Called Back to Duty: Activision on Iterating on Success

How do you follow up one of the most successful video games of this generation -- a mere year later, when it's still selling for full price? Activision senior producer Noah R. Heller is confident that Treyarch's latest iteration of the Infinity Ward-originated Call of Duty series, World at War, is up to the task.

Here, Heller discusses the technical and gameplay innovations that drove the development of the game at the Activision-owned Treyarch, which previously created Call Of Duty 3 and has just completed Spider-Man: Web Of Shadows and Quantum Of Solace, the latter of which also uses Infinity Ward's engine.

In this in-depth Gamasutra interview, Heller talks about working with Infinity Ward's engine technology, the collaboration between the two studios on the series, and the pointed mayhem that may become the hallmark of World at War -- both from an artistic and technical perspective.

You've used the Call Of Duty 4 engine for your game and, owing to that title's popularity, I'm sure you've incorporated many of the mechanics that made it such a huge success. But could you elaborate on those areas where you've struck out on your own in terms of design?

Noah R. Heller: Probably the biggest innovations revolve around the flamethrower. We built a really good fire system from the ground up. It does a lot of interesting things, propagating through the environment based on an actual wind vector.

So if you spray fire at the edge of a field, you can see it spread through the field, travel up a tree, and burn out the sniper perched in its branches. The entire environment is flammable. It's not a destructible game, but enough gets set on fire that we think it makes for really great interactive options.

The other thing we had to work hard on was communicating the Japanese military. Call of Duty has always been a pretty tightly-scripted experience, so we had to go a lot farther than we have done previously to create a Japanese enemy that was realistic.

They're much more aggressive. They have the ability to climb trees, to hide in foliage, to jump out of spiderholes, to ambush you, and to basically wait for their moment to pounce. They also have a disregard for their own well-being, which is a little bit different for Call of Duty.

Then, when it comes to cooperative play, we built the entire game so that the single player missions were also available as cooperative missions. That meant the levels needed to be bigger and feature alternate paths.

But we didn't want to take the soul of the levels away. Call of Duty is about you, with your buddies, trying to just get over that hill, trying to get to that objective. We didn't want something like: "Okay, you stand here and set the detonation charge and you stand over here and blow the charge." We wanted that gameplay to kind of evolve naturally, and that took a lot of environment work and crafting.

Then, on the multiplayer side, we really wanted to include vehicles. Treyarch has a long history of doing vehicles with Call of Duty 3 and United Offensive... We limited ourselves just to tanks because we wanted to get it perfect. [Ed.: United Offensive was developed by Grey Matter Interactive, some members of which were integrated into the Treyarch team following United Offensive's release and the studio's acquisition by Treyarch parent Activision.]

We didn't want to mess around with vehicles that might not work out. We built an armor system for the tank, so the tanks have detachable pieces of armor that you can blow off. And then, once you've detached that piece of armor, you can throw a sticky grenade in there and do extra damage.

Basically, wherever we could find a system where we felt like rewriting it from scratch, we did, in order to improve the engine.

So, on the point about it being quite scripted...

NH: And it's still scripted for this one. We had to build new and more intelligent scripts.

But one of the things that frustrated me with the previous game was the invisible checkpoints with enemies, where they'll just keep respawning until you pass an arbitrary, hidden line. Have you reworked that system?

NH: I can't say that there's no place in the entire game where enemies might be infinite, but believe me, it was a big concern to us. I'm pretty sure we reduced it down to nil in almost any situation. Except one, where from a plot perspective, it was important for enemies to keep pouring out to get the player to move on.

A lot of people are completists, and they want to kill every enemy before moving on, so we wanted to make sure that there was something for those players as well. Basically, we get it; we get that it's bad gameplay when you show how the engine works to people.

So anywhere that occurred, we tried to reduce it or limit it completely. I can't promise you you're not going to see it, but you're not going to see the clown car effect where thousands of enemies just keep streaming out of a house or something.

On a similar subject with the AI, in the previous game I think it was noticeable that, when you ramped the difficulty up to Veteran, the enemies basically cheated. They had prescient AI which knew when and where you were about to pop out, and so they'd just take your head off straight away. Have you handled difficulty increases in this game?

NH: There are a couple variables we can tweak, like how accurate the enemy is, how aware they are of you.

But the main thing we tried to do is honestly make the placement just more brutal. You've always got an advantage on the enemy; you've been through the level before, you know where they're going to be, but in Veteran mode you're going to find that they're not going to cheat.

You're really going to have to be going for headshots using the most effective weaponry. You're going to have to use that bolt-action rifle and aim for the head if you want to take an enemy out at a distance. It's a different sort of gameplay. We heard those concerns and we tried to address them.

You mentioned about the need to aim for the head at the game's highest difficulties. One of the inherent limitations of the World War II setting is surely the 1940's weaponry, which by definition is always going to be less accurate and more unusable than those in the modern or a future-set shooter.

That restraint from the premise translates into a weaker interactive tool for the player. How do you compensate for that weakness? In the most recent Brothers in Arms there were lots of situations where I found myself just plugging away for ages because my gun wasn't very accurate and that soon gets old...

NH: You'll have to forgive me for disagreeing, but I think it's a perception issue. We made sure that our weapons were doing very realistic damage. On normal difficulty a shot to the chest with a bolt-action rifle will kill an enemy instantly. On Veteran we might want to require a more targeted hit, but the weapons are very deadly and very accurate.

It's going to be more about choosing when to use what. The flamethrower is a great example, a weapon we want you to use not only to alter the environment but also to be a shield and to flush out an enemy from a region.

You're not going to use a flamethrower for a long-range engagement, and you're not going to use a bolt-action rifle to do up close and dirty work. You're going to use something like a shotgun.

We also wanted players to know that these weapons were nastier and dirtier. Not Geneva Convention approved. And that should be reflected in the kind of damage they do.

That's also why we added the limb-capping solution. You can detach a hand, detach a forearm, detach a shoulder, because we also wanted to show the power of the weapon very visually to the player, and give them the appropriate feedback. If you're taking a .30 cal to the enemy you're going to see feet and hands fall off. And that's good visual feedback, that you're doing some real damage.

What we don't have are engagements from 500 yards or a mile away based on super-telescopic sights, which are more akin to the modern era. We actually feel that it can be more visceral when the war is closer up. It's never fun to do a pixel hunt.

Other games do the hunt game, where you see a red dot over there and you know there's an enemy in that distance, and then you zoom in with your scope and you find him. That's not as interesting to us. Does that make sense?

Yeah, that makes sense. How's your flamethrower compare to the one in Far Cry 2?

NH: I'll say that I think our flamethrower is fucking awesome. You can see the flesh burn. We built this really cool procedural technology where you can watch the uniform burn away and the flesh just char as they struggle. We wanted to be very brutal and very realistic about it.

On the other hand, we gave the flamethrower an unlimited supply of ammo, because we thought it is a silly game mechanic to have you wandering around in a level, and you are picking up propane tanks or gasoline tanks and putting them on your back.

You know what, you've got a flamethrower. This is the flamethrower level. We're going to balance it appropriately for the flamethrower, because we're going to assume the player has it throughout the whole level. That lets us make better game design. The player can forgive us for unlimited ammo, because the alternate would be just as silly.

Flamethrowers are definitely in this season.

NH: There's a lot of flamethrowers out there. I will say that we spent a lot of time talking to the First Division Marines and the guys that actually use the flamethrower. I can't speak to flamethrowers in other games that aren't based on reality, but our flamethrower is.

How does the process of sharing tools and technology work between you guys and Infinity Ward? Are you in close contact with one another?

NH: From an engineering standpoint, we share our source code for the engine. They get source drops from us as we got closer to beta. As they were finishing up Modern Warfare, we got source drops from them. But that's where it ends. It's very important that both companies are insular when it comes to design.

Why is that important?

NH: Because it lets both companies have their own creative force. Honestly, it's not like if they see something in the game that they want to give comment on, that they don't have copies to review or anything like that. But it's very important that our design team is able to create their own vision, and that their design team is able to create their own vision.

Do you swap builds during the process and review each other's work?

NH: Well, at this point, because we're pretty much gold at this point, it's more us giving builds to them. Earlier in the process they were giving builds of Modern Warfare to us.

The engine is an Activision platform, something that we can all use. From a design perspective we don't want to step on each other's creativity, and the Treyarch team has its own insular design force.

Obviously there's got to be some level of competition between the two studios. How does that work? In the past your games have been unfavorably compared to their games, something you're sick of hearing I'm sure. Do you find that a motivator?

NH: It's tough, but at the end of the day we're all Activision. A great Call of Duty game just opens up opportunities to do other great things with Call of Duty.

That's nice, but you want to beat them this round, yeah?

(laughter)

NH: I'll say this: it's awesome that we're able to trade and we're able to do a Call of Duty game this year. They were able to do one last year, because on Call of Duty 3 it was nine months in the pipeline to get that game done and no team should ever have to work like that.

If this is the system that's set up so that we both get two years to make the projects, it's a really healthy system. And frankly, it's more gratifying to compete internally than it is against any other company, because at the end of the day we both win if we make a great product.

You said it was nine months dev time for COD3. How did you use the extra year for this title? How did you spread your time?

NH: I think it's been about iteration. This has been a game where we can get things wrong and then figure out how to get them right. I wasn't around with the company for Call of Duty 3, but I've heard, for instance, the quick actions during melee fights is something the team wished they had been able to spend a little bit more time on.

In this game, for instance, the dogs in multiplayer: that took many iterations to get right, to feel right. Early on it was terrible. It was very hard to make the dogs pathfind correctly or feel brutal and vicious. Some people just didn't like the idea, but it takes a good designer and a good programmer sitting in a room together iterating to make a great game.

We have a fun little bonus level that unlocks at the end of the game, that I can't say any more about. But again, that was the result of a designer and a programmer sitting down iterating and artists contributing until the idea took hold. The entire team was able to contribute on it.

That, to me, is the hallmark of good game development, and when you have a shortened cycle you never have that luxury.

Over the past 12 months COD4's multiplayer has been gigantically popular. You're introducing things like the dogs and vehicles into a finely tuned framework. How have avoided upsetting the balance?

NH: We've got the multiplayer team helmed by David Vonderhaar, who's our online creative director. They are a cutthroat group of guys that spend every waking hour playing multiplayer.

Every balance issue is put in, taken out, and the game is iterated on. We'd sometimes see six or seven builds in a single day when it came to multiplayer as that team would iterate. And then we brought in lots of focus testing and just tried to see what would feel right.

Tanks are another great example. When we first built the tanks, they weren't controlling right, and then they were too powerful, and then they were too weak. And then the team came up with the system of putting the armor together, and it balanced itself.

It all comes down to iteration. I think what's interesting about Call of Duty: World at War is the addition of the war game type and Capture the Flag. Those are much more strategic and much more slower paced than Team Deathmatch, and so the balance for them has also been very important.

What do you think your game is going to be remembered for in six months' time?

NH: Six months' time from now? I think it's going to be two things. From the single player/cooperative side it's going to be the game that took war to a level of realism that was exciting and yet a little bit disturbing. If we get that kind of reaction, I think that's a really good thing.

I think also it's really going to be showing people the way cooperative play is done. Co-op needs to be about a team working together, not special game modes, not special scripting. It's about a shared experience, and I think we're really going to deliver on that.