EPTsorry for the boring title, it is accurate
Part I: Introduction
Welcome to my 1500th post on Team Liquid! For this joyous occasion I present to you some of my thoughts on how units are designed in a competitive RTS game, such as the upcoming Heart of the Swarm expansion. This blog will not be discussing the balance of units, or even other important aspects of unit design such as the role a unit plays in its race, how well the tech tree flows, and overall balance in terms of unit counters. Instead, it will focus on what I feel is an under-appreciated aspect of unit design, which is unit control design.
![[image loading]](http://i.imgur.com/DfV7U.jpg)
If it's the hatchery that gives birth to larva, is it weird to anyone else that it's the queen (female) that's doing the injecting?
Well what is unit control design, you ask? The simplest answer is that it is the design of how a unit moves and fires. In a competitive RTS game like StarCraft, unit control design governs the micro-ability of units. Unit control design is what makes direct-fire units fun to use even when they do not have any unique abilities. Why does Blizzard make certain units so fast and agile, while others are slow and sluggish? The answer to this, and more, lies below!
Part II: A Few Notable Historical Examples of Unit Control Design
Brood War did not invent the concept of unit control, or micromanagement. It did, however, take this concept to an unheard of level, much the same way Jimi Hendrix pushed the boundaries of the electric guitar as an instrument. Here, I would like to show you some noteworthy examples of design decisions I've found in my own gaming history in many RTS games, and how they have subsequently affected the gameplay.
![[image loading]](http://i.imgur.com/hEDXd.jpg)
That is many tank, yes?
Anyone who's ever played the first Red Alert game will tell you that it's all about the tanks. Tank cannons were effective against everything that their treads could not squish. But tank on tank battles in this game did have some interesting micro mechanics. This was all due to the fact that cannon projectiles took time to reach their target, and the targetting AI never lead the shots against a moving target. Therefore, when the usual focus firing occurs in a tank vs. tank army battle, it was possible to move the tank being focus fired and dodge 90% of the damage, wasting entire volleys of tank shells. This mechanic wasn't very hard to exploit, but it was cool to see.
![[image loading]](http://i.imgur.com/0GDOT.jpg)
The least likeliest unit to die in this battle is probably that sheep in the lower left hand corner.
Age of Empires 2 took micro-ability of units to the extreme by giving all units, from foot soldiers to horses to ships, the ability to do instantly change direction without a turning arc or loss in speed. Furthermore, acceleration did not exist, meaning if a unit moved at all, it would be moving at the fastest speed it could go. While this meant micromanagement potential shot through the roof, it also meant there was little diversity in terms of how different units were controlled.
![[image loading]](http://i.imgur.com/5XFbx.png)
Fuck you, APC!!! Y U NO TURN
Command and Conquer 3 came out in 2007. It was a beautiful game to look at but a poor game to play. Unit control design was tossed completely out of the window in favor of "realism". All units except for infantry had their own turning arcs, which made micromanaging groups of vehicles an exercise in frustration. Units seemed to take player orders as mere suggestions, often overriding commands with their own programmed AI.
![[image loading]](http://i.imgur.com/LfLJS.png)
Those silly European knights thought Mongol cavalry archers were dangerous. The Mongols never flew around on bat wings shooting arrows that bounced from target to target.
Starcraft (and Brood War) took unit control to a different level. Although like Age of Empires 2 most units could change direction without losing speed, most units had unique attack animation timings and all air units had their own acceleration and deceleration rates. This added a lot of depth to how different units could be controlled.
Additionally, there were a handful of units (wraiths, mutalisks, vultures) with very fast attack animations that could attack in between moving in such a way that no speed is lost between the attacks. The rarity of this feature made mutalisk, wraith, and vulture micro very special and exciting to see.
Part III: Unit Control Design for StarCraft II
What can StarCraft 2 learn from all these other games? We might argue about how unit control can potentially affect balance, and discuss how certain units need to be limited in their speed and firing animations so they do not become too overpowered.
But I am going to argue that unit control should be designed above balance. Leave balance to the numerical statistics such as attack, armor, hit points, build times, and costs. Unit control should be designed with only one goal in mind: a unit should be fun to control.
I make this statement with the design philosophy of a very specific person in mind: Shigeru Miyamoto, the creator of Mario and designer of many of Nintendo's most famous games. Various anecdotes I've read tell of how before Miyamoto gets into the specific details of any game, he first makes sure that the most basic mechanics of the game is fun for gamers. He would go around the Nintendo office with a prototype of a game with placeholder graphics, and ask people to play through it to get an opinion of whether or not the basic game play is fun. This should be the same design philosophy that goes into unit control for StarCraft 2.
![[image loading]](http://i.imgur.com/eiBGR.jpg)
Games about turtles and mushrooms are the shit.
Let's take a look at the mutalisk as an example of what I'm talking about. Any Brood War player knows about "muta micro". This is a mechanic in which mutalisks could be grouped with an overlord that is far away and safe from harm, allowing the mutalisks to be stacked on top of each other, making it hard to target individual mutalisks. This, coupled with the fast movement speed and attack animation of the mutalisk, made muta micro the premier way for Zerg to harass mineral lines. It was fantastic to see and extremely satisfying to use.
For the uninitiated. You guys missed out, man.
Now we compare the Brood War mutalisk to its StarCraft 2 counterpart. Movement speed has been decreased, mutalisks no longer have the ability to move in a tight ball, and the firing animation seemed to have been slightly lengthened. By comparison, the Brood War mutalisk is more fun to control. Isn't this, then, a step backwards in terms of unit design?
Some readers may disagree by pointing out that mutalisks are already incredibly powerful in StarCraft 2 and do not need additional, "abusive" micro mechanics. But I argue that balance can always be shifted by changing the numbers such as hit points, attack strength, armor, and so on. "Fun" is not affected by these numbers. It is however, affected by unit control mechanics. I mean, Terran already has the Thor which punishes Zerg for bunching up their mutalisks at the wrong time. Why not empower players with units that become more powerful the better they are controlled?
On the other hand we can also compare the Marine from Brood War and from StarCraft 2. The only differences are that Marines have gained 5 hit points (15 with combat shields), and their firing animation has been reduced. While it was still possible to do THIS with marines in Brood War, in StarCraft 2 they are even more powerful. Stutter stepping in StarCraft 2 is not only extremely satisfying to do, but it is also quite easy to learn. Marine unit control has been made even more fun in StarCraft 2.
Of course, increases in speed and attack capabilities are not the only criteria to making a unit fun to control. The Ultralisk has been slowed down in StarCraft 2, but is given appropriate weight in its lumbering step and massive size. With their (tentative) burrow-charge ability in HotS, groups of Ultralisks will definitely become one of the most intimidating sights in the game. In a previous blog, I talked about how units can be made more fun to use simply by changing something as simple as the sound of their attack, or what the units say when they are ordered around.
I can also point at the Phoenix as an example of where speed and firing animation has failed to make a unit particularly exciting to use. As a self-described "air superiority fighter", the Phoenix is a poor performer. Although they are statistically stronger than mutalisks and vikings, more often than not they are out ranged by vikings and outnumbered by mutalisks in actual combat. Although their superior speed and ability to fire on the move were designed as a mutalisk counter in mind, in reality they are not produced in nearly enough numbers to defeat how many mutalisks Zergs usually make. It is perhaps the lack of effort a player needs to control phoenixes that make them not as exciting to use.
Maybe this is how Blizzard tests unit balance internally. Nahhhhh.....
The bottom line is, care must be taken in unit design to make sure that units are fun to control. If a unit seems to be "too powerful" because of how it can be controlled, it is, in most cases, wiser to re-balance through other avenues than to nerf how the unit moves and fires.
Part IV: Unit Control, Design vs. Accident
Let me first tell you a story about two games.
![[image loading]](http://i.ytimg.com/vi/9TCm1TsDaW0/0.jpg)
Douchebag Samus ignores your "employees only" sign
One game is called Super Metroid. It came out in 1994. If you beat the game fast enough, you could catch a glimpse of the heroine Samus Aran in a swimsuit. Toward that end, thousands of Metroid fans tried to find methods to beat the game as fast as possible. Many fans found ways to exploit the game's physics engine to reach powerful items far earlier than the game's designers had intended, opening up new ways to play the game. Speed running was born, sequence breaking was born. Everyone lauded Super Metroid as one of the greatest platforming/exploration games ever made, due in no small part to designer oversight.
Fast forward ten years. Metroid Zero Mission came out in 2004. Nintendo, having discovered the sequence breaking and speed running Metroid communities, designed this game with these two concepts firmly in mind. The game world was a sprawling maze filled with hundreds of hidden shortcuts and secrets. There were multiple ways to reach virtually every item before they were "intended" to be found. Everything was immaculately designed, some of the secrets were truly ingenious. But in the end all secrets were discovered within months of the game's release. Metroid fans all agreed that Zero Mission was a fantastic game, but Super Metroid was still better.
The development of these two Metroid games has many parallels with the development of Brood War and StarCraft 2. Granted, StarCraft 2 still hides many untold secrets in terms of how the game play is still evolving. But those who have played Brood War and StarCraft 2 will all agree that there is still some magic in Brood War that is missing in StarCraft 2.
The moral of this story is that sometimes accidents (or coincidences) in game design can achieve far more than what even the smartest game designers can achieve. This makes sense in some part, because the designers possess the brainpower of only a few, while the player base possesses the brainpower of millions, all of them seeking to exploit the game mechanics for their own gain.
So the last part of unit control design I am suggesting is this: perhaps Blizzard should take some risks with design. Not random risks, but smart risks by granting certain units extreme control. Maximize how fun it is to control units, and leave balance to other characteristics of the unit that do not affect how fun they are to use.
That's my 2 cents on unit control design, brought to you free of charge. Feel free to discuss and disagree.
EDIT: I have zero experience designing games. I've only read a lot of articles on game design, and when I play games I often find myself analyzing how the game is designed as I play. It's really just a hobby.
------------------------------------------------------------------------------------------------------------------------------------------------
Additional Discussion from comments below
On November 19 2011 03:06 Newbistic wrote:
I see what you are saying. Higher skill ceiling should allow for more exciting spectator experience, right?
IMO that's a pretty tricky design problem so far as SC2 goes. Blizzard has made the decision in SC2 to modernize UI controls, giving players MBS, smart casting, unlimited selection, etc. I agree with this decision because it helps attract a larger player base, and because (see below) UI clunkiness shouldn't be the main reason why a game is hard. On the other hand, the high mechanical requirement for Brood War was one of the reasons why the game was so successful as a professional competitive game.
Theoretically speaking it is possible to reallocate the "lost APM sinks" into other, more combat related tasks in the game such as unit control. This re-allocation of APM is a step forward in game design because unit control is a much more apparent and flashier display of APM than something like macroing really well without MBS, especially for casual viewers. We already see some of that with marine micro and stalker micro. Unfortunately, there probably isn't enough of these micro-able units to make the perceived skill ceiling approach Brood War standards.
The obvious solution to this problem is to simply speed up accelerations, movement speeds, attack animations on a bunch of units, or add more special abilities to existing units, but this solution does have some problems. First, if too many units have uber agility and fast attack animations, then the overall rhythm of combat will become unbalanced. Even in Brood War, only a select few units had incredible micro-ability. This allowed games to have a bit of amazing micro, a bit of large armies clashing, and not just all bog down to micro wars between super-mobile units all the time. Second, if too many activateable special abilities are added, then the game loses its simplicity, which IMO at least is something that StarCraft is famous for (complex interactions between simple units).
I think the best way to tackle this design problem is to give Zerg and Protoss maybe one or two more units that are super mobile and micro-able, then look for other, more creative ways to increase micro. What I specifically have in mind is something like reaver/shuttle micro, where the APM usage is more reactionary than the rhythmic "fly in fly out" of mutalisk micro. SC2 Colossi drops aren't quite at that level yet, and none of the HotS units shown so far seem to allow for something like this (maybe the shredder?).
I see what you are saying. Higher skill ceiling should allow for more exciting spectator experience, right?
IMO that's a pretty tricky design problem so far as SC2 goes. Blizzard has made the decision in SC2 to modernize UI controls, giving players MBS, smart casting, unlimited selection, etc. I agree with this decision because it helps attract a larger player base, and because (see below) UI clunkiness shouldn't be the main reason why a game is hard. On the other hand, the high mechanical requirement for Brood War was one of the reasons why the game was so successful as a professional competitive game.
Theoretically speaking it is possible to reallocate the "lost APM sinks" into other, more combat related tasks in the game such as unit control. This re-allocation of APM is a step forward in game design because unit control is a much more apparent and flashier display of APM than something like macroing really well without MBS, especially for casual viewers. We already see some of that with marine micro and stalker micro. Unfortunately, there probably isn't enough of these micro-able units to make the perceived skill ceiling approach Brood War standards.
The obvious solution to this problem is to simply speed up accelerations, movement speeds, attack animations on a bunch of units, or add more special abilities to existing units, but this solution does have some problems. First, if too many units have uber agility and fast attack animations, then the overall rhythm of combat will become unbalanced. Even in Brood War, only a select few units had incredible micro-ability. This allowed games to have a bit of amazing micro, a bit of large armies clashing, and not just all bog down to micro wars between super-mobile units all the time. Second, if too many activateable special abilities are added, then the game loses its simplicity, which IMO at least is something that StarCraft is famous for (complex interactions between simple units).
I think the best way to tackle this design problem is to give Zerg and Protoss maybe one or two more units that are super mobile and micro-able, then look for other, more creative ways to increase micro. What I specifically have in mind is something like reaver/shuttle micro, where the APM usage is more reactionary than the rhythmic "fly in fly out" of mutalisk micro. SC2 Colossi drops aren't quite at that level yet, and none of the HotS units shown so far seem to allow for something like this (maybe the shredder?).
