Interview: The Making Of Resistance 2's Co-Op Mode

Released worldwide and exclusively for the PlayStation 3 last month, Insomniac Games' Resistance 2 will likely be one of the biggest sellers for the system this holiday season and the go-to title for PS3 gamers wanting an online multiplayer shooter for many months to come. The game features three different modes -- single-player, multiplayer competitive, and multiplayer cooperative. Insomniac's co-op lead designer Jake Biegel spoke with Gamasutra in-depth on the development of Resistance 2's multiplayer co-op mode. Biegel not only shares how the studio took inspiration from World of Warcraft's "spirit," but how the game scales enemies and bosses to handle both small and large groups of players, and how it mines data from the Resistance community to tune the experience. Obviously, co-op is something that you guys wanted to differentiate from other modes in Resistance 2. It's very different from anything else that's out right now. Could you talk a little bit about where the inspiration came originally? Jake Biegel: We definitely looked to some games for some places to start. Our initial intent was to make a multiplayer game that people actually played together, so we looked at multiplayer games in which people are forced to play together, not apart. We looked at experiences like Team Fortress 2, in which there are dependencies on classes, and games like World of Warcraft, in which there are large amounts of people working in tandem, creating this kind of epic synergy to overcome these encounters that wouldn't be overcomeable as an individual. We really wanted to focus on teamwork and the buzz that comes from people working together online, because even though there's some downtime when you're not playing together, the buzz is that much higher when the group locks together and everything falls into place. You said the word "force" -- the idea is that the player won't feel forced, they'll feel excited, right? JB: Right. So, it's an encouragement thing. Did you analyze what design elements worked in those games, exactly? JB: For instance, one way that we had to really distinguish ourselves from the other modes in Resistance 2 was regenerating health. Co-op is the one mode that does not have regenerating health. That's a subtle change, but at the same time, it puts a huge amount of pressure on the Medic, and the group is more likely to stick together. It was funny with that one particularly, because there was push and pull, as it was the only mode that didn't have health regeneration. The moment we turned health regeneration back on, the moment people got injured, they would run to cover. They would run away from the group, because they wouldn't rely on the Medic to heal them. Just that small mechanic [helped make teamwork] that much more organic by having those dependencies. We have a lot of crossovers. For instance, the Special Ops ammo packs give healing charges to the Medic, as well as a shield boost and ammo to the soldier. There are all these crossover dependencies so that each class has contributions that they can give to the other two classes. I'm assuming you did a great deal of prototyping with the way the classes function and the way the co-op would work. Could you talk a bit about that process? JB: A lot of it was paring down my original design, because my design was very aggressive in scope, and it still is to a large degree. In the end, we treated it like a flagship mode, in terms of the depth we tried to add to the mode, and we had to taper back from that. That's where we got the gravity of the mode, by treating it like a flagship mode, and then pulling back from there. The development of it was very much a very aggressive design. It definitely got scaled back from what it originally was. You started with a paper design? JB: A paper design of the way I wanted a multiplayer experience to be -- the incentives, rewards, and progression that would be there. From there, we probably spent the most time developing the subtleties of how the classes would work together, really testing that plane to make sure the group would work together, because we had a lot of false starts and stuff that was left on the cutting room floor that just didn't work. There's a fair amount of content that's unique to all three modes, which is a challenge just in terms of -- I don't even want to use the word "justify." I'm talking about the push and pull, "If you do this, you can't do this," and resources can only go so far. That was definitely a challenge, but in the end, the upside to that is we have three modes with very distinct flavors. You can either gravitate towards one, or if you get bored with one, you can move to another, and there's that Orange Box feel to the whole thing. I didn't think about it, but you're right, they are almost like different games to an extent. They're all inside the Resistance universe, but they have their own rules and themes. They have their own play style and play mode. JB: Each mode really does have custom features. There are bosses in single-player that you won't see in co-op, although we try to use a lot of the big enemies, and you'll see them and Easter eggs in the co-op levels. We couldn't work everything into a co-op thing. There are 'berserks' you'll only see in co-op called "Specific Functionality." We have competitive specific 'berserks.' Each mode has things that are absolutely exclusive to that mode. How was the push towards developing modular content? The 'bosses' in particular are what I would call "design heavy." They are very aimed towards the protagonists taking them on and fighting them in sort of a set piece. That's difficult to make work, I would say. JB: It was difficult to make work. The challenge was more in getting people to play together, and the challenge was to go back to the modular content. We have this concept of these special zones that we flip in for the end. If you get one that's the end, we swept out a new objective so it feels more epic. We felt like we were pioneering, in terms of creating content that's dynamic and ... getting eight 'bosses' to work with eight players. Any entity that you point eight guns at all at once is going to need some special love for co-op. The AI ramifications of what happens when you have eight targets vs. one are huge. That's why our co-op team worked around the clock to add on functions or modules that would make it work in co-op. That's a good point. It's not so much just pumping up the hit points and hoping for the best; the behavior pattern of a boss is going to have to be drastically different. JB: We struggled with that. As we had eight targets around these enemies, we had to make additions and modifications to make sure that the enemies could handle not being overwhelmed and still offer a challenge for players, because when you have eight guns pointed at anything, it falls pretty quickly. So, we had to make them dynamically scaled in relation to the strength of the group. That's something I want to talk about, because it's max eight players. What's the minimum, two? JB: Right, minimum two. When you miss a class and you're playing just two, it's going to be tough. We knew that going in. We didn't want to cut any corners when it came to challenging the group. We wanted these things to feel like they were in these epic experiences that had to be overcome by a group, so we didn't really want to water it down. With that being said, two players is very challenging, especially because of the dependencies between the classes. Not having one class available [or having] a small unit is very challenging. But our most hardcore fans have enjoyed it because of that challenge, depending on which kind of challenge that they want. Eight is a pretty high requirement, for having to get eight people together at the same time to play through. How long do these levels last approximately? Because it's not a single player story, you play through a mission, and then it cuts there. JB: Right. Because of the variability, and because of the maps and all that, it really depends on what you're doing. I'd say that they take, at minimum, 10 minutes to get through with a well-oiled eight-player group, and it can go as fast with four if you have a really well-oiled group. Just to answer your question, with the variability, it's really hard to pinpoint how long these missions will take, because that really is based on the players' skill, the collection of missions that they get with any given play pack, and what map they're playing on. Did you worry about the high player requirements or the high skills requirements in building co-op? Did you want to make it a challenge for the cream-of-the-crop, hardcore audience? JB: It's an interesting challenge that we have because there are a lot of different facets, And again, balancing this thing is really a beast because of the variability and because it's this ever-changing thing. Trying to throw reigns on it is infinitely hard. It's hard to say, because ... if you get into a group with a very well-oiled team, it's really not that hard. The difficulty goes because it's like an instance or raid. You can go into that same raid with a horrible pick-up group, and you can be like, "Oh, my god. This is the most painful thing; I'm pulling my eyes out." And you go back with a group of skilled players, and all of a sudden, it's like, "What was I worried about before? This is entirely easy." And that's what we have. You might get beat up when you're playing with a bunch of noobs, but if you get in with some more skilled players or even with people who just want to play together and are open to supporting each other, you'll see that it comes much easier. We have seen people in the beta just pick this up, and it's been so reassuring to us. We've seen people just calling out orders and just getting it, plowing through this stuff. [In] the first day, we saw people who made it halfway through the level unlocks. That's not halfway through the co-op campaign, but that's just plowing your way through. There are people out there pushing the limits. Were you surprised that people picked it up as quickly as they did, in general? JB: Yes, it exceeded my personal expectations for how intuitive it was going to be, because there's a lot of subtlety there. The thing that I am probably happiest about [with] the co-op is the teamwork that it invokes, organically. People just play together, and it's really good for us and inspiring for us to see that. How many classes are there? JB: There are three. So, if you have a group of three players, you can have one of every class. JB: The game is balanced to be played with any singular class, it allows for that. If you want to experiment and go in with all Medics, you can do that. If you want to go in with all Soldiers, you can still do that, because our enemies drop ammo occasionally, so you can make it through without Special Ops. Everybody has the ability to revive [each other], but the medics revive much faster than any other class, so again, it doesn't force a Medic to be there to progress. It sounds like you got inspiration from World of Warcraft raids. Did you follow these archetypes, like Tank, Priest, or direct damage ... JB: I think it's more of the spirit than anything that we looked at for specific classes. There are no totems or anything like that. It's definitely the spirit of that forced dependency and really relying on each other, and that buzz that you get from working together to overcome things, less than actually following the classes. And even though it's similar to a couple of games, I think it's very unique from games like Team Fortress. Our Medic functions much differently than the Medic in TF2. We really did go out of our way to make sure that we had our own flavor for the Resistance franchise. It's interesting, because I'm personally very interested when designers solve problems by observing outside their genre. At the same time -- and I'm not talking about Resistance 2 -- I'm less impressed when designers think, "Hey, I really like World of Warcraft, so I'm going to turn this racing game into WoW." JB: I don't think it's like that [with Resistance 2]. I think that [when people] play it, it feels like a raid. To some degree, there's nothing new under that sun. People are borrowing from each other, but there's nothing that we ripped off. Everybody knows that people borrow things from other companies, right? It's all about the execution and how you bring it. I consider what we do to be an art form, and the greatest artists of all time have been inspired by other people's art. The thing about game development is, yes, it's a creative endeavor, but at the same time, you're also building software, and software is built around solving problems. It's an interesting synthesis. Looking at how other people solve problems is the only intelligent way to go, right? JB: Yeah, and also, as far as the history that I've seen, as long as you put your take on it and expand the sensibility, bringing something new to the table, then it's not a big deal to people. People understand that you're standing on the shoulders of giants and taking it further. What people, including myself, get really bitter about, is when they just rip this thing out and smack it in, just thinking something is going to work. And they don't care about it, or they don't understand what they're doing, they're just copying. But we didn't do anything like that. It was much more the spirit than anything we directly copied. As long as we're talking about it, the game scales, too, right? When you have two people or eight people, it scales in terms of both what types of enemies and how many enemies spawn in, right? JB: We have a variation for all those things, how many enemies and how much their hit points are, depending on how big the group is, mostly because of that analogy of having eight guns pointed at certain enemies. We don't scale health much, but just enough so things don't feel like they're paper when you're playing with a group of eight. Because when we had eight players and we weren't doing any health scaling, things were just paper. There are player numbers, player types, and even setups that we'll swap out to make sure that you get an epic ending. ... That's why you can go through a zone and be like, "Oh, this is playing out entirely differently than the last time I was here." And you might go through a level and never see some of these Easter egg bosses that we put in, and then [you'll play] 30 or 40 hours in and happen to get this combination. Did you do a lot testing to determine what combinations or settings were relevant to the different numbers of players? Did you have to do a lot of play testing? JB: Yes, [we did] a huge amount of play testing, but it still wasn't enough. To go back into the complexity of this mode with the dynamicism, the variability, the different classes, the progression, and all the upgrades that you get ... Like I said, we treated it like a flagship mode in terms of complexity and depth. And with that, it turned into this monster that we we threw as much play testing as we could at, and luckily, it's a multiplayer game. We continue to be very active on the forum. And we will be patching and making adjustments that balance it, just like any good online game does. [Marketing director Ryan Schneider] was saying that you can, to an extent, use the MyResistance registered users and mine their data to see how their game play goes. Is that something that you're going to be using? JB: Absolutely. I know we have one of the most robust stat tracking systems out there. We're tracking tens of thousands of lines and states. And a lot of that stuff we pre-requested to have data mining for -- like how many people are dying in a zone, how much grey tech is collected in a run, what's the class makeup -- all those things, we can look at that. A lot of these stats are stuff that are transparent to the players and the users of MyResistance. JB: Yeah. Most of the stuff that we use explicitly aren't stats that the player would want to know; it's stuff that we ask for specifically so we can tune the experience. That must be super valuable. JB: Yeah, it's really important to me. We made sure that we had a very lengthy list of variables that we could check for to make sure that we really have our finger on the pulse of what's happening.