Social: War Commander

By David Scott

I've always been a huge fan of real-time strategy (RTS) games: It started with Dune II, continued into Total Annihilation, and peaked (for me) with Warcraft III and Command & Conquer: Generals. I played those games to death, and I met several good friends through those games -- including KIXEYE co-founder Paul Preece.

Inspired by the mods we played in Warcraft, in 2007 Paul and I had developed a few tower defense games in Flash that were successful enough for us to quit our jobs and make games full-time. Fast-forward three years, and we were working on a version of desktop TD for Facebook (Desktop Defender) and I was looking for some browser-based RTS games to play in my spare time.

The games I found were simple HTML affairs: You had slots you placed buildings in (the location of buildings didn't matter) and attacks were instant; all you saw was a battle report informing you of how many troops you lost and resources you looted. For someone whose favorite part of a RTS session was the epic battles toward the end of the game, this was a huge letdown.

We wanted to merge Tower Defense and RTS into one game; a game where the placement of your buildings and defenses mattered and, more importantly, where you got to watch and take part in the attack in real time.

In early 2010 I started work on Backyard Monsters, an MMORTS that built core RTS mechanics around a friendly and playful art direction (monsters building bases, instead of tanks and guns, in your backyard). As Backyard Monsters launched to rave reviews and our confidence (and player base) grew, we updated the graphics to be more realistic and edgy.

Emboldened by Backyard Monsters' success, I set to work on creating War Commander. It was to be an MMORTS game with nothing held back; we would have tanks, guns, airplanes, helicopters, suicide bombers, the works! In short, it was to be the game we were too afraid to make only 12 months earlier -- and since then, it has become the most popular MMORTS on Facebook, with the average player spending 8.6 hours in-game per week.

 But when you go from "We should totally remake Backyard Monsters with tanks and guns," to a live game in six months with a four-person team, you end up cutting some corners with the intent of re-adding them after launch. We hadn't anticipated getting one million installs in our first three months, and we weren't ready for how quickly players would progress through the game, and so we had to spend the first few months after launch rapidly adding new content.

What Went Right

1. Keeping the Team Lean and Keen

We launched War Commander in six months with three people; that's pretty efficient! It took a year to get the headcount into double digits, and today we're still pretty lean compared to other game teams. We start all our games with just one or two people designing the game and making prototypes to prove out the mechanics, and slowly add tech and others to the team as the game gets closer to launch. Keeping the team small keeps communication overhead down to a minimum and, as a result, early progress is extremely fast.

Another thing we do right at KIXEYE is fully embrace the studio model. Each team is fully independent with the executive producer acting as the CEO of their own little company. What works for one team may not work for another, so we don't enforce certain methodologies cross-team. We do have shared learnings, but it's more organic, with people chatting over lunch to their counterparts in other teams, and weekly presentations to break down past releases and discuss upcoming features.

2. Making Tracks!

War Commander is two years old and still has weekly updates with new content and features. We now have five tracks in development: The first track is for large features that may take over a month to develop (the world map or customizable units, for example), the second track is for smaller features (anything that requires one or two weeks of dev time), and tracks 3, 4, and 5 are dedicated to improving infrastructure, fixing bugs and exploits, and improving operational efficiencies.

All five tracks work independently of each other, so if one is delayed it doesn't impact the progress on the others. We move developers between the tracks to keep things fresh for them -- it's good to have everyone on the team working on something that is player-facing now and then.

3. The Vertical Slice

With the exception of the world map (explained later in this article), we made the right calls on what to cut to get the game out on time and on quality. We had to sacrifice a few buildings to launch with air units, cull the number of infantry units to add a wider variety of weapons, and cut music to have the units speak when you give them orders, but it was all worth it.

Making a vertical slice was key to this success. It forces you to develop each part of the game a little, instead of getting bogged down in one area. As an indie game developer, I know it's very easy to narrow your focus down to a very tiny and insignificant area of a game and just keep iterating on it. If there's only one thing you take away from this article, make it this: Build a vertical slice of your game ASAP, and play through it end-to-end before developing out any one aspect.