Dev Diary: Dark Athena's Martin Annander On The Polishing Phase

If you've been reading our site much the last few weeks, you'll know we're pretty excited about the April release of The Chronicles of Riddick: Assault on Dark Athena. You can take a look at my in-depth look at the new Dark Athena campaign, and a hands-on report with the online multiplayer, to see a few reasons why. Now we're pleased to show you the first of four diaries written by the super-talented developers at Riddick's developer, Sweden's own Starbreeze.

Starbreeze may be trailing behind GRIN for the moment in terms of providing us with horrific Swedish candies for on-air taste-testing, but we'll settle for this series of inside reports on Dark Athena's production instead. (For now.)

What Is Polish?

Martin Annander
Gameplay Designer
Starbreeze Studios

Starbreeze Studios. Uppsala, Sweden. Late 2008. Nearing gold master, the last remaining team members, sweating by their coffee-stained keyboards, are trying to fix the remaining issues on Assault on Dark Athena. Some of the issues could be considered trivial, while others may mean the difference between a hassle-free release and submission ping-pong. But all of them need to be fixed. Call it what you want, but to us it's important to give every issue its due attention. It's also when you have the opportunity to think back at all the work that led to this point.

As a gamer, I used to wonder why games couldn't be released the day after their awesome-looking teaser trailers appeared. It's already awesome--why can't they just release it?

Now I know. It's because the game is far from done. Looks can be deceiving. Naturally, it's not that hard to make something look awesome in the scope of a few seconds. But when it comes to creating the real interactive product, it's an entirely different story. Especially with hype and expectation breathing down your neck.

One of the unfortunate truths of software development holds true to games as much as anything else. Possibly more. It's that you never know where the bottleneck will be. You can never point to a future task and say, "This will be our major hold-up."

Working as strictly as everyone can manage, trying not to freak out at the first sight of a tightly scheduled deadline and not let stress take over when the producer says "tomorrow" -- these are the keys to the polish phase! The game already has all its content, features and basically everything that makes it what it is. But it's far from done. If it was released like this, it would be a jumbled pile of code and media. Not a game. At least not yet.

One of the first things to happen in this phase is that a lucky few are invited from outside the company to play the game for the first time. Without any of the illusions that you get from spending two years locked in an office with only the game as your company, their insights are surprisingly important. At this point, there are still a few crashes, some of the voices are replaced by automated text-to-speech placeholders, and more than a few glitches beg us to assure the outsider that "this won't be in the release version."

Note that this outsider is a gamer allowed to dive into a new product. Meanwhile, you have the nail-biting developers--us--hoping that the shadow glitch which suddenly erupted won't show and that the new level will have time to compile before the scheduled playthrough.

It's gamer excitement versus developer madness, more or less. It also doesn't matter how many times we tell this gaming outsider that the product has almost a year before it hits the shelves. They'll instantly comment on the glitching textures and placeholder sounds anyway.

But once that's out of the way, they'll run straight into the trap--they'll get stuck. Bad. Over and over and over again, having all of us mentally face-palming our overlooked hows and sometimes wondering about the general intelligence of humanity.

But this is also where the polish comes in. This collision with the real world is a very important lesson. How do we make our product--this wonderful game of ours--into a game that everyone can pick up and enjoy?

This kind of playthrough shows exactly where there's a need for another tutorial message, a clearer button to press where you should start the elevator, or a voice snippet that foreshadows an approaching enemy. It shows bundles of design errors in a jarringly obvious way that will have everyone reply "Oh, I was just about to fix that."

Just then, after first showing the pretty teaser trailer that tricked everyone into believing the game was almost done and then luring an unsuspecting lab rat into the trap, that's when all of us first realized how much work there was ahead of us. It's the stage where we make the game shiny or damn it to review oblivion.

Of course, if it were left to some of us to decide (myself included), the game would never leave this phase at all. There's always a small event you could add, a script you could tweak or a piece of code to rewrite.

Still, late in 2008, fixing the last few bugs we could stick into our schedule, that's when I could think back to all the iterations, the changes, and all the polish that has gone into Assault on Dark Athena.

And it still makes me smile.