Enemy Territory: Quake Wars

Quake Wars: Enemy Territory first seeped into our collective consciousness at last year's E3. Although only available in the form of a brief teaser video, the trailer excited extreme anticipation as we saw human Global Defence Force fighters take on the nefarious Strogg in massively multiplayer battles featuring all manner of vehicles, weaponry and futuristic carnage.

Developed by Splash Damage, the UK developer team behind the admirable Wolfenstein: Enemy Territory, Quake Wars has gone on to become one of our most wanted online team shooters (it's set to debut before the end of the year although we've nothing more concrete than that right now) and it should provide a serious contender to the likes of EA's all-conquering Battlefield 2 and the forthcoming Unreal Tournament 2007.

Advertisement:

High time then to high tail it down to the wilds of Bromley, South London where Spash Damage are located and subject lead designer Paul Wedgwood to a spot of C&VG; Stroggification to learn the innermost workings of what Splash Damage are plotting.

Here's what we discovered:

Can we kick off by getting a bit of background on the game - how it came about?

Paul Wedgwood: We'd just finished Wolfenstein: Enemy Territory with id Software and I'd worked through that whole project for a year and a bit with Kevin Cloud from id Software. He was lead designer on Quake 2 and executive producer on Return to Castle Wolfenstein and executive producer on Wolfenstein: Enemy Territory and we'd spent a lot of time discussing design ideas and things we'd like to have done in Wolf ET. So in much the same way as Wolfenstein: Enemy Territory came about as a result of the wish list of things that didn't make into Return to Castle Wolfenstein multiplayer, there were ideas for Wolf ET that we really felt would improve the game.

We developed a high concept that was fired back and forth between id and Splash Damage, basically with the goal of a kind of general advancement in team play and technology and then situating it in a universe that was appropriate for the kind of gameplay goals that we had in mind. Because id Software owns the Quake universe and the Enemy Territory universe, it seemed like a perfect match because it gave us the asymmetry we wanted between the two teams which meant that the player would have the choice to fight as a member of a conventional military force or an invading Strogg army with high-tech weaponry.

So what's the story, and how does it fit into the Quake universe as a whole?

Paul Wedgwood: The game focuses on the initial Strogg invasion of Earth which takes place in around 2060 to 2065. In essence you could think of this as Quake 'Zero', as a prequel to Quake 2 and Quake 4. The Strogg arrive via slip-gates on Earth and immediately start invading the planet. They're a malicious and marauding force that's just interested in gathering resources and 'Stroggi-fying' races. The Global Defence Force are a paramilitary organisation that's seen a series of world disasters that have led to poverty and all kinds of problems and have really have become a kind of reactionary force, so they're kind of cobbled together and that's why a lot of their equipment is damaged and knackered and looks like it's in really poor condition.

However, because of the tactics and strategy you can employ, there isn't really the imbalance that you'd expect from an invading alien high-tech force against the humans.

Outside of the universe, what are the key ways in which Quake Wars will differ from the likes of Battlefield 2 and Unreal Tournament?

Paul Wedgwood: The first and most important thing is that Enemy Territory: Quake Wars has a focus on teamplay and pursuing specific military objectives. Each map is unique with unique objectives, a unique theme - like temperate or Arctic or arid environment - and a plot that drives you through these objectives, re-telling a story from the Strogg invasion. Unlike sports or arena-style games where you're Deathmatching or you're capturing flags, Enemy Territory: Quake Wars has a lot of focus on a specific frontline that progresses through the map. Players always know where the majority of combat is taking place, they always know what the next objective is that they're pursuing and as a result we believe that there's a greater level of satisfaction and enjoyment that comes about as a result of co-ordinating with other people and achieving these objectives.

What influence has Wolfenstein: Enemy Territory had on Quake Wars, both in terms of the experience you gained from that and the gameplay itself?

Paul Wedgwood: We learned a lot of lessons developing Wolfenstein: Enemy Territory. Obviously, because Wolfenstein: Enemy Territory was based on Return to Castle Wolfenstein, we started with a really big catalogue of assets and a pre-existing game to build upon. With Quake Wars it's a brand new title so it doesn't share any assets or technology with Quake 4 for example, we really just started with the basic Doom 3 engine and then started developing the additional technology in tandem with id. But the real advancement has been in gameplay. Wolfenstein: Enemy Territory was a really strong teamplay-focused game - we wanted to evolve that into larger, open areas. So in a sense it's not really a direct sequel but it is a spiritual successor to that kind of game design philosophy of a focus on teamplay and pure multiplayer combat.

You mentioned the MegaTexture technology there and how it's being used with the Doom 3 engine. Could you give us a few more details on that?

Paul Wedgwood: During the early part of the project we realised we wanted to be able to render these huge outdoor areas and John Carmack is undoubtedly the best graphics programmer on the planet. He proposed, and then devised, the idea for the MegaTexture, which is a single unique texture that covers the entire landscape.

It's 32,000 pixels by 32,000 pixels, it creates a six gigabyte source file which we then have to compress to ensure good disk space usage but actually only ends up using 8Mb of video memory on a video card. We took this basic implementation of the technology and then started developing it further so we had it working on a 3D mesh, we introduced a single parallel light source for lighting, the ability to put other models and things on the landscape, foliage, tools like mega-gen which generates the texture, geometric texture distribution, the road tool that lets you just plop roads down along a route.

As a piece of technology, it's really good because it generates really good visuals, and that helps with player immersion. But almost more importantly, it's great for gameplay because you're finally unlocked from polygons. You derive all of your properties - vehicle traction, particles, audio effects - from the MegaTexture, even things like the stuff distribution of debris, foliage placement. All of these things can be derived from texture masks and so you no longer have to have a strip of polygons that separate the road texture from the grass texture. So it also helps with performance because we can have huge terrains that use less polygons and this disconnection between the polygons and the MegaTexture means we have more effects and more efficient texture usage as well.

Can you tell us about the class system in Quake Wars and about problems of balancing the Strogg and the Global Defence Force?

Paul Wedgwood: There's a difference between traditional class-based multiplayer combat and what we really see as being specific combat roles. In some games, you select a class but really what you're doing is just selecting a different weapon or a different piece of equipment. In Quake Wars, each combat role is as distinct as a wizard is from a warrior in a role-playing game. When you pick that role, you're not only just getting different weapons and different grenades but you're also getting different items, different tools, different special abilities, different rewards as you advance your character.

In terms of balance, because the two teams are asymmetrical and they rely on completely different propulsion types for their vehicles, and different methods for managing the balance between weapons and stuff, there's no direct mathematical way of balancing those elements. Traditionally, if you had two weapons on two teams, you could do a damage over time equation and say that this weapon can fire 200 rounds a minute and do 10 damage so if we have this one to 20 damage per shot and only fire 100 rounds per minute, then we have an inherent balance between these two weapons - providing the amount the clip and the reload is the same.

With Quake Wars you can't do those equations, it's very difficult to balance an Icarus Hover Pack against a Husky Quad Bike because of all the differences and route constraints and stuff. There were two factors firstly that were introduced in Return to Castle Wolfenstein that really assists with this. Firstly, because the maps are asymmetrical as well, one team assaulting and one team defending, Return to Castle Wolfenstein multiplayer introduced the concept of spawn times.

This meant that all players spawned together in a wave and were arriving in combat together and fighting together, and they would have 20 seconds for the Strogg, 30 seconds for the GDF, which meant they met at the point you wanted them to arrive at. You also had the power bar in Return to Castle Wolfenstein that meant that in addition to collecting ammo and health you could control the balance between tools and items with an approach that had nothing to do with the damage they did or the effect that they had on things simply on kind of like a gut instinct.

Wolfenstein had about 11,000 QA hours at Activision to balance the game and make it work. With Quake Wars we're probably going to have to use the same kind of heuristic evaluation method or approach of having two teams fighting each other and then balancing things like power bar usage or spawn times and just gut feeling about how games are playing - how they're working just through a massive amount of play testing rather then being able to do any straightforward damage over time calculations.

How do the persistent character promotions work and the rewards system for continuous play?

Paul Wedgwood: In Wolfenstein: Enemy Territory we had a campaign system and experience points were accrued for the duration of the campaign and rewards were retained until the campaign finished. The everything restarted again, and that's important because gameplay effecting rewards, if they have any real value, can unbalance a match. Our general feeling is that we want to pursue persistent character advancement but that gameplay effecting rewards should be campaign length and status rewards that tell people how good a player they are, how good a leader they are, are the persistent elements.

For example, persistent military ranks are a great idea but giving somebody an additional weapon is not necessarily a good idea as it isn't fair for other players playing on the same server. In an RPG like World of Warcraft, you tend to go off and fight in specific dungeons which are appropriate for your skill level, but unless you have a really good solution to player matching on servers that's not a constraint you can apply to people running public servers. And so you end up with an unbalanced game if persistent rewards are too gameplay effecting.

You're encouraging teamplay, there's the emphasis that you must combine into teams. How's that going to work?

Paul Wedgwood: There are two elements to it. When you first start playing, I think there's a potential for the game to be a bit overwhelming if you wanted to know what every single character class does. So we've got the solo assignment system, and independently of the main objective your team is going for, the solo assignment system, based on the status of the match and certain eligibility criteria - like how many experience points you've got, what weapon load-out you've got, what special reward items you've got, how far you are away from the thing it wants to propose to you - it will assign an objective to you. And that objective will be to go somewhere, and it'll tell you exactly where you have to go, what you have to do when you get there and what you're reward's going to be for completing that objective.

That's basically a kind of a in the field, soft introduction to the game without you feeling like you're offline and some sergeant is giving you and over-expositionary explanation of how to fire the shotgun.

The second level of that, as players become more advanced an break into fireteams and what to go after objectives as a co-ordinated team, is that using the same system, we have a one-touch, context-sensitive radio key going in and you point and something, and depending on the entity that's under your crosshair, it sends the appropriate order to the appropriate person on your fireteam.

So if we're with Mac (who's demoing for us), and Mac's not very good at playing, he dies a lot and you're a Medic and I point and click at him it will send a revive request to you to go and help Mac. Equally if there's a destruction objective, and Mac's gone into to try and destroy it and obviously failed, you're a soldier, I can point at the objective and it will suggest that you go there and place your heavy explosive charge.

What's your aim for number of players supported, maximum and optimum?

Paul Wedgwood: The real sweet spot for fighting over a single objective across a frontline is 24 to 32 players. Any more than that and it really descends into a kind of street fight and co-ordination starts to break apart. In Wolfenstein: Enemy Territory there was no hard player limit, and there won't be in Quake Wars either, so if a server administrator wants to run a 50-player server or a 64-player server, then they're absolutely able to.

In the initial announcement of the game there was talk of accurate simulation of weather and atmosphere and so on. Will that actually have a direct impact on the gameplay?

Paul Wedgwood: To a degree. In the demonstration I showed you how the Mobile Command Post can be completely camouflaged in one atmospheric setting and in another it's really obviously present. Shadows and dynamic lighting play a big part in whether you can see stuff. We have a HDR-like solution called Bloom which is much cheaper and that gives us the really bright skies, lets us do sunsets and stuff.

You're certainly not holding back on the vehicles. What's been your approach to vehicles because sometimes it can be a problem - you want a mixture of infantry and vehicular combat rather than everyone just hoping in a vehicle.

Paul Wedgwood: Firstly I'd say that the general temptation in a game that features vehicles is to think of vehicles as simply replacing your avatar with improved mobility, firepower and armour. The problem with this, is that it's then not much fun when you're on foot. Wolfenstein: Enemy Territory was very focussed on foot combat and that's really critical for us, so we wanted our vehicles, rather than simply swapping out your avatar and giving you something else to fight with, to be and extension of the character classes and the way that they approach and attach specific objectives.

You really need a Titan tank to escort an MCP if you're going to establish an outpost. But we have a lot of urban environments, tunnels, caves, footways, routes that you generally are much better doing on foot. Also, you can't complete objectives except when out of a vehicle. So there's a ton of on-foot combat. The players are better equipped to fend off vehicles as well. For example, when flying in an anti-ground effect vehicle you have to stay fairly low because if you get into range of an anti-air turret belonging to the Strogg it will lock on automatically and shoot you out the sky.

Strogg tank character classes that carry the Obliterator can lock onto you and shoot you out the sky. You can't just use a vehicle to absolutely hammer the enemy from a distance and rely on it as a ranged attack. You always have to have people on foot supporting you, repairing the vehicles and helping them move forward.

What's your favourite vehicle to play around with and why?

Paul Wedgwood: I think at the moment it's the Goliath, the heavy walker, just because it's big and really powerful. The temptation is to choose your favourite vehicle as the current one that's the easiest to destroy everything with, but I think that's the one that I enjoy most when I'm playing on the Strogg team at the moment - because two of you in a Goliath is just amazing. And then I love the Anansi ground effect vehicle for the humans because I love pulling off stunts in it.

I'm really pleased about the fact that newbies are going to be able to jump into our game and fly Anansi helicopters really easily with straightforward player controls, but then see those controls evolve as their skills improve. You really have this kind of system where you reward player competence with added complexity, which is always something I've been fond of.

And you were also saying it'll be an advantage to have a three-man crew in vehicles rather than just one guy in one, or three guys in three separate vehicles?

Paul Wedgwood: Absolutely. We're playing with the idea of gunner and pilot in concert - so whether or not you grant people permission to use weapons and how that system works out. But regardless of how the gunnery system works, we absolutely are going to have a system in the game called target acquisition and nomination. This allows a pilot to tell a gunner about a target that he might not be looking at and for the gunner to inform the pilot of targets he's firing at that the pilot might not have seen.

If you take the Titan tank as an example, a guy manning the general purpose machine gun effectively plays anti-personnel role so he stops the Strogg from running up and placing plasma charges on the tank. The tank gunner can deal with vehicles and the pilot can try and get into locations where the vehicle has cover. Like, we'll build tank concealment into the map so you can just drop into these dips and just have the turret sticking out the top. When they start to communicate these targets to each other, they play a much more effective role and we discovered this just playing at our desks.

Two or three of us jumped into a tank and by shouting commands at each other you've essentially got three pairs of eyes. The vehicle becomes considerably more powerful and if you add into the mix the idea of an engineer being on board who can repair the vehicle when it's damaged, you have a really strong, small fireteam in a vehicle.

Any plans for including voice over IP?

Paul Wedgwood: Yeah, VoIP is in already. We have the same kind of 'v' chat we had in Wolfenstein: Enemy Territory, so you can type letters very quickly and it fills out the whole sentence and sends it as a fast message. We've got the traditional team chat and global chat as well.

Once the game's out, will you release tools so people can build maps?

Paul Wedgwood: It's an id Software game, and they have a fantastic reputation for releasing software development kits and level design tools. We only exist as a result of id Software releasing those tools. For Quake Wars, we started from the beginning wanting to produce a toolset that level designer and mod makers in the community would find really easy to use. As far as I'm aware the plan from id Software is to release these tools in exactly the same way as before.

Will that be around about release?

Paul Wedgwood: Generally speaking, you still want to put your tools through some sort of QA cycle but it's a slightly lower priority than the game. But for id Software... I mean, with Doom 3 it was almost on the release date of Doom 3 that they came out with the tools and things so it really does depend on id.

Marvellous, thanks for you time

John Houlihan