tapping the case of the last disc with his finger. "I'm guessing we will have 153 versions of the game by the time we are done. So two more to go." What isn't said is implied: Prince of Persia: The Sands of Time, one of the most highly anticipated games of the year, is almost finished.

Right now, all that stands in the way are a few dozen bugs. Already, the team has fixed more than 13,000 bugs in the past three months. "It's amazing it has all come down to this," Mallat says, as his eyes scan the large open area that houses his team of 50 young French-Canadians. To the untrained eye, the room looks a bit like a rowdy frat house crossed with a COMP SCI 101 classroom. Look at the team members, and you can't help but notice how young they are: Almost all of them are in their mid-20s, and none of them have a hit game on their resume. "It feels like we've all come so far in such a short period of time," Mallat says with a hint of pride.

That may be understating the fact. In just two years, French publisher Ubisoft has grown from the house that Rayman built into a blockbuster game publisher with hip franchises like Splinter Cell and Prince of Persia. And more importantly, the so-called "POP Team" in Montreal has gone through a radical evolution all its own. Two years ago, the team members behind Prince of Persia were the folks responsible for forgettable kids' games based on the Playmobil toy series and Disney's Donald Duck. But now, almost in the blink of an eye, these young nobodies have become known entities. And their game, Prince of Persia, is a title many have earmarked as one of this year's best.

Of course you don't need to tell this to Patrice Desilets, 29, the game's creative director. As he sits with Mallat and goes over a bug list for the game, his eyes are saucer-wide, and he gives off a thousand-watt smile. It's as if he still can't believe he's about to finish a cool game, not a kids' title that he'd be embarrassed to tell his brother he made. Desilets, a professional improv comedian on weekends, flips through the bug list and discusses the remaining problems with Mallat. While 24 bugs aren't many, everyone knows that the last bugs are always the most elusive. "Yesterday was the last time I said to myself that we should push the game to 2004," Desilets says. You get the distinct impression that he's only half joking.

But everyone knows that missing the release date is no longer an option. In the next few days, the PlayStation 2 version of Prince of Persia has to be finished and sent off in preparation for its early November release. When it's finally done, though, the game will mark something of a milestone for the team: their first hit product. After all, the team is well aware of the downside of the games business. They've all experienced the frustration of standing at a trade show and realizing that no one will even give their game a cursory glance. And worse, they've felt the heartache of spending two years of their life working on a game that is almost instantly forgotten upon release.

This time, however, success is within arm's reach. Everyone on the team knows it too. But they also know that getting to this point hasn't been easy. In fact, the journey they've taken is a fascinating tale filled with twists and turns, surprises and drama. And now, for the first time, the remarkable story behind the creation of Prince of Persia: The Sands of Time is about to be told.

Page 2: One Bitten, Twice Shy

Sitting in the backyard of his Spanish-style home high up in the hills above Los Angeles, Jordan Mechner is wearing shorts and a pale green T-shirt as he gently tosses a worn tennis ball to his dog. As the sun begins to set in the distance, Mechner fondly recalls the early days of Prince of Persia, a game he built all by himself in the late 1980s. "Back then, the biggest cost for developing the game was the rent for my apartment," Mechner says with a chuckle, speaking slowly and methodically as he always does.

When Mechner was younger, he never dreamed of becoming a great game designer. Growing up in New York, he was a smart but shy kid who liked comic books and animation. So when his parents bought him an Apple II computer at age 15, Mechner was eager to create animations on the computer. "At first, I was just doing simple things like creating a little red block that would be chased by a green block," he says.

Over time, Mechner's fascination with computers grew stronger. While at Yale for college, Mechner programmed a karate action game called Karateka. Despite the fact that he was still in school, Mechner sold the game to Broderbund, and it went on to sell an impressive 500,000 units. After graduating, Mechner started to work on Prince of Persia, an action game set in the Arabian Nights universe. The game took four years to create and became a bit of a family effort: Mechner's dad wrote the music, and his brother performed acrobatic moves in front of Mechner's camcorder. (Mechner then digitized the movies into a computer to create amazingly fluid animation for the prince.) Mechner was just 25 when the game came out in 1989, but Prince of Persia went on to sell more than 2 million copies. Across the industry, game designers took note of Mechner and Prince of Persia. "Jordan's early games were cinematic in this very undefined way," explains Will Wright, creator of The Sims. "They felt something like an interactive movie, whereas at the time most other games were simple puzzle toys like Space Invaders."

After Prince of Persia and its sequel, Mechner went from the world of interactive movies to real ones. He attended film school at NYU and subsequently shot Waiting for Dark, a short documentary about the harsh living conditions in Havana, Cuba. The irony of an action video game designer making documentary films is not lost on Mechner. "When I meet other documentary filmmakers," he says with a chuckle, "they ask me what I do. And when I tell them I also design video games set in fantasy worlds with a guy who wields a big sword and kills people, they give me this look like 'How does this guy sleep at night?'"

Even though he enjoyed making documentary films, Mechner still stayed somewhat active in the game industry. But unfortunately, Mechner found it hard to repeat the success of Prince of Persia. The Last Express, a $5 million adventure game that took Mechner and his team four years to build, was a critical smash but a commercial flop in 1997. And in 1999, Mechner worked as a creative consultant on Prince of Persia 3D, a disappointing attempt to blend Prince of Persia with a Tomb Raider style of gameplay. Today, Mechner sounds almost embarrassed to discuss that game, even though he still proudly lists it on his resume. "Can we skip ahead to the good part?" he asks somewhat hesitantly before trying to change the subject.

After the disappointment of Prince 3D, Mechner distanced himself from the game industry. He began working on Chavez Ravine, a documentary film that tells the story of a Mexican-American village that was destroyed to build the Dodger Stadium in Los Angeles. To his friends and admirers, it seemed unclear when--if ever--Mechner would return to making games. "I was in one of my several-year-long 'in-between-game' phases," surmises Mechner. "I needed to charge up my batteries by doing something completely different."

But then, in the midst of working on his new documentary, Mechner received a call. The president of Ubisoft, the company that owned the Prince of Persia name, wanted to discuss the idea of doing a new game in the franchise. With wounds still healing from Prince of Persia 3D, Mechner wasn't sure he wanted to do another video game, much less attend a meeting in France. "The first thing on my mind was, 'Well, it would be great to do another good Prince of Persia game,'" remembers Mechner. "But I was also thinking, 'It wouldn't be great to do another mediocre one.'"

Nevertheless, in March of 2001 Mechner flew to Paris to meet Ubisoft and hear what it had to say. The fate of Prince of Persia, one of video gaming's most classic franchises, hung in the balance.

As soon as Mechner arrived in France, he headed to the Ubisoft offices in Paris and met with Yves Guillemot, the president and CEO of the company. Through a series of complicated business transactions, Ubisoft had acquired the Prince of Persia name from Broderbund, a company that imploded in the late 1990s. The catch was that while Ubisoft owned the name, Mechner held the intellectual property rights to the franchise. In other words, the only way Ubisoft could make another Prince of Persia game was by convincing Mechner to cooperate.

During the meeting in Paris, Mechner says he was impressed by how seriously everyone treated the Prince of Persia franchise. "They had this big PowerPoint presentation about the future of the franchise," he recalls. But fancy slide shows aside, Mechner was still skittish about approving a new Prince game. In order to even consider the possibility, Mechner told the Ubisoft executives that he wanted to personally meet the team that would produce the product. "I needed to judge if the team was going to be able to do a good game," Mechner recalls. The suits at Ubisoft agreed: The goal was to assemble a team in Montreal and have the producer meet Mechner at the Electronic Entertainment Expo in Los Angeles.

In Montreal, the task of pulling together a team for the Prince of Persia game fell on the shoulders of Yannis Mallat, a producer who had games like Rayman Advance and Dinosaur (based on the Disney movie) under his belt. Mallat, like Mechner, had a streak of public-interest work in his past: After going to college in France, he worked in West African countries to help rural citizens increase the productivity of their crops. Eventually, however, Mallat realized he wanted to do something different with his life. After obtaining an MBA in Quebec, he took a job at Ubisoft.

Mallat says he was certainly excited by the possibility of working on a new Prince of Persia game. But in early 2001 that's all the game was: a possibility. Mallat knew that in order to get the game green-lit by Mechner, he'd need to come up with a great concept and assemble a fantastic team. So he set to work on building a team and coming up with a game design. Leading up to E3 in May, Mallat brought three designers on to the project who began working on a design (or "pre-conception") for the game.

A few weeks later, Mallat flew to E3 to meet Mechner. The meeting lasted two and a half hours, and by all accounts the two men hit it off. By the end of the meeting, it seemed like Mechner was beginning to warm up to the idea of a new Prince of Persia game. "But I also kept reminding myself, 'Hey, Prince of Persia 3D sounded really good at first too,'" recalls Mechner. In other words, Mechner still needed more convincing. So at the end of the meeting, it was decided that Mechner would fly to Montreal to meet the team and see the initial design work.

There was, however, one big problem. Back in Montreal after E3, Mallat started to realize that the pre-conception work on the project was not going as expected. "The designers on the project were coming up with a very basic game--you know, you run around in 3D, there are traps, and it was all very obvious," Mallat says. With Mechner arriving in Montreal in only a few short weeks to approve or kill the project, Mallat knew that time was running out. Before long, he realized that drastic measures were needed to save the fate of the game. In short, the entire Prince of Persia team would have to be fired.

Page 4: The Princess of Afghanistan

With the clock ticking and Mechner's visit only weeks away, the decision to fire the team was certainly risky. Mallat says he had no choice. "We needed designers who could think about the animation and the style of the gameplay. We needed inn-o-va-tion," he says, emphasizing each syllable of the last word. According to Mallat, the first team on the project simply wasn't delivering any new ideas. "It was the old school of game design where the designer uses only one tool: Microsoft Word," Mallat says, his voice quickening as he grows more agitated. "Typing, typing, typing--that's all they do. You know, 'The Bible' as they call it. It was just all words on a page. I didn't want a 300-page design document--I wanted something more alive." Mallat says he hemmed and hawed over his decision for a while but eventually realized the harsh reality. "We had to completely clean house and start from scratch with a new team," he says.

Building up a new game team in only a matter of weeks would not be easy. But since Ubisoft Montreal has more than 650 employees in development--the second largest studio in the world, right behind Electronic Arts in Canada--Mallat didn't have to go outside the company to find new team members. Still, those inside the studio were beginning to wonder if the Prince of Persia project was destined to fail. "People from other projects would walk by the team and say, 'Hey, that's Prince of Persia--it's going nowhere, it's a dead franchise,'" recalls Phillipe Morin, the game's camera designer, who is also the oldest member of the team at 31.

As he assembled a new pre-conception team of six members, Mallat began telling everyone that what mattered most was the gameplay. "I didn't want people thinking about the license at all," he recalls. "I said it didn't matter if we are making Prince of Persia or Princess of Afghanistan--we just needed to come up with an interesting game design that would be fun and relevant to the gamer in 2003."

That interesting design soon came to the surface. "The real step forward was when we came up with the idea of a Persian ninja," he explains. "Basically it came down to the idea of a very acrobatic gameplay environment." So instead of having a prince who could only jump over traps, the pre-conception team imagined a prince who would perform moves right out of a movie like Crouching Tiger, Hidden Dragon. Perhaps the most exciting idea was what the team called "vertical walling," the ability for the prince to run up the sides of walls, just like in The Matrix.

With Mechner's trip to Montreal quickly approaching, the team knew there was no time to create a playable demo. Instead, the core team decided to create a series of animation tests to show Mechner. Animator Alex Drouin--who himself is skilled in kung fu--worked on a series of simple tests that showed a silhouetted prince character running around a blocky flat-shaded environment. The tests weren't flashy, but they showed off vertical walling and gameplay ideas, like the prince being able to swing on bars like a gymnast.

When Mechner arrived in Montreal he didn't know what to expect. But he knew what he didn't want to see. "I absolutely did not want to hear about making another Tomb Raider clone," he recalls. Luckily for the team, the animation tests looked nothing like Tomb Raider. But still, no one was sure how Mechner would react. "A year later, Alex told me, 'You know, we thought you were going to come in and say, "No, no. This is all wrong. We're not doing a game like this,"'" says Mechner.

Not quite. Mechner was absolutely blown away by the tests. "The masterstroke was taking my running and jumping animation from the first game and expanding that vertically so you could run up walls," he says. When the reel of animation tests finished playing, Mechner turned to the team, looked them all in the eyes, and said one thing: "Guys, what I've just seen has reawakened the joy of making video games to me."

Not long after returning from Montreal, Mechner struck a deal with Ubisoft for the new Prince of Persia game. (At first Mechner planned to be involved only on the periphery as a creative consultant.) Word quickly filtered back to Montreal that Mechner had given the thumbs-up. The team was understandably thrilled. But at the same time, everyone realized that the hard work was just beginning. "It's one thing to say you are going to have this amazing acrobatic game--you know, the prince will jump over enemies, run on the walls, swing on ropes, and things like that," says Mallat. "But now we had to translate that into a real working game."

In September of 2001, real design work on the new Prince of Persia game began. While Mallat acted as a producer, most of the day-to-day work fell into the lap of creative director Patrice Desilets and the rest of the pre-conception team. Unlike the previous design team, Desilets--who moonlights as an improv comedian--wasn't interested in writing a big 300-page design document. "In a way, improv inspires the way I design a game," he says. "In improv you have to say yes a lot to what people give you because if you say no, the improv is over. Game design is like that for me. I try to say yes to everyone's ideas and build off of them. I would never sit here and say, 'Well, that's a good idea, but the design document I wrote back in February says this, so we can't change it.'" All told, the design document for Prince of Persia never grew to more than 10 pages of paper.

The way Desilets tells it, the core game design for Prince of Persia was largely inspired by sports games. "I love how games like NHL have great context-sensitive control. So I wanted to blend the mechanics of a sports game with a story-driven adventure," he explains. "In the early days, I thought of this as an action adventure game like Tony Hawk, only with a prince. I wanted a character you could throw around the environment, and he would do acrobatic tricks. Our prince was going to be about a set of possibilities rather than specific abilities, so as you played the game, you were going to surprise yourself with what the character could do in certain environments."

For months, the team brainstormed different gameplay ideas. For a while, ladders were supposed to play a large part in the game, with a design that would allow players to kick and swirl them around like in a Jackie Chan movie. A magic carpet was another concept, as was a sequence where the player would ride on top of a horse and fight enemies in the desert. "We had a lot of great ideas, but we also wanted to focus the game design," says Mallat. "We didn't want a ton of average stuff--we wanted a focused amount of great stuff." After a few months of initial design work, it was decided the game would largely be composed of three core elements: fighting sequences, acrobatic puzzle sequences, and a few adventure-game-like puzzles. Horse riding, magic carpets, and swirling ladders would have to wait for another game.

Still, Mallat wanted to make sure that this Prince of Persia game had something that made it stand out. "We needed our own technological breakthrough or a gameplay keystone," Mallat says. "Much like people talked about Jordan's animation system for the original Prince of Persia, we needed to raise the bar in some way with this game."

It turned out Desilets had the perfect idea--and he owed it all to Donald Duck.

Page 6: A Bittersweet Moment

Yes, it's true: One of the most original features in Prince of Persia: The Sands of Time has its roots in a game based on Donald Duck. When Mallat started prodding the team for ideas of a "gameplay keystone," Desilets brought up an idea he had proposed to the producer on his last Donald Duck game. In a nutshell, he wanted to let players rewind the game after they died. "I remember playing the Donald Duck game, and when I died, I wanted to just go back in time a few seconds instead of having to replay the whole level," says Desilets.

The producer on the Donald Duck game dismissed the idea, but Mallat seemed more receptive. "I think we all began to realize that by rewinding the game, the player would have a never-ending checkpoint," says Mallat. "We also knew a feature like that could become really important in an acrobatic game where you are always making jumps and trying to avoid traps." But while the idea sounded great, no one knew if it was technologically possible to let a player rewind the game on a PlayStation 2.

For the answer to that question, Mallat turned to Claude Langlais, the lead programmer. In a way, it was the perfect time to be exploring such questions: The programmers were in the midst of deciding what technology would serve as the foundation for the game engine. For months the team was gravitating toward using the Unreal game engine, but ultimately the programmers and Mallat decided on using an internal Ubisoft technology called JADE, developed in France for the game Beyond Good & Evil. Best of all, Langlais thought he could get the rewind feature working with the JADE technology.

After deciding on the game engine, work began on the first prototype (or playable version) of the game at the very end of 2001. At the time, there were still less than a dozen people on the Prince of Persia team. "By keeping the team small at the start, everyone develops a really strong working relationship," suggests Mallat. The team would eventually grow to more than 50 employees, but Mallat was intent on not starting full production on the game until the core team was happy with the prototype.

It didn't take long for the first prototype to emerge. By early March, the core team had something up and running on the PlayStation 2. The graphics were ugly, but the team managed to get the rewind feature working. Best of all, there were already signs that the acrobatic gameplay was going to work. Animator Alex Drouin and AI programmer Richard Dumas had built a prince character that was incredibly responsive to player controls, just like in a sports game. (In fact, the prince character would ultimately end up having more than 700 possible animations.) In terms of the levels for the prototype, lead level designer David Chateauneuf--who is only 24 years old--was able to build test levels that highlighted the acrobatic gameplay possibilities. "It was almost like a big acrobatic playground," he recalls.

With the first prototype up and running, Mallat knew it was time to give the executives at Ubisoft a preview of what his team had accomplished. So in a Montreal conference room in March 2002, Mallat loaded up the first prototype version. (For a never-before-seen look at the first prototype, watch the video clips on this page.) The prototype was rough, but it showed an incredible amount of promise: The game ran smoothly, it was easy to control, and the acrobatic gameplay seemed fresh and innovative. Best of all, the rewind feature was already working flawlessly. By the end of the demo, the executives said they were floored. Some called it the best first prototype they had ever seen. "This is so good you need to go into full-scale production today. Bring in 50 people immediately," one of them told Mallat.

While Mallat was pleased the executives had liked the demo, he was dead-set against ramping up production so early. "We needed to finish the art style and find an art director. We needed to finalize the fight system. There were so many things we still had to do," he says. In other words, "The game simply wasn't ready for production--we needed at least three more months with the core team."

Nevertheless, the executives kept pushing Mallat to scale up the team and move out of the preproduction phase. By the end of the meeting, what was supposed to be a moment of sheer euphoria turned into a bittersweet experience. "I remember leaving the room and being furious, absolutely furious, with the executives," explains Mallat.

In a way, the Prince of Persia team had been too successful too fast--and now it looked like their success was about to come back to haunt them.

Page 7: Send in the Sand

Mallat was upset with the outcome of the meeting with the executives. But he didn't give in easily. After many heated discussions and debate about the future of the project, Mallat finally convinced the executive team that three more months of preproduction would ultimately help the game. So while the team would continue to produce new prototype versions in the interim, the build-up of the team would be delayed until later in the summer of 2002.

In the intervening months, Mallat and the core team worked to iron out some of the remaining issues. First on the priority list was hiring an art director to create a visual style for the game. "We probably should have brought on an art director earlier," Mallat admits. After much searching the team found Raphael Lacoste. He immediately began to work on the game's overall visual look and the design of the prince character. Dozens of designs were made for the prince--everything from a young boy in a turban to edgier looks with a buff and shirtless hero. Over time, the team agreed that they didn't want to make a cartoonlike Aladdin world for the prince. "Everyone just kept saying, 'More edgy, more mature,'" says Lacoste.

The idea of creating an edgier game extended to the look of the game environment. "We started with bright and soft cartoonlike worlds," says Lacoste. But over time, the game's art style took on a darker and more textured look, which was inspired by the work of French painters Jean-Leon Gerome and David Roberts. "We didn't want the game to look like a rigid polygon game, but rather it should feel like a painting you can play," says Lacoste, as he rubs his index finger and thumb together. Lacoste would also begin to push for more visual effects in the world, including fog, streams of light, and cloth that could sway in the wind. Slowly, the visual quality of the prototype improved.

The last element of the game that had to be nailed down was the fight system. The team had perfected the prince's acrobatic moves, but creating a similarly smooth fight system--with the prince using a scimitar and dagger--proved difficult. "We started saying things like, 'Well, if the prince can run up walls, why can't he run up enemies too during a fight?'" says Richard Dumas, an AI programmer on the core team. Ultimately, the team says the fight system was inspired by documentaries on capoeira--a Brazilian fighting style that melds martial arts with stylized dance.

As the various elements of the prototype came together, Mechner visited the team in Montreal for a week to check on the game's progress. He was immediately dazzled by the prototype version, especially the rewind feature. "It became a bit of a process of seduction," Mechner recalls. "I found myself interested in getting even more deeply involved with the game."

After spending an evening at a hockey game with Mallat and Desilets, Mechner proposed that he could write a unique story tailored to this Prince of Persia game--a story that would integrate the rewind feature into the narrative. By the end of the week, the team and Mechner had decided this Prince of Persia game would be a prequel to the previous titles and feature a new female character, Farah, who would aid the prince. To integrate the rewind feature into the story, the team came up with the concept of the "Sands of Time," a mystical power the prince unknowingly unleashes on his kingdom. The power would allow him to manipulate time, but it would also turn the inhabitants of medieval Persia into demonic creatures.

By July 2002 all the pieces were finally falling into place. Prince of Persia: The Sands of Time was ready for full-scale production. Almost overnight the team would quadruple to more than 50 employees.

Page 8: From Vision to Reality

Walk into most game developer studios, and you see rows of cubicles and offices. But at Ubisoft in Montreal, the Prince of Persia team works in one huge war room with not one office or partition. Even Mallat, the producer, sits at a desk at the side of the room, flanked by a large white board that displays the current production schedule. When team-wide meetings need to be held, Mallat simply stands in the middle of the room and talks to everyone. There's not much privacy, but the members of the Prince of Persia team say the environment helps spur collaboration. "In other companies you build a level in isolation and only really talk to the producer," says David Chateauneuf, the lead level designer. "Here, you can just call someone over to your monitor and say, 'Hey, what do you think of this?' If you are doing something cool, you might attract 10 or 15 people who will give you feedback."

With the prince's moves and abilities locked down in the prototype stage, the main focus of production was building the actual levels--the art, the design, and the puzzles. At the start of production a large bulletin board was put up in the office to give an overview of the entire game. Each level was represented by a colored index card that could be moved around on the board as the level order changed, much like a TV executive uses index cards on a board to juggle around his prime-time schedule.

Since Prince of Persia: The Sands of Time was conceived as a highly acrobatic game, the level design needed to be more than a bunch of large nondescript rooms. Rather, the levels would become physical puzzles that required the prince to avoid traps, scale walls, and swing from poles to progress. In order to build such complicated levels, the team went through a series of what were called blueprint sessions. "Those designs were literally done from 30,000 feet in the air," explains Desilets. "We'd look at the top-down view of a level and say, 'OK, over here there will be a fight, and then over here there will be some acrobatic gameplay with columns.'" When the blueprint sessions were done, the team had more than 250 rooms created in rough form. Once the team was happy with the blueprint, the level then began production: Artists would create textures and in-game objects while the level designers began the detailed work on creating the actual gameplay.

As in any 3D video game, the in-game camera became an important design issue. For the first half of the project the team had the camera stay in a fixed position like in Devil May Cry or ICO. But by the end of 2002, everyone began to realize that the camera should move around with the player. "At some point we realized we could move the camera and bring the player into the experience even more," says Phillipe Morin, the game's camera designer, who used to work as a second assistant director on motion pictures.

The camera change was a minor switch, but at the start of 2003 the team would face a much greater hurdle. In early January, the programmers realized that the JADE graphics engine was no longer capable of supporting the game's huge levels and detailed visual effects. "The programming team came in and basically said, 'Yannis, we have some bad news. We need to create a totally new rendering system for the game,'" says Mallat.

While changing the graphics engine wouldn't influence the level design or the core gameplay, Mallat says it was scary to make such a huge change to the game only three months before its first public unveiling at the Electronic Entertainment Expo. "Here we were only months before our debut to the public, and we no longer had a graphics engine. I definitely was getting worried."

As the deadlines began to approach for creating an E3 demo, the JADE rendering engine was removed from the game. It was replaced by a 3D rendering engine created by members of Ubisoft's Shanghai, China, design studio. Almost immediately the team began seeing results. The game was running faster, and, most importantly, the visuals were improving, thanks to effects like in-game fog and streams of light in the environment.

Still, as the team prepared the E3 demo--which would feature a polished selection of a few game levels--no one knew how the game was going to be received. Mechner, a veteran of many E3s, says he told the team that they were in good shape. "I kept saying 'Your work is top quality,'" he says. "But I don't think anyone on the team emotionally believed what I was telling them."

In a way, the team pessimism was understandable. Given the young age of the team members and their history of projects, no one was expecting to have hordes of people playing their demo at E3. The members of the Montreal team were used to practically being ignored at the show. "The only guys who came to our booth before," explains animator Drouin, "were the people looking for somewhere to rest, because there was a place to sit down near the Donald Duck game."

But as soon as the team arrived in Los Angeles for E3, everyone began to realize that things might be different this year. "I went to E3 expecting we were going to have this little stand in the back," remembers Richard Dumas. "Then I walked into the booth and saw we had a whole section for Prince of Persia." Even better, there were Prince of Persia booth babes. Things were definitely looking up.

Once the show opened, the news only got better. The team, dressed in red Ubisoft shirts, eagerly waited at the booth to see early reactions. The hours leading up to E3 had been long, and team members were looking for some kind of validation. They got it in spades. By the end of the first day, the Prince of Persia section of the booth was filled up with dozens of onlookers who were elbowing each other for a chance to play the game. "I remember turning to Patrice at one point and saying, 'Is it always like this at E3?'" remembers Dumas. "And he just said to me, 'No way man. It has never been like this before. It's so awesome!'"

Equally awesome was the praise lavished on the game at the end of the show. As E3 closed, both online Web sites and print magazines were whispering that Prince of Persia was perhaps one of the most impressive games at the show, mostly because of its lush graphics, unique gameplay, and excellent control. A few weeks after the show, the buzz was replaced by awards. Prince of Persia: The Sands of Time was named The Best Action/Adventure Game of the Year by the E3 Game Critics Awards. In 2002, the award was given to Splinter Cell, another Ubisoft title that went on to become one of 2002's best-selling games.

After returning to Montreal, Mallat gave the entire team the week off after E3 to celebrate their success. But upon returning from vacation the mood suddenly became more somber. "We came back, and I told the assembled team, 'OK, we have a good demo...and that's all we have right now,'" he says. Mallat didn't want to be the bearer of bad news, but the simple fact was that Ubisoft was expecting the game to be on store shelves by November, only six months away. No one on the team thought that release date was realistic. "I basically realized after E3 there was no way we could finish the game this year, and we'd have to slip it to 2004," explains Mechner.

But that, it turns out, was not going to be an option. All of the sudden the success of E3 would be replaced by what Mechner calls "the darkest hours" on the project. "It was a make-or-break moment. The whole game could have easily come crashing down."

Page 10: The Darkest Hours

Mechner, whose involvement in the project continued to grow over time, decided to temporarily relocate his family to Montreal after E3. Arriving at Ubisoft in late May, Mechner says he expected the first meeting to be about setting a new release date for the game in 2004.

After all, while the E3 demo was impressive, everyone on the team knew that the rest of the game was nowhere near complete. With actual production only in its ninth month, the ambitious scope of the game meant that at the current rate, level production wouldn't be complete until sometime early in 2004. "Can I be honest?" Desilets asks. "I didn't see why we couldn't just delay the game. Everyone knew who we were after E3, so why couldn't we just take some extra time and come out next year?"

But as these things often go, the businessmen at Ubisoft had a different idea. "Word came down that the game was going to ship this year come hell or high water," says Mechner. While Mallat agrees the game appeared to be behind schedule after E3, he also thinks a lot of the younger team members overreacted. "OK, these guys had a lot of fears," he says. "I think it's because they felt so close to creating a masterpiece. It was going to be their first taste of success, so they really didn't want to mess it up right at the end."

Still, Mallat says that in order to try for a holiday 2003 release, the game's scope had to be cut back by about 15 percent. "Every morning we had to come in and say, 'OK, the time for dreaming is past. The time for basking in the glory of the E3 reaction is over,'" recalls Mechner. "I always knew that the game would have to be cut back at some point. It definitely was too large. But of course it's never easy to figure out what gets cut out."

At first, the team came up with the idea of slicing the game into two halves, much like Quentin Tarantino split Kill Bill into two separate films. "The first half was going to come out this fall and then the second half next year," explains Mechner, who wanted to ensure the cuts didn't destroy the integrity of the story he had labored on for months. Eventually, though, the idea of splitting the game was dismissed. Instead, the team decided to go through and cut out levels that were either "very early in production or not up to a high level of quality," says Mallat.

Still, some of the cutbacks were hard to make. Perhaps the most difficult decision of all was scrapping a series of boss fights with a creature named griffin, a large sand-bird nemesis. "You were going to have these three encounters with him throughout the game," reveals Desilets. "The first two times you wouldn't have a good enough sword to beat him, but at the end you finally get him." But with a holiday deadline looming, the team had to make the difficult choice to pull the griffin battles out of the game.

Even with the cuts, "it became a bit of a death march in the summer," admits Mechner. But there was some good news: The intense preproduction process on the game helped avoid any further delays or cuts. "I'm really proud that the first map we did for the game is still in the final version," explains Richard Dumas. "We had to refine it three to four times, but we didn't have to redo or scrap a lot of content on this game."

Of course in the software industry nothing is guaranteed, so even though the production rolled out through the summer, there was still the distinct possibility that Prince of Persia might not make its release date. But by the end of the summer the game seemed to be coming together. Despite the cuts, early play-test reports suggested the game still took at least 17 hours to finish.

It's a warm fall day in Montreal, and the team behind Prince of Persia knows that the game is within inches of being done. The past few months have been filled with fixing thousands of small bugs, but now the project is in its final stage of development. As Yannis Mallat walks through the fifth-floor office, he looks over to the "POP4" sign on the wall, which is made out of Penguin mint boxes that have been taped together. There's a hint of tension in the air, but most of the team now realizes that the cuts they made in the summer have helped ensure that the game will come out for the holidays. "We were always very realistic about what we could and couldn't do," Mallat suggests. "The project was ultimately better off because we didn't try to cram a bunch of average stuff into the game. Everything that's in there now is triple-A quality."

But at the moment, what Mallat is about to put on the TV monitor isn't triple-A quality. Rather, he's pulled the first prototype disc from the black cabinet and taken it over to his PS2. Sitting on a couch, Mallat loads up the prototype from March 2002 and patiently waits for the first level to appear onscreen. As soon as it does, he gently smiles and shakes his head. "Wow, this really puts it all into perspective," he says. "We really have come so far in only 18 months." As he sits there playing the prototype, he calls out to various employees in French. Some of them break out laughing as soon as they see the first version on the screen. "It's like looking at the first painting you did in kindergarten or something," one passing employee says.

The team may have come a long way in 18 months, but right now 24 bugs are left in the game code. Three of them are serious A bugs, or "crash bugs"--errors in the code that cause the game to crash during play. (Already the team has eliminated all the Z bugs, which are bugs that prevent the player from successfully finishing the game.) And while there are 24 bugs left, most of them are very obscure. One, for instance, involves the disappearance of the head of the female character, Farah, if you leave the game on for more than 12 consecutive hours.

With the bug list shrinking as the hours tick by, Mallat knows the end of the project is near. As evidence of this fact, 24 bottles of champagne have been placed on a table in the middle of the war room. The message Mallat wants to send to everyone on the team is clear: We're almost there--get ready to put the champagne on ice.

A few hours later, the team gets to the point where the last bug appears to be fixed. Not surprisingly, the final bug is another obscure one: If a player goes back and forth between two levels without killing any enemies, Farah, the computer-controlled sidekick, will forget what she has to do in a level. It takes only a few minutes to fix the bug. After that, Mallat tells his programmers to burn what could be the final version of the game. The team thinks it has fixed the last bug, but the final verdict rests with Sony, which will test the game one last time before it is cleared to be released on the PS2.

Sony's approval will take a few days no matter what, so in the intervening days Mallat works hard with the team to finish up the Xbox, GameCube, and PC versions, which will all arrive slightly after the PS2 release. At the same time, Mechner--who returned to Los Angeles two weeks before the final hours--patiently waits at his house for word on whether the game has "gone gold," industry slang for finishing the product. "Your fate is in the hands of the console company, so you can't really do anything. You just wonder: Will they approve it? Will they find any bugs? There are a lot of unknowns," he says.

But on Friday, October 15, what was unknown becomes known: Word comes back from Sony Europe that Prince of Persia for the PS2 has been approved for duplication. (A few days later, the US version is approved.) When word reaches the team in Montreal, the champagne bottles are popped open, and the team celebrates its accomplishment late into the night. For some, it's unbelievable that the game is actually done. "This is the first game I've worked on where I'm really proud of what we've accomplished," admits Richard Dumas.

After the celebration ends late at night, Mallat returns home and goes to bed at 3:00am. So relieved that the game is finished, Mallat sleeps through the day Saturday and doesn't wake up until early Sunday morning. And when he does wake up, he realizes that at a duplication plant somewhere in Europe, hundreds of thousands of discs of his game are being made. In a matter of a few short weeks, The Sands of Time will be on store shelves all over the world.

Page 12: A Long Way From Donald Duck

It's a few days after the game has gone into duplication, and no one on the team can believe they are done. "It's been a hard couple of days," Desilets admits. "That game was my entire life for 26 months--every day, every single minute. And now suddenly it's gone. I honestly feel a bit of a hole because I don't know what to do with my life."

At least in the short term, most of the Prince of Persia team will go on vacation. Desilets isn't sure where he's going--like the unpredictable improv comedian he is, his plans have changed half a dozen times in the past few months. First there was a trip to Thailand. Then there was the idea of going to Vietnam for two months. Now, he's thinking two weeks in Cuba and then maybe going to Australia for a month. ("The women there are absolutely amazing," he says, showing signs of a young man who's finally ready to think about something other than his game.)

But even though Desilets and most of the POP Team will go on vacation in the short term, Desilets is already eager to get working on another game. "I'd love to do a game where you play it in real time for 24 hours, or maybe a horror game where the entire game is you battling against just one enemy," he says. Already, ideas are flowing about the future. Somehow, though, it's probably a safe bet that the days of Donald Duck games are over.

When you hear Desilets talk about the future, you can't help but ask him if finishing Prince of Persia has, at least in a way, fulfilled the dream every teenage boy has to one day build his own blockbuster video game. Ask Desilets this question, and there's a long drawn-out pause. He runs his fingers through his hair, and his eyes drift to the ceiling.

"No, no," he says softly. "You didn't dream about that when I was growing up in Quebec. When I was a teenager, video games were done elsewhere--Japan and America. I loved my country and I didn't want to move anywhere else, so I basically thought I would never be able to make a video game." But sometimes, it seems, the passage of time changes both circumstance and opportunity. For Desilets and the other young French-Canadians who built Sands of Time, what seemed like an impossibility 10 years ago has now become a reality.

As the sun finally sets on Mechner's house in Los Angeles, the wind picks up and he begins to cross his arms to keep his body warm from the breeze. When told what his fellow teammates in Montreal had to say about the end of the project, Mechner begins to nod his head in agreement. "This project reminded me what it can be like to work with a great team," he says. And while Mechner says it's unlikely he will get deeply involved in another Prince of Persia game anytime soon, Sands of Time has rekindled his interest in making video games after a string of crushing disappointments. "This was so much fun I'm already working on a brand-new video game project," he hints.

In the end, Prince of Persia: The Sands of Time is a game that will surely be lauded for its rich and addictive gameplay. But on a deeper level, what stands behind the polygons, beyond the levels, and past the dazzling visual effects is a group of 50 young men and women forever changed by the experience of building the game. For the team in Montreal, Prince of Persia was a project that showed each of them their true potential. No matter what the team does next, they will never have to worry again about being ignored at the next E3. And when it comes to Jordan Mechner, creating Sands of Time changed him too: It showed him that no matter what may have happened in the past, the passage of time brings with it new opportunities and unlocks new possibilities. Prince of Persia is back, and so is Jordan Mechner--and this time, hopefully for good.