KKND 2: Krossfire Development Diary - 5

Well, with the game creation process well underway—with the game engine semi-working—the next step is to check the balance of units, to make sure one side or unit is not more powerful necessarily than another. The best way to do this, perhaps obviously, is to create missions, and tweak mission and unit properties until the balance is deemed to be right.

First things first: we need missions.

A little while back, Justin put together a mission design competition open to Beamers and their friends and family. The competition outline contained specifications for all the units in the game—although things will still be changed, even now—and we were encouraged to submit missions with the modest enticement of seeing our own mission in the final game and, for the best overall mission, a 3Dfx video card. Well, that encouraged me! I submitted two missions and although I didn't win the big prize, one of my missions will make it into the game, so I'm satisfied.

Now the mission outlines have been parceled out to the level designers - Brian, Greg and Rowan—and Justin himself so they can turn our ideas into actual missions. John has created the backgrounds for KKND2 in a series of map "tiles" and these are what're being used to build the mission maps. First, the sketches from the mission designs are interpreted and scaled up to create a map large enough for the mission - the mission maps can vary considerably in size. Then, after the important details are entered, the map compositors have to flesh out the maps with interesting features, alternative routes and potential ambush points so any important paths appear no more important or interesting than the rest of the map.

The map building work is done in two layers, the first that units can walk or move in front of, and the second which units must move behind. It was at the stage I realized this two-level process was taking place that I really appreciated the work John had done—rooftops were created separately from the bottoms of buildings, and yet they have to mesh together perfectly. Even when different roofs are placed on the same building walls, they still have to mesh perfectly. The amount of pixel-perfect work involved in making these tiles has been enormous! Even with this work, the designers have to ensure edges are smoothed, parts of tiles overwritten and tiles mirrored in order to achieve a map that has no discontinuities and looks (and acts) like realistic terrain.

Once the maps are compiled, the designs are interpreted for important points, such as the tech level the enemy can attain, and what units or buildings each team begins the game with. Once that is clarified, the designers add to the maps enemy and player buildings and oil wells that are in existence when the mission begins, and then run the mission real-time, adding more units, enemies, terrain adjustments and so forth as they go. The order of unfolding events might be changed to vary the difficulty of the mission, and the aim of the designer is to create something that's very playable and compelling, while being far from a cakewalk. I've been watching Rowan construct a mission from a design just now. He says he tries to make enemy attacks arrive from unexpected directions, and tries to outthink himself, giving the enemy plenty of terrain opportunities for ambush and so forth. It sounds like intense brainwork but a lot of fun—how could you pass up the opportunity to play games—and make them better—and get paid for it?

So, they were talking about Kollateral Damage as a name for KKND2. The best of a bad lot so far. But Marketing is rejecting it because "not everyone knows what 'collateral' means." Can you believe that?