Rift: Building A New Realm

[Trion Worlds' first launch, Rift, is an ambitious MMO in a space crowded with also-rans from major studios -- so what sets it apart? Executive producer Scott Hartsman discusses the game in detail -- from design to launch logistics.]

Earlier this month, Trion Worlds launched its first game, Rift. The company, founded in 2006, had long trumpeted its dynamic server technology as the wave of the future for MMOs and other online games. It had also publicly debuted this project in 2008.

The game was then under the stewardship of Might & Magic creator and Trion co-founder Jon Van Caneghem, who left the company in 2009 and moved to Electronic Arts. Scott Hartsman, a veteran of the EverQuest series, is now the general manager for Trion's studio and executive producer of the project.

Gamasutra recently spoke to CEO Lars Buttler, who talked of Trion's huge ambitions -- to be at the forefront of the transformation of games to fully and permanently online experiences, as he sees it.

"This is the most social game in the history of gaming," Buttler told Gamasutra, and is a game that "perfectly fits today's landscape."

Said Buttler, "If people think there's a revolution going on in casual games right now, wait until we're done with core games."

In this interview conducted prior to Rift's launch, Hartsman goes into depth very specifically on the development process behind Rift, which is the company's first launch and comes from a studio packed with industry veterans. Everything from game design to the logistics of launching to player expectation is touched upon below.

You're not just the executive producer of Rift, but also the general manager of the studio.

Scott Hartsman: I am involved in some level with all of our games here. ... At least for the short term, it's been 99 percent Rift and 1 percent everything else.

Launching a large-scale mainstream fantasy MMO has not been the smoothest thing for anyone who's tried it for the last several years.

SH: Correct.

When you look at that landscape that you're launching into, what are your thoughts?

SH: It's kind of interesting. The first things that come into my mind are the expectations in the space for what is even competitive are through the roof. There's no doubt about that. It's not 2001 anymore. We're not in the Wild West.

The mere fact that you have a game up and running is no longer impressive. It has to be functional. It has to be smooth. It has to be worldwide. Ideally it's localized. And then on top of that, it's got to have plenty of the traditional stuff that people want in an MMO as well as a whole lot of things that make it unique. And so hitting all of those things at the same time is a hell of a challenge.

I've worked on all kinds of different software my life, and MMOs are above and beyond the biggest, most disparate types of technologies and types of services that exist really.

You touched on something, which is you have to give people what they're expecting, and also give them unique -- which I think a lot of the games that came out, sort of did one or the other, in recent times.

SH: Yeah, yeah.

And didn't really hit all the marks.

SH: I agree, I agree. Again, I don't think it's because other developers are dumb or they're bad. I think that it comes down to you need a specific mix of support, investment, talent, and vision.

And when I say talent, experience is in there, is a large part of it. It's one of the things that made me want to come to Trion in the first place. I've been here for about 16 or 17 months now. The support was there all the way up through the execs of the company, the board, everybody. The resources were there to actually stand a chance to succeed at hitting all those things.

The team that was built up already, which we've continued building over the last year, has experience with more single-player games and MMOs combined than any team I've ever heard of. Just before we unveiled the game, I took all the teams' names and added up all the MMOs and online games and online platforms that they worked on, and you know, we've got everybody from original EverQuest to Xbox Live to World of Warcraft to Warhammer and so on and so on.

That's like 25 MMOs, online games, and platforms. And when you have that group of people bringing in their combined experience, we've been able to short-circuit a lot of mistakes that other new companies kind of run headlong into.

One thing that you mention is vision. What is the creative vision for the game? How did you arrive at that? Because that is the crux of that tension between giving people what they want and giving people something new.

SH: Exactly. Trion, for the early [stages]... at the time you were probably last talking to [CEO] Lars [Buttler], the description for Trion was all about the technology, the platform, and the underlying cool ones and zeroes that make everything work. In the last, say, year and a half to two years, it's been more about, "And here is the expression of all of that cool technology and why you think it's fun."

For us, it's been more about creating an interesting world that has stuff that people will expect in it, you know, like their overhead exclamation mark farms where you get quests, and your instances where you can go into a group and fight dungeons. But that's a very binary way of looking at the world, and that's really how most people see it right now.

What we're doing is we're using this entire incredible dynamic content engine -- creating entire new types of content in between those two binary extremes.

So, for example, the signature thing you can do in our game, the pinnacle thing that everybody comes on board and says, "Holy shit, this is amazing," is you come out of the newbie experience, we think you've gotten a good handle on who you are and how to play the game -- these zones are under attack.

There are gigantic, massive attacks that are only possible because of our server architecture. Five, six, seven hundred people all in the same area fighting these gigantic invasions that are templated by designers and triggered and run by the gameplay system itself.

So, you can be sitting there, level 8 in our game, in your first raid, fighting off this area, defending your part in the world, and if you fail, it will be taken over. Stuff does change. So, it's not the same. It's not "log in and get the same gameplay experience. Oh, I'll just move on to the next exclamation mark farm."

We have them there, but that's the baseline to us. That's not the interesting signature. When a big colossus comes tromping through Silverwood and starts killing all the quest NPCs, takes over an entire quest camp, sets up a foothold of his own -- and starts summoning other invaders and they have now taken over this area, what are you going to do? You know, that doesn't happen in other games.

Like I said, we're spending most of our efforts, in terms of our signature features, in what can we do that's interesting that way? What can we do to mix it up in a way that you're not going to really see in any other MMO?

So, if somebody starts playing Rift in six months after like these changes have come to the world, they'll come into this changed world?

SH: Oh, absolutely. The backend of the system technologically has the ability to know how many users there are, different levels that are online, and where they are in the world.

To us, there are situations where you might walk up to something, and you might not be able to figure out how to get past it right away. You might not be able to do a quest. Oh, no!

We used to spend a lot of time worrying about that until we actually tested it out with people, and people went, "Holy crap. This is really different." It kind of wakes them up and breaks them out of the days of just running back and forth.

And so to us, it's more about making sure that players have a way. It's about player agency. We need to make sure that if we put them into situations where they can't quite do it on their own, we give them abilities to get past some of these things, because that's where the fun is.

So, we have everything from what we call Ascended Powers, which are abilities that let you play this PvE land control game where you can build up turrets, build up your own defenses of your own quest camps, and you can actually take part in defending them and get cash and prizes for it.

As well as giving the players the ability to self-organize much, much more easily than any other MMO. I could walk up to you in the world, and if you have left yourself available to join public groups, I can walk up and push a button, and we're now in a group together. We're able to just go fight right there. Many of our events like the Rift events, the Invasion events, and all the other events don't even require us to be in a formal group. We just share credit. We share reward. And it all just works.

Is it based on your race, or your class, or that kind of stuff? Or like what realm you spawn to when you start the game?

SH: We do have two sides. The game is fictionally about the Guardians versus the Defiant. We don't have a strong good versus evil theme. We do have two very strong outlooks on reality. The Defiant were largely considered the rejects. They have abandoned their faith. They believe that saving the world is on their shoulders, so they're very technologically oriented. And the Guardian side is the handpicked letter jacket-wearing ones of the gods.

[laughs]

SH: They were the ones that were fictionally picked for being really great at something, not necessarily because they're good people. So, you can be whatever type of person you want to be, and there is a reason you fit on the side you're own.

To return to your gameplay system, the way these events are triggered, is it procedural?

SH: We have both kinds. The system itself is a very abstract concept, which is why we don't spend a whole lot of time bluntly talking about the underlying systems anymore. We talk about the cool stuff you can do in it because people understand that.

So like, these big zone events that happen tend to be functions of time, population density, and are procedurally triggered. That said, we still can and do go in and trigger them by hand. So, we can juice them to do whatever we want, but even left alone, they're there to provide a tuned, entertaining experience.

There are other kinds of content that sit on the same engine. One of the examples is something we call Ancient Wardstones, which is a PvP land control game where players of both factions can go take over all the Ancient Wardstones in a zone and forcibly trigger a gigantic colossus invasion event, where they can then attempt to defeat the colossus to gain raid rewards and all that stuff.

And then we have different scales in between. And then others are procedurally triggered but then also trigger other things. [He arranges items on the table] My soda is an unsuspecting quest camp. Your business card is a Godzilla and is going to come over and take over this area. Once it takes it over, this area that's been taken over is now alive on its own. It wants to spawn invaders. It wants to continue trying to spread. So, it will spawn its own invaders that go take over other areas of the world.

And we have six planes and two sides, Guardian and Defiant, as well as the six planes. Sorry, all eight of these forces are trying to take over the world, so they will fight with each other as well. So, it really lends itself to some interesting emerging styles of gameplay.

How do you from a design perspective stop the world from just getting screwed, you know what I mean? Can a server get so infested and miserable that it's...

SH: That's it unplayable forever?

I don't know about that. I mean, I would hope not.

SH: That's exactly it. The system is smart enough to understand, like I said, what level of players are online, which players they are. Is it possible for an area to become useless? Absolutely. That's the fun.

That's what we were worried about at first. We were really worried that, oh God, what if that happens? What if that happens? Fortunately, we had the foresight to test it before spending too many months of handwringing, because it turns out that walking to an area that's been frigging destroyed -- it's a blast.

People love being the ones who are coming back in and making things awesome, and then putting up their emplacements, reinforcing their wardstones, summoning their angels to fly over them and watch over the area. It's a true PvE land control game.

I guess you are trying to put the E into PvE.

SH: Yeah, exactly. Instead of player versus punching bag, it is in fact player versus environment. These things are definitely more than your standard bag of hit points.

And it's also putting the M in massive, because when was the last time you played an MMO where you were in a raid at level 8?

[laughs]

SH: Because that happens in our game, and newbies get it right away. When one of these events start, you get a pop-up. "You want to join a group?" "Yes." Bang. Boom. You're in a raid. You look at the map. "Oh, there's something the size of the Empire State Building walking toward me. Perhaps the 20 of us should go see what he wants." You know, that kind of thing.

When you say newbie, on one hand, yeah, but you also have people who have probably been playing World of Warcraft for six years starting your game, so they kind of have a rather grounded understanding of that endgame that they've just been experiencing for most of that time. And what they want out of a game might be different. It seems like it might meet audience expectations in an interesting way.

SH: Yeah. And that's exactly it. Our biggest challenge is, how do we make a game that's interesting enough for somebody that doesn't really play MMOs, but does get games? For starters, I should say that this game is not targeted at people who don't play games. We're not going after the gamer newbie. That's just not it.

I think in the year 2011, there are enough hundreds of millions of people that have played games in the world to where we don't really need to worry about going after the FarmVille crowd. FarmVille takes care of them just fine. So, we are going after gamers and giving gamers new experiences.

So, yeah, if you look at a game like World of Warcraft, how many tens of millions have touched that game in its life? If you look at EverQuest, how many gamers have touched that over its life? You can build a really ridiculously healthy business by just making that group of people happy. If nothing else, we definitely know the audience that we're going after.

One thing that just intrigued me was this concept that you're not going for good versus evil -- which is like, it's about time somebody said that, I think. [laughs]

SH: [laughs] Good.

Because, you know, that sort of stark morality... It has its own appeal, but I think it very quickly became done to death.

SH: I agree. Again, a lot of it goes to the average audience maturity level. People get it these days. They understand that... You don't have to paint a black and white picture for them to understand or be interested in something.

And I definitely think there are some interesting story possibilities. ... For example, the back story on our Guardian side... Some people will take the shallow look and go, "Oh, they must be the good guys."

So, you've got this series of gods that realize their own universe is being screwed with. The gods need an army, and it finally got to the point where they didn't need to care who they took. They couldn't afford to be picky and only take the pious. You were the best assassin in your neighborhood and you got killed? Guess what? The Guardians want you. We actually do tell some stories in-game, about people, their fall from grace or their reascension. There's all kinds of good fiction you can build up around that, because we really don't want to turn it into like a bad cartoon.

And "good versus evil" can really be wallpaper in MMOs. I don't think that Alliance versus Horde... I mean, people do role play, but in general, with World of Warcraft players by and large, I don't think that really has much influence on the people who pick what side. It's what side your friends are playing.

SH: True, exactly.

Do you think that the back story will affect people's decisions on which side they're interested in?

SH: It has so far. It really has so far. If you wanted to hypothetically role play a person that was the overly pious type and/or the overly oppressive type, lots of people would choose Guardian. We see lots of people who want to be the free-spirit rebel types getting very attached to Defiant.

We have been very, very fortunate in that the actual balance once we started throwing, you know, hundreds of thousands of people at our beta, that the populations were staying even enough for the game to still work. Because we were very bluntly afraid for quite some time until we started really explaining who the Guardians were and why they were cool that everybody was going to want to be a Defiant. But once we started getting it out there recently, it's been remarkably even, which is great.

How do you tell story in a game? What is your thought and approach on telling a story, and how story should be approached in an MMO? How much story should be thrown at people in an MMO?

SH: There is a lot of ways to try to tell story in MMOs. I worked on one MMO that was 100 percent fully voiced. That's not the way to do it because people will skip over all your voice. Click, click, click.

That you spent hundreds of thousands of dollars on.

SH: I wish it was only hundreds of thousands of dollars, but yeah, that you spend millions on. They'll just click all through it because they don't care. And so our take on it is we want you to play through the story.

So, for example, when I say these events that are designer [created]... The events are all hand-crafted by designers to tell a story. They're just system-triggered. So, they're not procedural events. The first one that we showed people was in a zone called Silverwood. The overall story of Silverwood is Prince Hylas has his keep at the center of the zone, which is where culminating zone event for the solo quest content and the storyline content takes place.

The event that gets triggered is Prince Hylas, his pact with the Plane of Life, he has now channeled hundreds of rifts opening up, dozens of invasions now spewing out of the world, and you're getting little bite-sized bits of story through his broadcast through the event, which I think is a hell of a lot. And those little broadcasts, yeah, those are voice, which is fine because it's like 10 lines.

It's not, "voice a whole game and upend the way you do content development." And so that seems to be, for us anyway, a really good way to do story because, pardon me, I guarantee you that if you're sitting there questing in that zone, A, you'll know that Prince Hylas, there's something wrong with the frigging guy because he's trying to kill everybody.

You'll know that the Plane of Life is involved because the entire zone just changed. You'll understand where he is in the zone. Then that will make sense when you go to those quests. And you'll understand who he's allied with and why it's bad. That's then also reinforced through our storyline of updates that we'll be rolling out over the next couple of years.

So, yeah, for us, we want you to play through the story. We don't want to sit there and tell you a story. We don't want you to be forced to read a story. We want you to play through the story.

I'm interested in your plan for updates. People have tried all kinds of methods. "Do we do roll-in content updates? Do we do sort of mini -- not expansions -- but chunks every so often?"

SH: Yeah.

"Do we do big expansions, huge expansions, profoundly change the world?" What is your philosophy?

SH: So, I've worked a couple different ways before. We're trying to... Again, a lot of this comes back to what other people on the team have worked on, what I've worked on, how all of our experiences blend together.

Our initial plan -- given that we have this dynamic content engine, we can add content incredibly rapidly, which is phenomenal for us. And we can do it without having to wait for new world space to be created. So, we can always be telling new stories.

We can always be introducing new enemies. We can always be pushing the story forward just through these events that you're playing. And so for us, our goal is to do things like that monthly and then punctuate it with larger key releases at the time it's right to tell the next chapter in the story.

So, yeah, there is this overall story that gets told over N number of months/years where we do have the general chapters already planned out. We know who the bad guys are. We know all of that.

The only thing we really don't know yet is what specific events, what specific pieces of content are, and what specific live features are players going to be wanting as most important... You've talked to many developers. I'm sure you hear a line that goes something like this. "We've got our content planned out for the next five years." Bullshit. I'm sorry. You don't want to do that.

We have a skeleton of where we want it to go. We know the chapters. We know the story we want to tell. We hope that that ends up being the story that is going to be the one we should be telling. But a lot of it at the end of the day is going to come down to, in terms of features and events, where the players in our game that need more stuff to do, what level are they, all that stuff. We still are very... We're very data-driven as well as having our own ideas.

I've heard that a lot, but more routinely and depending on whom I am speaking to, I hear, "We let our players tell us exactly where we're going to go, and we respond rapidly to what they want, and we test the fuck out of everything."

SH: Right. Sure.

Your game sounds like it's architectured in a way that makes it more possible, but in the past, that was not something people were planning for. But then when we see what's happened with social gaming and the way people can respond rapidly...

SH: Yeah. That's one of the things we've architected our team around in general. There are not many companies that I know of that would be putting on an open beta yet still so confident in their development that they were able to update it every day, which is exactly what we're doing right now.

Every day that we've run a beta, we have updated it each day with feedback and data from the previous day, and we've continued to do that through this beta. So, we are definitely doing some of that as well.

The storyline stuff I was talking about was more of the high-level story arc. But for instance, the entire concept of our evolution to a public group system was spurred on from beta feedback and beta play. We had individual developers of the game assigned to specific servers through our own beta to keep an eye out. Data is good and data is great, but data tempered by human experience is even better.

So, by having our own data coming in from the servers, how many people are participating in these events, how many people didn't get the opportunity to even participate -- what's the rate, what's the frequency -- what we had done is we adjusted the rates of all these things such that a given user who logged in and played for a half hour had a 95 percent chance of seeing one of these events happen.

So, it was using real data to feed back into this, then once we got that rate into a place that we were happy, we realized the next barrier they were running into was, "Man it's such a pain in the ass to actually form a group."

We launched this huge time-sensitive zone event. People spend 10 minutes trying to assemble for it. So, there's two ways to do that. Yes, you can extend it out in 10 minutes. Where's the fun in shouting for groups and doing all that crap? Instead, we'll put a system together.

I can press a button over your head if you have it set that way, and I can join you. [Trion PR] Katie can join you. [Trion PR] Chris can join you. You can walk up to a raid, you can press the merge button, and join your group into their raid, and now we can fight the Godzilla that's approaching. A lot of that has been driven by both objective and subjective data and feedback from the betas. And if we can keep doing that forever, I will be a happy camper.

You hear a lot about how more games are more solo these days than they were in the past. They are that way not because players don't like other people; it's because people just don't want the pain in the ass. And so, pardon me, when grouping up and trying to do stuff becomes a zero-friction activity, people enjoy the crap out of it.

And that's where we decided again we want to try and make the Rift-iness of Rift to come through. That really is what it is. It's about how can we pull together people in ways to where they're going to find themselves, "Where did the entire day go? Holy crap. That was amazing."

So, launching.

SH: Yeah.

Those are hard, I've been told.

SH: They're very hard. This is why we're practicing the shit out of them. The way we're testing the game is even built on one of our best experiences in the past. So, what we're actually doing right now is something none of us have tried -- but it's what we all want, given everything we've learned. "If we can do it, here's how we would do it" -- this way.

We have been running a closed test on two or three servers since April of last year. So, that's where we do our rapid daily iteration in front of a live audience, and the best things I think we did was get to the point where we can do rapid daily iteration.

Because taking an MMO from "in-development-zero-audience" mode to "we can update this reliably every day" is hard, especially when you have a 120-person development team. It's just hard.

And especially when you've got a company that's not done if before as a company, and you're just building on your ops infrastructure, your patching infrastructure, your customer service... Anyway, it's work. We've had that rapid iteration test running. Then we have the beta events that we've been doing. We did the beta events this way for a couple of different reasons.

Number one, it gives the development team time... It gives us the opportunity to look at a given beta test, react quickly during it, then take a breather -- don't have the pressure of running live then -- then bust ass on getting some core key stuff that people really want like that public group system. That's like one of like 15 features that we've actually done during the downtime between betas. So, that's great for us.

The other thing that that lets us do, obviously, on the marketing side, it lets us theme them as events, and then we run little contests around them to try to give people... "Hey, we like you. Log in, win prizes. It's great." And then we're able to give them each their own little story, which is cool, too.

But I think the most important thing it does for us is it lets us practice launching, because launching is usually something that companies do one time, and sometimes it goes well, sometimes it doesn't. We've been launching every two weeks for the last three months.

At this point, I'm used to on launch day having to be the guy sitting there being stressed out, trying to coordinate a hundred thousand people, doing all the stuff in the office. These days, we've done our seventh launch right now through beta. Our ops guys know their jobs. Our production staff knows their jobs. Our engineers know their jobs.

This open beta launch day, it started at 10 AM, I walked into work at 9:40, and went, "How we doing?" The head of development for the studio went, "Want to open them early? We're good." "Yeah, open them early." "Alright." Didn't have a single crash, didn't have a single problem for the next 24 hours. You know, it just worked. We're hoping like hell our actual launch works that well because we're certainly good at running betas.

There are two things that can happen at launch. You can exceed technical problems. Those are tough, but you can get past them if you have a compelling game. What you cannot get past is not having a compelling game at launch.

SH: Correct. I agree.

Because if your launch sucks and you launch a game that isn't robust enough, it will pretty much poison the internet.

SH: I totally agree. It's one of the same things that we talked about -- fun versus optimization, or fun versus balance, right? The guideline that we give our systems designers: "Make fun. We can balance fun. Don't give me something that's perfectly balanced, and then tell me we need to make it fun because you'll never succeed. You can add balance to fun. You can't add fun to balance."

So, yeah. It's a lot of why our Soul System turned out the way it is. You know, we've got this unique class system that gives people abilities to do all kinds of crazy, insane combinations. Mathematically speaking, the number of combos you can make is in the thousands. No, we are not hand-balancing every one of those thousand combinations. That would be asinine. We're more concerned about anything a given character can do much less any individual soul.

We knew we had a winner with the system when people started spending more time playing and experimenting with just the souls and classes sometimes than they did actually in the world adventuring. So, yeah, we knew the system was fun. We knew we could make it balanced enough. It's not going to be perfectly balanced. It doesn't need to be. It does need to be perfectly fun, though. I'm with you.

Obviously that becomes a big thing with MMOs. This gear is totally not balanced. it has to get nerfed. People start freaking out. That kind of stuff. But you're not too concerned as long as people are having fun...

SH: As long as it's not egregiously broken. Like, for example, the hypothetical example, if I am doing 100 points of damage every time I swing a sword and you're doing 1, we have a problem, but we already have it well within the range where it's not a big issue.

Yeah, I mean, of course, like any MMO, we'll correct things as they come up. That's why we're very fortunate in that we've been testing these abilities and these souls and these classes in front of users for the better part of a year. A year of time of getting beat up in front of people who are smarter than you are -- the players -- is good.