The Failures of StarCraft 2, Pt 1

Written by: EsportsJohn

Table of Contents

  1. Defender’s Advantage
  2. Damage Numbers
  3. Macro Mechanics

Decided to do a bit of a short article about my thoughts on StarCraft 2…and then it grew into this thing. I want to be very clear that StarCraft is quite honestly one of the best things that has ever happened to me; it completely changed my view of the world, and I’ve never been the same since I first discovered it. Nonetheless, I think it failed to live up to its potential, and it’s important to look back on the history of the game, how it evolved, how it came to be, and really think critically about how it was handled. It’s important to do this sort of analysis, not just because I just want to disagree with David Kim, but because I truly believe the developers didn’t think through their design decisions properly, and thus the finished product was botched beyond belief.

I’ve been through a lot of ups and downs with StarCraft 2, and since I’ve taken the long way around to playing Brood War, I feel I’m qualified to explain the design failures of StarCraft 2 accurately with a sober and impartial approach free from “arguments of nostalgia”. StarCraft has been a huge part of my life, and I feel it’s necessary to put down on paper all of the things that have been bothering me about its design and development from day one—and hopefully in the process, I can explain why some of these things came to be in the first place.

Note: I have not played Legacy of the Void since the first few weeks of release. I know a lot of people believe that the final expansion has made a lot of progress and created a much better game, but through my own observations, watching tournaments, and seeing others play, there are still plenty of core issues that still persist and taint the potential of this beautiful game. I want to avoid making too many Brood War vs StarCraft 2 arguments, but I think it would be willfully ignorant to gloss over the things that BW did correctly just to avoid a comparison argument.

Defender’s Advantage is Dead

If you play Brood War for only a moment, you will immediately notice the insane power of defender’s advantage. With the right units, you can hold a base forever against your opponent. For instance, literally no number of marine/medic will ever break three lurkers on top of a ramp, and Protoss can camp out on one base with Templar, Dark Archons, and Arbiters safely for pretty much eternity.

This sort of defense doesn’t exist in StarCraft 2. It’s not necessarily bad that the sequel got rid of high ground advantages that relied on RNG, but the effects on the gameplay were numerous and adverse.

Perhaps the largest difference is the emergence of the “deathball syndrome”. I don’t necessarily mean the emergence of large armies, but rather the phenomenon where players will always expand outward from their main while using a rather mobile army bouncing between bases to defend. From this issue arises a whole slew of other problems, from hard counters to uninteresting economic models to unit design issues.

Here’s an example: in Brood War, one of the key concepts, particularly for Zerg and Protoss players, is to expand to other corners of the map and create two “main bases” to work outwards from. This means that you can defend one base from your opponent’s attacks while slowly building up a force at the other base. When the big doom push comes knocking at your natural expansion, you can stall out with defilers or templar while continuing to amass forces at the other corner of the map.

If you attempt this sort of strategy in StarCraft 2, one or both of your bases will likely be overrun very quickly unless your opponent doesn’t scout it. This was attempted many times in the game’s infancy, and there’s a reason why the tactic quickly died out.

A Thought Experiment

Think abstractly for a moment. Two kings are at war with one another. King Raynor has only one castle, but King Artanis has two castles placed a reasonable distance apart. If Raynor wants to take over Artanis’s empire, he will want to invade both castles. He can either split his forces and risk being unable to break either or he can overrun them one at a time; naturally, Raynor will decide to dedicate all of his forces toward one target to avoid splitting his damage too much.

Assuming unlimited resources, the king with two castles will always win. Artanis can stall out Raynor’s siege for a very long time while gathering his forces at his other castle, eventually gathering a critical mass that will allow him surround and crush the invasion or attack Raynor’s base directly; Raynor will have to either sacrifice his castle (which he can’t) or retreat with his forces intact. Either way, Artanis with his two castles comes out ahead in the war.

If you remove the defender’s advantage—say, the two kings own camps on large fields—there are few incentives to creating large camps far away from each other (though you do have the perk of being able to relocate easily). Instead, the kings will tend to clump up their resources and rely more on mobile troops who can switch very quickly between attack and defense to guard their land. History will show that this is often the case in less advanced regions, with examples such as the Huns during Atila’s reign or the Iroquois Indians in the plains region of North America; the group that was proficient on horseback and owned many horses was always on the winning side.

The second example is much closer to the accidental design of StarCraft 2. It’s not necessarily bad, but it does create a situation where bases must be tightly clustered and multi-purposed units with a lot of mobility reign supreme. If you need a more concrete example, look at the one exception in Brood War: ZvZ. In that matchup, Sunken and Spore Colonies simply don’t attack quickly enough to deal with swarms of mutalisks or zerglings, therefore negating a lot of the defender’s advantage. As such, players constantly had to match their opponent’s army in order to defend against potentially fatal attacks.

You could argue that ZvZ was borderline chaos. StarCraft 2 took this a step further into to the extreme when things like instant reinforcement (Protoss Warp-ins, speedlings on creep) and hyper utility units (like the Queen or the Mothership Core) were added to the game and even further weakened the defender’s advantage. The road since then has never yielded us a comfortable design that felt manageable. Without the proper checks and a stable set of rules, this sort of mobile warfare devolves from a brilliant allocation of troops similar to Risk into absolute chaos.

The Deathball: An Unintended Side Effect

The thought experiment above is actually great for understanding different systems of warfare and even understanding some of the asymmetric balance that occurs between the races in StarCraft, but as you can see, it comes with some serious considerations. If bases aren’t spread out, what’s the point of spreading your army out?

Deathballs were something that emerged almost immediately in Starcraft 2‘s storied past, beginning with the horrific 1 food roach swarms during the beta. Many reasons were stated in the past as to why this particular phenomenon seemed to crop up: it was the fault of “unlimited” unit selection, damage density, hyper-mobile units, weak AoE, boring unit design, economic mining behavior, etc. There’s no doubt that these things may have exacerbated the problem, but at its core, it all began with a lack of defender’s advantage.

If you have a weak defender’s advantage and have to rely primarily on numbers, then positioning becomes much more important. In the late game, a large army can only be defended by an equally large army. It’s difficult to spare even a single unit to defend outlying bases, much less split your army in two. Thus, it makes more sense to move your army in a large ball between bases, using small groups and vision to deter possible counterattacks.

Blizzard’s Attempt to Fix the Problem

Legacy of the Void has attempted to artificially solve this problem by starving players out (“expand or die”) and forcing them to take blind chances with their positioning; they must split up their army and do harassment on several different fronts to protect their own economy while slowing down their opponent’s. You will always lose something, so it becomes a battle to see who can lose less—it’s skillful, but not necessarily fulfilling. For multiple reasons, I don’t believe this is fun (though I know others believe differently). More objectively, however, it creates a world in which a “perfect game” is impossible, a sentiment that many Korean players and coaches have shared with David Kim and the design team over and over—it’s not just very hard to play well, it’s literally impossible.

One of the beauties of Brood War is that it can actually nearly be mastered. Basic macro and positioning is difficult to do, but very much achievable with many intermediate steps along the way. Most of the difficulty is in the PvE aspect, so you feel great if you played a game with high APM, great macro, and a well-executed strategy. From there, it’s a battle with your opponent to see who can out-multitask the other. That’s where the endless challenge of Brood War lies, and it’s an endless pursuit as long as players play the game competitively.

On the other hand, Legacy of the Void has an extremely low barrier of entry but forces you to make blind decisions regarding your tech, scouting, and army positioning. While this can be entertaining from a spectator’s perspective (for those “big moments”), it’s nigh impossible to practice properly because of the game’s ever-changing nature depending on the opponent, their build, and their playstyle; you cannot become proficient without either having innate godlike twitch mechanics or an uncanny ability to read your opponent and guess their next move.

To reiterate, this is a band-aid fix for a problem that runs much deeper than the surface. It’s not necessarily accurate to give the game an inherent property that actually means something, but for a game that is based on economics, Starcraft 2 fails on the premise of making economics meaningful. Unit interaction and throwing a wrench in your opponent’s plans take up a far more meaningful role than building bases and managing resources.

There are some potential fixes that could have helped to fix deathballs (such as better defender’s advantage, stronger space control, or some sort of innate base defense that can defend against small numbers of units), but a starvation economy and an increased focus on harassment has done nothing but destabilize the game.

Damage Numbers Are Out of Control

One of the key features of StarCraft 2 has always been its quick pace and smooth graphics. Compared to Brood War (or really any other RTS that came out around the same time), it runs on a beautiful, efficient engine. Everyone who’s seen a dragoon take 20 minutes to find the entrance to a ramp knows exactly the frustration that older generation RTS’s posed in terms of unit movement and animation. StarCraft 2, on the other hand, was revolutionary.

For the first time, units would glide over the terrain with precision and accuracy. Micro tricks like marine splitting, blink stalker micro, and ling/baneling wars were the apex of the game’s achievements; nothing in the world takes your breath away like watching a pro player split marines like a god. Anyone who argues for the wonky glitches and awkward unit interaction from older generation RTS’s is living in a fantasy world. Either way, we still have to face the fact that the smoothness of the engine did cause some unintentional problems.

The first inherent problem is the tendency for units to clump up. If you select a large group of units and click at a designated location, the engine will give each and every unit a command to walk to that exact spot on the map, hindered only by unit collision. Not a big deal, but it does create some issues in that groups will always travel in clusters. Add in “unlimited” unit selection, and you’ve got yourself a good old-fashioned “deathball”. One of the beauties of older generation games was that units moved in waves or small, kind of square-like groups that was messy and required micro management to keep it in line.

A ball, however, is the perfect shape for damage. With ranged units, it applies equal DPS on all sides and naturally protects itself from surrounds by eliminating the gaps in between ranks and reducing surface area. Most importantly, it greatly increases the damage density.

Damage Density is Dangerous

Damage density is the damage per second per square inch (or foot or meter or what have you). In other words, clumped up units do more damage per second.

So what makes this different from any other game? Critical mass. If you continue adding to the ball, eventually you reach a point where the diameter of the ball exceeds the range of the unit. When all of the units cannot fire at once, the ball has reached critical mass and cannot generate a higher DPS except through a concave. What happens when you can select up to 100 units at a time in StarCraft 2‘s ultra smooth engine? The critical mass almost ceases to exist in a realistic game.

Some have speculated that increasing unit collision size or refining some of the movement behavior through unintuitive engine rules might fix this problem, but it’s unlikely that these changes would ever create a more stable or glossier interface that we have currently; we do not want to go back to a clunkier system.

Assuming that the engine mechanics are here to stay, we can only influence the behavior of deathballs (which is difficult for reasons stated above) or find a way to prevent the critical mass from sublimating everything in their path.

Quality of Life Improvements Are Too Good

The second major problem arising from StarCraft 2‘s engine is the ease of utility and the smoothness of the way the units move and behave. Again, these are great improvements in quality, but they can cause some serious issues if left unchecked.

Things like smart targeting, lack of overkill, and smart casting all play a part in making the user’s experience easy and consistent. In addition, the animations in the game are clean and functional without creating visual clutter. It’s honestly a marvel in game development how few bugs and glitches StarCraft 2 has. However, these quality of life improvements also make it really easy to focus damage and gun things down very efficiently.

Smooth unit movement also makes it incredibly easy to close distances with melee units or move armies up and down ramps like a flowing river. It makes everything more mobile, more slippery, and above all, more dangerous. Added to the quick speed of the game, there’s hardly time to react to unit movements and you will almost inevitably take some damage if you’re not paying close attention. It’s not uncommon to look away at your base and look back to find your army melting to colossi beams and Psionic Storms.

To put it simply: the fluid unit movement and attack animations in StarCraft 2 are simply too good for the current damage numbers. Damage numbers have grown out of control. Again, we definitely don’t want to relive the past, but we must adapt to the new technology better than we have so far.

Reducing the Overall Damage Output

The most elegant solution is a damage nerf across the board. Oracles should not be able to clear an unattended entire mineral line in seconds. A group of marine/marauder/medivac shouldn’t be able to level a base in the blink of an eye. A group of 12+ roaches shouldn’t be able to one-shot basically any unit in the game.

The game of StarCraft 2 is actually played a notch faster than originally intended, but as the standard game speed increased, the damage numbers stayed the same. As a result, the hectic race of trying to drop in two places while maneuvering your army in an intelligent way on top of macroing perfectly has always been a delicate balance. Many games have been won and lost by a single mistake, a single moment of inattention, and it’s largely because things just die too fast. For the most part, we got used to it, but the insane pace set by Legacy of the Void sped up the game even more and created a frantic atmosphere of drops, small skirmishes, non-committal expansions, and crazy strategies. It’s become a game of making less mistakes than your opponent rather than executing thoughtful strategies perfectly.

If you ask me, the base attack of most units in the game could be toned down by 20-50%. It would feel weird at first, but giving players more time to react, micro, and play around attacks might create an illusion that the game is not so chaotic as it seems sometimes. Large spell threats like Psionic Storm, Ravager bombs, or Widow Mines could remain the same to retain those big moments where attention is absolutely necessary, but preventing critical masses from mowing down everything in sight instantly could create much more interesting game dynamics than we see currently.

Macro Mechanics Were a Bad Idea

I don’t think there should be any argument here, to be honest. Macro mechanics were designed as a way to keep players doing things and paying attention to their bases, a problem the developers appropriately identified when they simplified/smoothed out a lot of the UI. Increasing the ease of play by allowing workers to be rallied automatically, shift-clicking buildings, and increased maximum unit selection were all good things (it would be ignorant to say otherwise), but they had one major drawback: they made the game a little too easy to play. Working with the smaller maps and confined spaces to build at the time, the developers calculated that something needed to be worked out so that players had to look at their bases occasionally. The result was macro mechanics.

At the outset, it didn’t seem as if they posed any large overarching problems. Hilariously small maps like Steppes of War and Slag Pits were dominated by proxy cheeses which Terran and Protoss benefited from most, but macro mechanics allowed non-stop action during these elongated one base vs one base fights. On large maps, there seemed to be no adverse side effects other than 4-gate rushes and speedling openings. It was apparent that one-base tech was coming out a little too quickly, but that could always be solved with research time adjustments (like the ones for the bunker, warp gate timing twice, banshees, reapers, etc.); large scale macro games, however, showed no real signs that the macro mechanics caused issues. It was difficult at this time for the creators to actually gauge whether macro mechanics or some of the more common things like unit design, timing, and maps were the issues with imbalance.

In hindsight, it’s strange that they overlooked a core aspect of the game for more variable objects. While numbers or functionality of a unit can be changed to affect one circumstance, macro mechanics affected all parts of play in every circumstance. If something so core to the game isn’t accurately vetted and tested, there’s no telling what the long-term effects of it will be, and in this case, all it did was artificially speed up the game.

The Inject Larva Arms Race

When we finally reached open mapmaking that gave fair opportunity to all races and Zerg could freely take third bases, Inject Larva started an arms race. This is when we began to truly see the “three base cap” and big deathballs emerge, and it was all because Zerg could instantly remax their army off of four injected hatcheries. I personally believe the first time that macro mechanics became truly problematic was Stephano’s roach max build. This wasn’t some chimerical idea that had never been thought of before, but it did change the way that many players looked at production and defense. After that, Terran players began to build extra CCs earlier, Protoss players began taking bases earlier, Zergs got even more aggressive with their expansions—the greed got out of control because whatever drawbacks the player took from expanding early were more than made up for within a minute or two due to the macro mechanics. The economic boost gained through Chronoboost, MULEs, and Inject Larva sped up the early/mid game to an alarming speed and ushered in an artificial late game with monstrous armies.

Within a few months, the game had evolved from a mosh pit of one and two base aggressive plays and awkward macro play to a calculated game of risk that balanced greed and safety on a knife’s edge while abusing macro mechanics. Pretty soon, everyone was able to get to three bases rapidly without any danger, and we began an era of 2-base all-in or max. A few odd turtle strategies like mech or swarm host play emerged, but generally the game revolved around one thing: getting a third base and maxing out.

Legacy of the Void and Macro Mechanics

These problems persisted late into the second expansion and into Legacy of the Void. As the game grew into larger maps and freer bases, the developers began to realize they had made a huge error. The attempt was made to artificially slow down the rate of expansion and maxing out with their economic changes as well as the introduction of several more units who could break fortifications or harass mineral lines with ease. Following an outcry that the game was too difficult, the developers decided now was a good time to address macro mechanics and maybe even remove them altogether.

Removing MULEs, Chronoboost, and Inject Larva was probably the best thing they could have done with the game, but a surprising amount of backlash from the community pressured developers into bringing them back. Faux arguments that macro mechanics showed skill, allowed more choices, or were an integral part of StarCraft 2 were all fallacies backed by nostalgia; all of them failed to recognize that the insane arms race generated by macro mechanics are the reason why the game needed an economic adjustment to begin with. Removing them provides far more meaningful decisions in regards to your army positioning, how you harass, and your opening build.

Think for a moment of an early game where variations of 4-gate timings aren’t two minutes apart. Think about how much more predictable that particular pressure will be. All builds would take a little longer to get off the ground, harass units like oracles or cyclones would come out later (and at a much more reliable time), and scouting in the early game would actually be somewhat difficult. Mind games and proper control become paramount, but no longer does each player need to take risks to account for an impossibly early rush that might kill them instantly. Bases are taken somewhat more organically as players take a bit longer to mine out. There’s more early game interaction between units and less positional guesswork involved.

That’s the sort of StarCraft that feels strategic.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s