Tribes 1 Legendz

At E3 in Atlanta in the spring of 1998, a newly announced game named Starsiege: Tribes took aim at the whole history of first-person shooters, a burgeoning genre whose hallmarks had crystallized quickly.

In front of a large convention crowd, lead designer Scott Youngblood led a live demo that introduced the PC title to the public. “It started off running around an interior that looked just as good as Quake,” he recalls. “And then go out to the exterior and they see this huge terrain, and then just jump off and then jet-pack to the ground. Like, ‘Holy crap.’ They’ve never seen anything like that.”

“It blew people’s minds,” remembers lead software engineer Mark Frohnmayer. “He was running around like, ‘Oh, yeah, first-person shooter,’ and then all of a sudden rounds the corner and there is this huge terrain in the distance and he runs out there with a jet pack on. Like, ‘What the fuck.’”

In the era of Destiny, Battlefield, Far Cry, and Fortnite, it’s difficult to grasp how limitless Starsiege: Tribes would have looked to the E3 audience only a year after Quake — Id Software’s gloomy, Gothic, muddy-looking successor to Doom — became widely available on Windows. “Now it would be totally normal,” Frohnmayer says. “But back then it was the first time that you really had a high-fidelity interior environment and a high-fidelity terrain environment, and they blended seamlessly.” In a preview posted after E3, Gamespot raved that Tribes “renders scenes with sharp backdrops for miles” and, with appropriate awe, reported, “Instead of waiting … to load a massive outdoor map, the game just moves along without pause.”

Ask developers and players alike to reminisce about Starsiege: Tribes, which turned 20 on November 30, and those three words — “the first time” — form a frequent refrain. Much of Starsiege: Tribes seemed so far in advance of its own time that two decades later, it’s still similar in concept to contemporary titles.

Tribes didn’t just anticipate later trends. It also played an important part in popularizing them. By earning critical kudos, cultivating an active online community that persists today, and demonstrating that the previously undone was doable, Tribes influenced future developers even without being a massive seller. Before such features were standard fare for first-person shooters, it boasted a sophisticated character class system; an emphasis on teamwork, coordination, and team objectives rather than gleeful fragging alone; player-controlled single-person and team-oriented vehicles; a major publisher green-lighting a multiplayer-only release; and ingenious network code that supported an unprecedented number of players amid massive indoor-outdoor environments, in an era when most players were still stuck with dial-up connections. And, maybe most memorably, it offered a degree of verticality and freedom of movement that’s rarely been replicated.

“For us, it was a wide-open thing,” Youngblood says. “It was like, ‘Well, there’s lots of potential opportunities, but since no one’s done it before, what do you do that’s going to be fun?’”

In retrospect, the solution seems simple: When in doubt, do jet packs.

FrohnmayerFrohnmayer remembers a moment in the mid-1990s when the cofounder and president of Tribes developer Dynamix, Jeff Tunnell, addressed him and two Dynamix directors, Tim Gift and Rick Overman. “We need to do a Doom-killer,” Tunnell told them. “Dynamix needs to throw down with something that will beat Doom.”

The formative first-person shooter from Id was a huge hit and a cultural juggernaut, so wanting to top it — or at least grab a slice of the blossoming FPS market — was a worthwhile, if ambitious, goal. Eugene, Oregon–based developers Tunnell and Damon Slye had teamed up to start the generically named Software Entertainment Company in 1983, changing the name to the more distinctive Dynamix the next year. The company, which was acquired by King’s Quest creator and publisher Sierra in 1990, was well known in the ’90s for its adventure games, but it also had a genre-hopping history of flight simulators, tank shooters, and mech games, which came in handy when the studio decided to get serious about entering the FPS arena. Tunnell and Slye’s first game together, 1983’s Stellar 7, was a futuristic tank sim known for its difficulty, as Tribes would one day be. Author Tom Clancy, who in 1988 named the former one of his two favorite games, said of Stellar 7, “It is so unforgiving, it is just like life.”

Youngblood and Frohnmayer had been playing competitive Quake together, and as much as they admired its fast-paced action, it motivated them to break the deathmatch mold. “I was trying to do something more than the other shooters that were out there,” says Youngblood, who was working on a line of sports games at Dynamix until he joined the Tribes team when the incipient project was still codenamed “Fear.” “I looked at Quake and they had these level-based experiences where you could run around and shoot guys in the face, but there wasn’t a whole lot to do other than that.” Early FPS games were often called “corridor shooters” because limited processing power prompted developers to keep players confined to narrow, resource-saving pathways.

At E3 in 1997, the two Dynamix developers took part in “Red Annihilation,” a Microsoft-sponsored Quake tournament in which close to 2,000 players competed to win Id cofounder and Doom and Quake creator John Carmack’s 1987 Ferrari. The winner was Dennis “Thresh” Fong, generally recognized as the first pro gamer and the popularizer of the “WASD” keyboard-control scheme. Thresh dominated the tournament in part through his exploitation of a mechanic called “rocket jumping,” which players could use to propel themselves to far greater heights by jumping, firing a rocket downward, and being boosted by the blast. (Don’t try this in real life.) According to Id cofounder John Romero, rocket jumping was not an intended feature of Quake, although the company kept it in the game after programmer John Cash discovered it accidentally. It’s since become a staple of many other shooters.

Watching Thresh explode himself to success made a major impression on Youngblood. “That, to me, offered the players a whole other avenue of choice in terms of, ‘I can go left, right, forward, backward,’” Youngblood says. “But if I can go up, wow, that changes things a lot.” Dynamix decided that jet packs would become a foundational feature of Tribes, pushing the shooter genre further away from the Flatland where its early titles had largely lived. Not only would the jet packs diversify the game’s navigation and allow for greater size and variety in its multiplayer maps, but they would also make combat more compelling by adding unpredictability to opposing players’ movements.

Although jet packs, unlike rocket jumps in Quake, were a fully planned part of Tribes, the Tribes development process included its own serendipitous discovery of a physics-based bug that altered the game in a good way. About halfway through development, software engineer Dave Moore, who was working on collisions and character movement, walked into Youngblood’s office. “[He] said, ‘Hey, I’ve been working on this stuff and I have a bug, but I’m not sure if I want to fix it or not,’” Youngblood recalls.

Moore wouldn’t divulge what the bug was. Instead, he brought Youngblood back to his office and told him how to trigger it. “It was, ‘OK, you’re on top of this hill. Run toward the next hill, and when you get to the top, jump and then just keep jumping as you go all the way down,’” Youngblood says. “And I played with it and was like, ‘Holy crap.’” Pressing the jump button canceled the collision with the landscape, allowing the player to accelerate down slopes and frictionlessly launch into the air, attaining much higher speeds than were possible with the jet pack alone. “[It] was ‘wrong,’ but it actually made the game more fun to play,” Youngblood says. “I looked at him and was like, ‘Nope, don’t fix it. This is now a feature.’”
Sierra

The mesmerizing mechanic — which wasn’t any less realistic than rocket jumping — was dubbed “skiing,” and it became a core part of competitive Tribes. Youngblood, who joined the Tribes team just after completing an actual skiing game called Front Page Sports: Ski Racing, designed the remaining maps, such as the sprawling, open-air “Roller Coaster,” with skiing in mind. Skiing permitted players to travel so fast that it made the game’s vehicles all but irrelevant, but it also made Tribes a much more kinetic and tactically rewarding experience for players who were willing to invest the time it took to master it. “I knew it was going to be a hit when we had a LAN game going on,” says art director Mark Brenneman. “I sniped someone and I heard him scream across the office.”

Tribes was set in the same universe as Earthsiege, a line of Dynamix-developed mech-combat games that depicted a solar-system-spanning, far-future conflict between humans and sentient robots known as Cybrids. Ostensibly, Tribes was a sequel to Starsiege, the last game in that line, which actually came out a few months after Tribes. That shared world gave Tribes a tech-oriented, sci-fi aesthetic, but little in the way of plot. Although the game was initially supposed to include a complete single-player campaign, it shipped without one, making it multiplayer-only aside from a few offline training missions.

“We made the multiplayer first, and that was really all that we spent the time on,” Frohnmayer says. “And then we were like, ‘Shit, we have no single-player. All right, we’ll go multiplayer-only.’”

In 2018, many of the most popular games are exclusively devoted to online multiplayer, and even franchises that have traditionally offered or relied on single-player campaigns, like Call of Duty and Fallout, have gone online-only in their latest iterations. While plenty of players still value or prefer solo play, there’s little cause for concern about technologically limiting a game’s audience by going online-only in an era when 89 percent of U.S. adults (and 97 percent of U.S. adults ages 18 to 49) use the internet and 65 percent have broadband service at home.

Tribes debuted in a different world. According to a Census Bureau report, only 26.2 percent of U.S. households in 1998 had internet access, and a Pew Research Center poll found that only 41 percent of adults said they went online at all. Almost all of those who did had dial-up connections. In 1998, the 56k modem had just been introduced, and even in August 2000, only 11.2 percent of American home internet users claimed connections faster than dial-up. Making a multiplayer-only shooter with expansive environments and rapid travel in that bandwidth-deprived period was somewhere between gutsy and suicidal.

Even so, Dynamix persuaded Sierra to publish the title as-is, postponing the studio’s single-player plans until an anticipated expansion (which was canceled) and sequel, Tribes 2. Youngblood says Dynamix viewed Tribes less as a complete product than as a proof of concept that would leave players wanting more. “We kind of looked at it as, ‘We’re coming from nowhere. Nobody has any idea if we can make a shooter or not. So let’s make a shooter. Let’s get it out there.’ We didn’t even put copy protection on it, on purpose, so people would go out and pirate the crap out of it, which they did.”
Sierra

Quake, which came out in the summer of 1996, was originally designed for low-latency (or “low-ping”) LAN play, and until a subsequent code update, it was almost unplayable over the internet. X-Wing vs. TIE Fighter, released in April 1997, was multiplayer-only, but it suffered from severe lag in online play, and a single-player expansion followed later that year to mollify displeased players. If Tribes was to thrive despite the drastic step of ditching single-player and embracing online competition, its multiplayer implementation out of the box had to be better than state-of-the-art. Fortunately, the networking model that Dynamix developed was a marvel. As Frohnmayer and Gift later described in a 17-page paper, the network code that powered Tribes “support[ed] low-end modem connections and [was] designed to deal with low bandwidth, high latency, and intermittent packet loss.”

Frohnmayer explains (very slowly and as simply as possible) that the key component of the game’s network architecture was a “notification protocol” that streamlined communication between Tribes servers and players’ computers. Whereas UDP — or as Frohnmayer calls it, “Unreliable Datagram Protocol” (networking comedy!) — is a “fire and forget” system that sends information without verifying that it arrived, and PCP guarantees delivery but also insists on resending any dropped data before new information can be transferred, the network code that underpinned Tribes reported back whether the data reached its destination or not. That way, if information was lost because of a faulty connection, the server was notified and knew to resend the most recent state of the game.

“There were so many ideas that allowed that level of simulation to happen over a 28.8 modem,” Frohnmayer says. “I look back and I’m like, ‘It’s amazing that that much data was allowed to come through that small of a pipe.” Although the pipes are a lot larger today, the 2015 textbook Multiplayer Game Programming noted that “the model used in Tribes still has a great deal of validity” and remains “relevant to a modern-day action game.” Youngblood, who points out that the game could support up to 128 players under ideal conditions, says, “Nobody had the networking code that we had at that time, and arguably today I think a lot of games still aren’t as good as our networking code back then.”

NetworkingNetworking wizardry made Tribes playable in a world with slowpoke connections, but Ben Goldhaber, a founding member of Twitch and an esports industry veteran who counts Tribes as his favorite title, recalls that network response time still made a difference, with lower-ping (i.e., faster-response-time) players deriving an advantage. “The term that everyone used to use back in the day was ‘HPB,’ high-ping bastard, and ‘LPB,’ low-ping bastard,” Goldhaber says. “A weapon like the chain gun, you couldn’t really use it unless you had low ping, so all the HPBs would be salty about the LPBs chaining them to death.”

Certain classes that required more precision were harder for HPBs to play effectively, but it was a rarity at the time for a class system to be included at all. Tribes allowed players to select their armor and weapons, deploy remote turrets, and activate “packs” that granted particular power-ups. “Nothing like that had ever been done before,” Goldhaber says, adding, “Modern FPS games are all based around class systems, but as far as I’m concerned, that’s a natural extension of what Tribes pioneered with a system that allowed you to play a bunch of extremely different styles.”
Sierra

In addition to encouraging different approaches to the same game type, Tribes offered several team-based, non-deathmatch modes that demanded close cooperation and communication, mirroring most of the options available in modern equivalents like Overwatch. “Some of the game types that we invented became a staple for all the shooters that followed it,” Youngblood says. “We were the guys that invented the concept of ‘capture and hold,’ which basically turned into all the point-defense games.”

All told, a team of 15 to 20 people took about two and a half years to develop Tribes. Its release date never slipped significantly. To deliver a groundbreaking game on schedule, Frohnmayer says, “Many people [exhibited] Herculean levels of crazy creativity.”

For all its innovations, Tribes was overshadowed during its development by two other Sierra-published classics: Half-Life, which barely beat it to market, and Homeworld, which came out in 1999. At E3 ’98, Youngblood says, “We set up our six-person kiosk right next to Half-Life, and that was on purpose. Everybody came over to play Half-Life, and they would hear the commotion of what was going on over at Tribes and would come over and play our game.” That strategy didn’t work as well after release. Half-Life, which was the sixth-best-selling PC game in November and December of 1998, ranked fifth with more than 445,000 copies sold in 1999; Tribes sold fewer than 99,000, although there’s no telling how many pirates played it.

Tribes’ modest sales may have stemmed in large part from piracy and its commitment to multiplayer, but its learning curve probably bore some responsibility. Most of the qualities that set Tribes apart from other shooters of its time — jet packs and skiing; long-distance combat; classes; objective-based game modes; customizable logos, skins, and voice commands; even a top-down “commander screen” that allowed one player per squad to set waypoints and plan attacks — also made it more intimidating for noobs to pick up and play. Shooters were so new in 1998, Youngblood says, that “at that point, there was no concept of a casual shooter player.” If any such players did exist, they weren’t being courted by Tribes. Watch veteran Tribes players in action, and one wonders how anyone aims at anything that’s moving so fast.

The flip side of the game’s relative impenetrability to new players was the way it rewarded diehards, who welcomed the challenge, valued its high skill ceiling, and wanted to keep testing their prowess against experienced opponents. “Tribes actually was part of the foundation of esports,” Goldhaber says. Although it hardly had the impact of StarCraft or even shooters like Quake, Unreal Tournament, and Counter-Strike, Tribes spawned a dedicated competitive community, giving rise to hundreds of clans of 10 to 20 players that competed for position on 10-vs.-10 tournament ladders hosted at sites such as Online Gaming League. Long before Halo 3’s “theater” mode, Dynamix also included a “demos” feature that allowed players to record their gaming sessions and upload them for others to play back inside Tribes, which was useful both for bragging rights and for demonstrating tactics.

Thanks to its complexity, Tribes snagged another first in 1999 by inspiring the inaugural attempt at shoutcasting, or broadcasted commentary of competitive gaming. “The very first organization to ever do shoutcasting in a semiprofessional manner,” Goldhaber says (and other sources corroborate), was the Tribes Shoutcast Network, or TSN, which was later renamed Team Sportscast Network when it branched out beyond Tribes. Marcus “djWheat” Graham, the director of Twitch Studios — whom Goldhaber calls the “godfather of esports commentating” — started his shoutcasting career at TSN. “I think one of the reasons Tribes was important is that the game wasn’t as welcoming as others … so shoutcasting was one way to bring in new players through education and sick plays,” Graham says via Twitter DM. “This also helped push the competitive scene quite a bit.”
Sierra

If Tribes had trouble transcending cult status within the general public, fellow gamemakers, at least, were properly appreciative of its milestone status. Frohnmayer says, “We definitely got feedback within the industry that people were playing it and enjoying it, and so you would have to think that whatever it was that … originated [in Tribes] would have then spread through people who did actually play it.” In 1999, the Academy of Interactive Arts & Sciences, a nonprofit organization composed of industry professionals, nominated Tribes for Computer Action Game of the Year (it lost to Half-Life) and named it Online Action/Strategy Game of the Year, while PC Gamer US awarded it a prize for “Special Achievement in Innovation.”

“For people who played the game and understood what it was and how much influence it had on the rest of games that followed it, it was absolutely a success,” Youngblood says.

TribesTribes started a new subgenre, which has informally come to be called the “zFPS” — shooters with a heavy vertical component. But it also seemingly perfected the form on the first try. “I was totally enamored [of] the freedom of movement,” says Kris “MaLicE” Patton, a 38-year-old Tribes lifer who’s been playing continuously since he was 17 and estimates that he’s sunk somewhere between 12,000 and 15,000 hours into the title. “No game has really come close to giving me the same feeling.”

Several have tried. Dynamix released the well-received but less boundary-breaking and ultimately less long-lived Tribes 2, still sans single-player (and also sans mortifying Mötley Crüe theme song), in March 2001. Five months later, Sierra — which had been sold in 1996 and started laying off Dynamix staff in 1999 — shut the studio down as part of a “strategic reorganization.” After the publisher essentially ported Tribes 2 to PlayStation 2 as Tribes Aerial Assault, the Tribes franchise was licensed to Irrational Games, which made the financial flop Tribes: Vengeance in 2004.

Another studio, Hi-Rez, acquired the IP and announced a massively multiplayer online Tribes title, then scrapped that plan and instead released Tribes: Ascend, the most recent installment in the series, in 2012. Dynamix veterans have also produced spiritual successors with Tribes-like gameplay, including Legions: Overdrive (2008) and Youngblood’s Firefall (2014), which married jet packs with resource-management elements that had been once been slated for the first Tribes. Earlier this year, a startup studio called Archetype published a crowdfunded Tribes copycat called Midair, but the game didn’t make much of a splash, and the studio wound down development almost immediately.
Sierra

“All of the sequels were pretty disappointing, especially to [Tribes] veterans,” Patton says, adding, “People kept coming back because none of the sequels could capture the speed and physics … [the original] just felt better, and more fluid, and faster in general.” Youngblood argues that other shooters that have featured jet packs, including Destiny, have erred in making jumps behave differently depending on the class and input from players while they’re already in the air. “I feel like they’ve missed the whole mark in terms of what it means to have predictable flight and combat that derives around being able to predict where the players are going to land … which was a huge staple in Tribes,” he says.

The old Tribes games are now downloadable for free, and the community has both modded and enhanced the original and designed game variants that are better-suited to smaller teams. The player base has understandably dwindled over the decades, although a contingent of longtime fans returned to Tribes over its anniversary week. One of the anniversary event’s organizers, known as Dave in real life and “Lyon” in Tribes circles, says some 350 players participated, only about 100 of whom were regulars.

Twenty years after Tribes, the typical shooter looks much more like Tribes than it did in 1998. As a result, it’s difficult to distinguish between games that are explicitly influenced by Tribes, like Titanfall, and games that have seemingly absorbed aspects of Tribes by osmosis, like Call of Duty: Black Ops III, Overwatch, Fortnite, and Bioware’s upcoming Anthem. For the faithful Patton, the only successor that really rivals his go-to game is one that does directly trace its lineage to Tribes, although it isn’t a shooter: It’s Rocket League, the 2015 multiplatform mashup of soccer and rocket cars. Rocket League, which remains popular and has established itself as an esport, also relies on fast-paced, airborne action and is deceptively difficult to master. “It has drawn a lot of us that miss the old days of Tribes,” Patton says. Youngblood understands the appeal, saying, “It’s got the feel to it.”

Scott Rudi, a Dynamix alum who briefly served as the head designer of the aborted Tribes expansion and then worked as one of the lead designers on Tribes 2, is now a game director at Rocket League developer Psyonix. Although he didn’t join Psyonix until 2018, he also sees the parallels. “A striking similarity between Tribes and Rocket League is the speed of actions and overall velocity of each game when played at high skill,” Rudi says. “Both games have the highest demands I’ve seen in gaming on a combination of player reflexes and making split-second physics calculations that are constantly changing based on other players’ actions.”

Technical artist Ben Beckwith, Rocket League’s primary level designer, is a Tribes devotee. Tribes was one of the first competitive games Beckwith played, and moving around its massive maps particularly appealed to him; instead of fighting, he says, he would often “ski into infinity.” That experience instilled a love of games in which one could “express yourself with style and tricks while maintaining flow and controlling the chaos,” which Beckwith notes is “now, of course, a big part of Rocket League.”

Rocket League’s Tribes DNA goes deeper than that. A couple of years after Tribes came out, Beckwith and his brother Adam (who also works for Psyonix) discovered a mod called Tribes Football, in which two teams battled over control of an ammo pickup that stood in for a ball. “The rules were basic while the player-movement mechanics were deep and skillful, just like Rocket League,” Beckwith says. Tribes Football led Beckwith to discover Deathball, a similar mod for Unreal Tournament 2003. A field he designed for Deathball later inspired the Rocket League level Beckwith Park.

In September 2003, Beckwith began working on “Onslaught,” a vehicle-based gametype that Psyonix developed for Unreal Tournament 2004. Beckwith calls it “a condensed Unreal version of Tribes,” and he drew on his history with Tribes to design its environments. By building in jumps and using a vehicle’s handbrake to slide into them, he could perform midair spin tricks. Those tricks were then programmed into the game, and the code behind them formed the basis of the physics for Supersonic Acrobatic Rocket-Powered Battle-Cars, the 2008 predecessor to Rocket League.

In its early iteration, SARPBC was an entertaining vehicular battle game, but Beckwith says that he and Psyonix founder and CEO Dave Hagewood worried it was too shallow. They decided to prototype additional game modes, and so Beckwith and his brother tried to replicate the Tribes Football experience, constructing an indoor soccer map and dropping in a ball. “When Psyonix’s production level plummeted because we couldn’t stop playing, we pretty much dropped everything else and focused on the new mode,” Beckwith says. “And that’s how Starsiege: Tribes influenced almost every aspect of my work on Rocket League and my entire career.” According to Psyonix, Rocket League has been played by more than 52 million people, totaling more than 3 billion matches. Without Tribes, it might not have existed (or caught on) at all.

If we can learn anything from a list of the best games that recently turned 20, it’s that there’s no one way to make a classic. Some started long-running series; some extended an existing series; others were one-offs. Some were immediate sensations; others never set the sales charts on fire. Some were products of excruciating, protracted development and delays; others hit their targets. Frohnmayer describes the Tribes development process as “pretty pure,” citing an esprit de corps that contributed to “one of the most [collaborative] game-creation leadership experiences I’ve ever had.” Losing single-player eased the developers’ pain, but Frohnmayer attributes Tribes’ birth to the fact that the team was making the game that they all wanted to play.

“We were all into it,” he says. “We were like, ‘Oh shit, this is a fun game. Oh wait, we can make it more fun this way. And then pretty soon you’re just like, ‘Holy shit, this is a whole new thing.’”

Well, no, not holy shit. But maybe holy shazbot.