Starcraft 2 Heart Of The Swarm's Senior Designer, David Kim, talks about balancing Blizzard's award-winning strategy video game
Starcraft 2 obviously has a huge community, but it also is played as a professional sport. How do you approach something as meticulously balanced as the multiplayer need to be and start adding game changers while not wrecking it?
Well, we have a lot of tools at our disposal, but you have to remember not everything is perfect. Everything has its flaws. We try our best to bring everything together and make sense of it.
The way we do this, initially is through getting feedback from our players and feedback from the pro players and the pro-casters (the commentators). We also check out tournament results, player ladders and a whole ton of stats to take into account. The way we started with Heart Of The Swarm is, we started looking at problems we couldn’t fix with numbers and stats…
Could you give me an example?
Sure, for example in Wings Of Liberty, the Zerg didn’t have a way to assault opponents who just sat back and bombarded them, until they got to the end-tier of their tech tree. So we introduced the Swarm Host and the Viper, both of which shut down distance attacks pretty effectively. So that’s the sort of problems we were looking at and that’s the sort of problem we tried to solve for Heart Of the Swarm, which is why the game feels so much more complete than Wings Of Liberty.
So when you’re thinking of adding a new unit, then, you think of the effect it’ll have first?
Yeah, that was what we did for Wings Of Liberty. For Heart Of The Swarm, we had a ton of pro-matches at our disposal, as well as tournaments that were going on at the same time as the development. By watching those games we figured out how pro players are actually using the units. We’d constructed them to be used a certain way, but often players would work out a way to use it in a way we didn’t intend. One example was when players just maxed out on SUVs and attacked their opponents, scoring an early win.
At what stage do the artists and animators get involved?
Well, the artists and designers work on the units together.
In tandem? Right from the beginning?
No. We try so many different units and so many different effects first. Take the Viper unit – that went through a whole lot of iterations. We try not to involve artists during that initial process because we don’t want to create everything and then have to change it 10 or 20 times before we’re done. So we try and wait until we’re at where we think is the homestretch for the unit – and of course, we don’t always get there – and we try and make sure that we’re sure of what we want before we hand it off to the art department. Then we work together to create something that looks really cool, but something that’s also easy to read on screen because not only will people be playing it, they’ll be watching it via eSports. Viewers have to be easily able to tell what a creature or unit is on screen during a match so they can follow what’s going on.
Do you involve the artists at later stage because you find that once a unit starts to take form, people start to get more attached to it?
We limit that as much as possible. We try not to take ownership of ideas and work as collaboratively as possible. You don’t want to start getting defensive about something that may get cut.
Speaking of which, how many units did get cut? And which ones did you miss the most?
That’s really hard to say because there are different stages to the process. Some units get cut right at the beginning. The way it works is this: say I bring an idea to the team, then they will try and attack from as many angles as possible. That way there are no flaws or holes in the balance of it before it moves onto the next stage. Units that make it into the game, most of the time, it’s pretty clear when we’ve played with them for a bit which ones are going to stay and which ones are going to get cut.
As far which ones we miss… well, they end up being bad ideas. That’s why they get cut. You don’t miss your bad ideas. Well… maybe the War Hound. Or the Replicator. Or the Shredder. (laughs).
Now, it’s pretty established Blizzard has a philosophy about making games – as developer, you concentrate on making something you’ll enjoy and since your track record is pretty well established, more often than not you’re a success. But you do know there’s a large, collection of eSports players who may get your new game and see their winning strategies completely eradicated. Is that something you allow yourselves to think about?
We do take that into consideration – especially for patches – so we try to make small steps where possible. We don’t want to eliminate every established strategy in eSports. We do want to get rid of ones that are boring or easy. Like say, you have one unit, and you find that lots of players race to build it and as soon as it’s in play, it’s game over. That’s boring.
A good example of that happened in the first six months of Wings Of Liberty where we noticed Terran players had a strategy where they just massed Reaper units off of one base and then a lot of times the Zerg just got completely wiped out. That’s not fun and it’s not skillful. Anyone can do that, right? So we got rid of that.
At the same time you have units that are key to strategies you want to keep in play. We recently nerfed (tweaked) the Zerg Infester units. Infesters were still a core unit in the Zerg army so we didn’t want to just eliminate it. So we made using it a little bit more challenging, but we didn’t get rid of any strategies that rely on upon using it.
We got our first taste of how eSports pros will take to Heart Of The Swarm at the launch. Can you see any immediate impact on how they play? Do you think rankings may change?
The diversity of the units was great and that’s the most important thing, right? They actually wanted to use the new units. If we put in a whole bunch of changes and new units and they just relied on old strategies and used what they were used to… that would be bad. The new units have to be strong. They have to be different and they have to be fun to use. Players have to want to use them.