StarCraft II Should Be Balanced Around Its Very Best Players

… with caveats

I like Substack because I find it to be a more casual and conversational medium. I was initially drawn to the platform as a reader, actually, to places like Slow Boring and Noahpinion. I find their informal but analytical approach to the news both informative and highly engaging. And after more than a decade of trying various newspapers and websites and feeds and what not, Substack turned out to be the first platform to meaningfully displace a good amount of my morning coffee with The Economist.

I like writing analytical commentary on various topics around real-time strategy every week, and the Substack format lends itself to exactly that kind of content. One of the great things about writing in an informal style is that it clearly communicates the opinionated nature of a piece, which is what today’s article is all about - my opinion on how StarCraft II should be balanced. And I highlight the fact that it’s my opinion right at the start to temper feelings and make it clear that I’m open to (non-enraged) counter-arguments.

Hell, I even start this piece with a couple fluffy paragraphs about Substack to lower the tension a little bit. I’m open to other ideas! (I was amazed to discover that responding in good-faith with an ask for feedback to a shitty troll comment on my Protoss underrepresentation article elicited upvotes.) I’ve been playing real-time strategy games competitively for about two decades now, and I feel like at one point or another I’ve held every opinion there is on this subject. So believe me, I get that there are other perspectives. Personally though, I just don’t think you can do better than what I’m about to describe.

I also mention all this so that folks don’t forget all the caveats. And there are caveats!

Scouts into Knights

Let me start with an example from Age of Empires II. I get that non-StarCraft examples tend to bug people, but I choose it because I want to talk about an RTS that’s actively developed and reasonably competitive but where most StarCraft players lack emotional investment.

Over the past few years, particularly since StarCraft II ceased active development, I’ve observed a sizable number of professional players, content creators and commentators try their hand at grinding the Age of Empires II ladder. I myself did the same, grinding from a genuine beginner to top 1% a few years back.

I’m always amazed to see Grandmaster-level StarCraft II players putz around in lower leagues of other games. I mean, sure - strong core mechanics are not sufficient to instantly master another game. But I sort of imagined that someone who’s good enough to be Grandmaster at the most competitive real-time strategy game in the non-Korean world would have at least some theoretical understanding of how to improve at these games. And maybe they do, but for whatever reason, they generally chose not to apply these lessons to Age of Empires II.

And the net result of this is that I can point you to several Grandmaster players who, after hundreds of games, arrived in the Age of Empires II equivalent of, like, platinum league.

Now, consider this idea: I could sit down with your average Diamond-level StarCraft II cheeser and teach them a simple, non-gimmicky, professional-level build order: 20-pop scouts into knights into either a committed push or a TC boom. And I’m not here to tell you that I’m going to instantly turn this person into a top player. But I will claim that after not too long, this cannon rushing pleb would be beating practically every StarCraft player that’s ever given competitive Age of Empires II a solid go, no matter how many more hours of practice they had.

Am I crazy? Nah. I actually think most people would intuitively agree. Professional-level build orders are the distillation of tens of thousands of hours of learning from the very best players in the world. Scouts into knights is a simple, solid, low-skill-floor-and-high-skill-ceiling macro build that’s not only competitive at the pro level, it’s miles and miles better than 99.99% of builds you see on the ladder. No amount of mindlessly massing ladder games for the benefit of your Twitch viewers is gonna do better than that.*†

What scouts into knights is not, though, is imbalanced. The build is just an easy and strong way of playing the game, even for civs that don’t specialize in cavalry. It might not seem fair that Mr. Diamond mechanics with his little bit of practice could slap around your favorite streamer with his hundreds of games on the ladder. But that’s just the way the game works - efficient build orders are a higher leverage form of skill than most of the things you might organically pick up just from playing the game a lot.

And that means, like it or not, Mr. Diamond would be earning his wins.

Balancing For The Best

By most objective measures, the Grandmaster player is a better player - better core mechanics, better game knowledge, better game sense and understanding. But they’re not better at certain specific measures that Age of Empires II values, and this creates a lot of trouble for them, despite their otherwise higher-level of gameplay.

And I feel that this is a challenging situation to analyze through the lens of balance. I think the conclusion you would inevitably draw is that following a build order is overpowered.

You might as well say playing the game better is overpowered.

And that’s kind of an unworkable conclusion, right? I think we can all agree that there are better and worse ways of playing any given competitive RTS, and that not *everything* is a game balance issue. To get around this, I feel that you need to first rule out skill issues before you can identify balance issues. More precisely - you need to check if a meaningful skill improvement from the losing player would have changed the outcome of a given match. If not, then there’s a balance issue at play.

Fortunately, for the vast majority of players, this is as simple as looking one rung higher on the ladder to see how better players handle similar situations. Follow this consistently, and you effectively rule out balance as a factor for anyone who has a non-trivially better player above them.

(An important thing to remember here is that learning the counter to a strategy is typically much higher leverage than simply learning how to execute that strategy better. Think of the diminishing returns of Phoenix/Adept builds, for example, in contrast with how strong the response to them got. Or, think of various scenarios where having a few extra units is irrelevant in the face of their hard counter coming out in time.)

Where this approach stops working is the very best players in the world. For them, skill improvement is theoretical - it may not be physically possible to micro better, to follow a more efficient build order, to have better game sense, to scout better, or to do any of the other things we collectively label “skill” in a real-time strategy game.

Now, having said that, even at the highest levels there are usually numerous avenues for significant skill improvement. But it takes time for professional players to train and generate said improvement in order to prove that any currently dominant strategy is actually counterable with enough skill, instead of simply being unbalanced.

Thus, the developer arrives at a state where they are left trying to determine whether said skill improvement will a) actually materialize, and b) materialize on a fast enough time scale for the health of the meta. And if the answer to either question is no, they can cut a balance patch to accelerate the process or shift the meta in a healthier direction, faster.

The net result is “balancing the game around the best players”. But what I want to emphasize is that I think this is a misleading way of describing what’s happening. Really, the developer is just managing the meta: bringing it to a good place, and keeping it there. For most players, what really matters is how a game designs how skill works.

Skill Expression

I think I would summarize my high-level perspective thusly:

Game balance comes into play when a player can no longer meaningfully improve to change the outcome of a match on a timescale that is acceptable to the designer.

One implication of this is that the impact of game balance is relative to the height of a game’s skill ceiling. Recall, if you’re familiar, the Mongol outpost rush in the early days of Age of Empires IV. Age 4’s skill ceiling in the first couple minutes of the game was simply not high enough to offer players mechanisms to counter this rush through meaningful skill improvement. This meant that anyone hitting that ceiling - I’d argue pretty much anyone playing the game seriously - was impacted by game balance when they came across this build.

StarCraft II, by contrast, features an unreachably high skill ceiling. Furthermore, even at the highest levels, small skill improvements still translate to meaningful performance changes. Think about how much even professional gamers have improved over the past few years! The reachability of the skill ceiling is much less of an issue in this game.

That leads me to the broader caveat I mentioned before: in games with unreachably high skill ceilings, most of what the developer is doing is managing the meta, not just balancing it. And I think where StarCraft II is really lacking is meaningful discussion and analysis around this very subject: how skill expression is designed and its implications on the meta. For example, I mentioned this, in jest, a few paragraphs back:

You might as well say playing the game better is overpowered.

I feel it’s not widely acknowledged - or at least not discussed enough - that playing the game better is an intentionally designed component of real-time strategy games. And that said intentional design can always benefit from tweaks and improvements.

For example, StarCraft II made a conscious early decision on how unit movement factors into player skill. Default unit AI and pathing is really, really good, meaning the simple ability to get units from point A to point B in a timely fashion is not, by itself, particularly hard. The game instead emphasizes other aspects of unit movement to enable players to express skill, from positioning to map awareness to concave/convex to splitting to what units go in the front and so on and so forth.

Or, think about my previous example regarding build orders. Age of Empires II features randomly generated maps and hundreds of potential match-ups. Good execution of an optimized build order is an important part of playing the game well, but so too is adapting your build to your map and your match-up. StarCraft II is different: with static maps and only a handful of match-ups, precision build order execution is relatively more important.

My point here is that what it means to “play the game better” is largely at the discretion of the developers.

Too often, I see people complain that certain mechanics are overpowered or imbalanced at lower levels. And, if you’ve followed me so far, you’ll know that I don’t quite agree with that take. But I think people deserve the benefit of the doubt, and I think part of what they’re saying in statements like this is that the skill expression design around certain mechanics would benefit from some tweaks, because it’s not sufficiently interesting to improve at certain parts of the game.

Balanced does not mean fun. It can be possible to meaningfully improve your way out of a particular situation without that improvement being something you actually want to do.

The funny thing about fun is that while it’s entirely subjective, it’s also undeniable. I can opine to you on the balance or design or marketing or financial or esports or whatever motivations behind any particular game mechanic. But I can’t force you to enjoy it, or to tell you that you’re wrong to prefer something else. And in aggregate, that feedback is a powerful tool to move a game in a better direction.

It’s here where I think StarCraft II could benefit from more constructive criticism. I write this article in part due to exasperation with the endless and futile balance arguments popping up recently around Protoss underrepresentation in professional play. (My own take on that is here.) These debates seemingly always boil down to people connecting their own individual experiences in lower leagues to the balance of the game at the professional level. I just don’t think that’s a very useful discussion to have.

But where people seem needlessly shy - especially when discussing non-professional play - is pointing out what they simply don’t enjoy as it’s currently implemented. Sometimes, it feels like people are ashamed for not liking certain aspects of StarCraft, as though it indicates some personal weakness on their part; that if they’re not enjoying widow mine drops or mass void rays or battery overcharge or abducts or what have you, then they simply need to “get good”.

Maybe - just maybe - they’re not good enough because the skill expression design isn’t sufficiently enjoyable for them to spend the time “getting good”.

“Protoss isn’t fun to play against, and they’re way over-represented in GM, so screw them and screw buffing them” is not a coherent balance argument. But it is a critique of the game’s skill design for people who play it seriously but not professionally. And it’s worth diving into this critique more analytically instead of getting lost in another pointless balance discussion.

We should be more willing to acknowledge folks who say that parts of the game are simply not good, at least from their perspective, even if we are unwilling to concede that this somehow implies a balance issue. And we should strive to improve in response to that feedback where possible (while acknowledging that opinions are opinions and not everyone agrees with what’s tedious and what’s not). To me, StarCraft II is a game that benefits from active meta management by the developers / professional player’s council / Zerg cabal. (That’s one of the reasons I was a vocal proponent of the game’s design patches, which I believe contributed to a healthy meta.) And I think active meta management hinges on understanding what type of skill people actually want to showcase: what they find fun, and what they don’t.

When people hear that StarCraft II should be balanced around its very best players, they think it makes their experience irrelevant. But I’m here to argue the opposite: that what this statement is really saying is that balance is irrelevant, and instead elevates skill expression design as the more important point of contention. And that’s where I think StarCraft II could benefit from more discussion and analysis and change.

Final Thoughts

I hope you enjoyed this one! I did my best to keep it as neutral and even-handed as possible, and I hope it manages to keep the discussion calm and collected, even on a topic as hot-button as this. All the best!

Until next time,
brownbear

If you’d like, you can follow me on Twitter and Facebook and check out my YouTube and Twitch channels.

* Brown Bear RTS is brought to you by the bottomless well of self-esteem that comes from watching vastly better StarCraft players hit Castle Age at the 27 minute mark after 300 ladder games.

† Some folks will interpret this as a variation of the “one trick” gimmick, in which you can gain huge amounts of MMR by specializing in a single build, even if you don’t really know what you’re doing. Scouts into knights isn’t that. It’s a macro build, not a timing attack; it’s not a gimmick; you can rock it all the way to #1 on the ladder; and the reason it’s a good starter build is that it’s the most generically RTS-y way to learn the fundamentals of Age of Empires II. I would actually make the much stronger claim that a person who spends a couple weeks learning scouts into knights will then go on to learn the rest of the fundamental openers and builds - 18-pop archers, 20-pop men-at-arms into archers, scouts into cav archers, drush FC, etc - substantially faster than someone who’s just massing ladder games.