EPTMandatory Brood War Digression
EDIT: This is questionable as an accurate overview, as vOdToasT explains in detail below. Read his post for a more enlightened perspective, but kindly don't make this a BW v SC2 thread in so doing.
As has been stated ad nauseum on this forum in dozens of threads over the last 2 years, Brood War did not allow the type of deathball, a-move tactics that can be quite successful in SC2. The reasons usually cited are that unit selection was limited to 12 units, that unit pathing was wonky and unreliable, and that units such as very high damage siege tanks or the high ground advantage made defending with less than your entire army a more viable option. The general consensus seems to be that deathballs were discouraged primarily because of how difficult the UI and pathing made balling up your units and sending them to attack, with certain details such as tank damage, psi storm being very nasty, and high ground advantage playing a lesser (but still significant) role. While this is often brought up as a solution to deathball-type play, it strikes me as an extremely inelegant one that will never stand the test of time moving forward in RTS design. So while Brood War certainly did not encourage or reward deathball play, the design features behind this discouragement were, in my opinion, far from ideal.
An Alternative Solution?
Were there a way to discourage deathball style play – or microless steamrolling generally – without making the UI intentionally difficult to work with, would that not be a superior approach? A solution which not only greatly rewarded intelligent and swift unit control, but made A-moving nearly suicidal while retaining an easy-to-use UI with no cap on unit selection? A solution which promised extremely dynamic, uniquely interesting battles that became increasingly more complex and intense as the skill of the players involved rose? I believe there is such an alternative. However, its implementation requires a fundamentally different combat system than that utilized in both Brood War and SC2, one that reached its greatest expression to date in an RTS game of a different flavor which, as it happens, was released on the same day as Brood War!
The Myth Franchise (minus Myth III)
The basic combat system I am envisioning can be traced back to that utilized in Myth: The Fallen Lords and Myth 2: Soulblighter. If you are not familiar with either of these titles I would not be surprised, as they have nowhere near the notoriety of Starcraft, Warcraft, Command and Conquer, etc. For the uninitiated I have included a video below exploring and discussing some key aspects of the two games’ design that seem to me to present an elegant solution to both breaking up the deathball and raising the skill ceiling of competitive RTS games greatly. But first, we need to look at how ranged units function in a game like Starcraft 2.
Ranged Units in Starcraft 2
In Starcraft most ranged units function roughly like this:
- Acquire target
- Attack animation ensues
- Projectile animation unfolds
- Damage is dealt to the target and only the target
Once the projectile is fired, it locks onto its target and will hit that target, regardless of how that unit moves or if the projectile animation actually happens to fly through the model of another unit positioned nearby. (Sometimes if a unit is killed as it fires the attack does not hit, but this is rare and really beside the point.) It is simply an animation leading up to the guaranteed hit on the unit being attacked. Units in Starcraft cannot even be properly said to fire projectiles, as it is really just an animation that auto locks onto the given target that synchronizes with damage being dealt.
There is no variable behavior or potential disturbance with, say, a marauder’s rocket: it will go to its target every time and never miss, so long as the attack was initiated and the attack animation completed. There is no potential for the attack to hit a different unit, be it allied or enemy.
Higher tier ranged units function very similarly, though they may also do additional damage in an area of effect around the unit that they hit. The colossus functions essentially the same as the stalker, it just happens to do damage in a given area around the unit that it laser beams as well. The colossus has no friendly fire, and will not damage allied units unless they are explicitly targeted. The siege tank is the most interesting deviation from the norm in Starcraft; while it still locks onto targets like the other ranged units, it damages allied units even if an enemy unit was targeted. It also has a minimum range, which gives it characteristic weaknesses if caught up to or dropped on.
As it happens, the siege tank is probably the most exciting and interesting ranged unit in the entire game. I would argue this is in no small part because it has this potential to backfire along with the notable weakness of a minimum range. It can be exploited in clever and exciting ways as a result; by dropping marines on top of a siege line, sending your zerglings/zealots to attack marines so that the tanks kill half of your opponent’s marines for you, etc. Despite this, the siege tank is still among the best units in the game and is incredibly useful in a wide variety of circumstances. Were the siege tank like the colossus it would, in my opinion, be substantially less interesting. Of the two units, there is little doubt that the colossus wins in the “A-movable” department and loses in the “strategically complex and interesting” one. (I play Protoss, so don’t go there.)
What does this general type of design encourage or make viable for ranged units? In combination with a smooth and more modern engine like SC2, this makes things like clumping up into a tight ball extremely effective in certain scenarios, such as when you have a pack of marines and are stutter stepping as you focus down a spine crawler in early game TvZ. The reason this sort of thing can be very powerful is because the number of units that can focus their fire, or just get a shot off on nearby enemies, is greatly increased when every unit is tightly packed together. The ranges of the separate units overlap to a great extent, and so you have a very potent pack of damage dealers that can shred things quite quickly. The system in Starcraft does not discourage this clumping at all due to its auto-lock attack system. There are obviously built-in ‘counters’ to this balled up unit style in the baneling, siege tank, high templar, and so on; but it is undeniable that balling up and staying tightly packed can be extremely effective in a wide variety of situations all the same.
Second, kiting is a highly effective (even mandatory at high levels of play) tactic against both melee units and ranged units with a lower range and/or higher attack speed. (Think stalkers kiting marines, marines kiting zerglings, roaches kiting zealots, etc.) The form this kiting takes is, by and large, characterized by a tight ball of ranged units moving back and shooting in sync. Because units tend to clump when you move them towards a common waypoint any arc or spread your ranged units had quickly dissipates unless small groups of units are kited back independently of one another. (This is fairly commonplace in high level TvP, where Terran players will kite back a small portion of their bio while the rest remains stationary and continues to fire.) Let us just note at this juncture that this clumping up of the ranged units rarely has any appreciable effect on their damage output, as having ranged units directly in front of other ranged units does nothing to impede their damaging enemy units.
With these considerations in mind, let’s examine what the differences are between ranged units in Starcraft and those found in Myth, and how these differences affect the way in which battles are waged.
NOTE: My mic cuts out briefly a few times and there is some lag on account of my recording program not meshing well with Myth II, but it's nothing too horrible. Just don't judge it too harshly on that score alone.
#1 - Actual Projectiles Abound
In Myth every ranged unit in the game fires an actual projectile of some sort which will, depending on where it was aimed, move across the map according to a physics engine. The crucial difference is that, since these are actual projectiles and not merely locked-on animations, they have the potential to hit anything that is in their path, friend or foe. If you fire a volley of 8 arrows into a mob of battling melee units, those arrows will behave as the physics engine and your targeting determine and strike whatever is in their path. You have the potential to shoot your own warriors in the back and kill them. The same holds true for AOE units such as bomb-tossing dwarves, warlocks, fetches, and so forth. If these units are not carefully looked after and ordered to fire from intelligent angles or at strategic locations then they absolutely can--and often do--destroy your own units. In a system such as this A-moving and deathball tactics would often be tantamount to tactical suicide: having your ranged units haphazardly lob arrows and bombs at the first pack of enemy units encountered would in many cases only serve to pepper your own melee units with arrows or, worse yet, dismember them with explosives.
#2 – Balling Ranged Units is Grossly Inefficient
Because one’s own ranged units can shoot one another they have to be spread sufficiently far apart that they can shoot around or over allied units. It often happens in Myth that if your archers are clumped in a ball then those in the middle and back of the pack cannot shoot because, were they to do so, they would simply shoot the archers in front in the back of the head. This necessitates being spread out in a line or in a deep enough formation that they have room to shoot over a nearby allied unit. The distance at which they need to be separated from an allied unit in order to shoot over them will naturally vary based on the height of the allied unit. A dwarf can be right under an archer and the archer can still fire, but if a large unit like a trow (my namesake!) is in between an archer and their target, they have to be quite far back before they can shoot. (The AI automatically detects this and moves the archers around to find a clear shot when an allied unit is directly in front of them, they do not shoot your own units in the back of the head from point blank range.) So one would rarely see a ball of ranged units as is ubiquitous in SC2, because when the units are firing actual projectiles they must be spread in a line or be staggered far enough apart that they can shoot over one another, as you would expect in a real battle. However, if the target is far enough away a tighter ball of archers could all fire due to the extreme upward angling of their bows. It is closer targets that demand a more straight-line trajectory which would really make the ball formation unworkable.
#3 – Ubiquitous Minimum Range and Lower Movement Speed
Every ranged unit in Myth has a minimum range and the vast majority are relatively slow, much like the siege tank. If a melee unit gets up to striking distance on an archer or a dwarf, they cannot attack that unit (save for an optional and extremely weak knife attack that has to be manually activated, a feature I don’t find particularly interesting). This is in stark contrast to the normal ranged units in SC2, where a marine will riddle a zealot with bullets from 6 inches off despite the psi blades carving through his torso, or a stalker will shoot a laser into an ultralisk’s maw even as it is being devoured by its pincers. Personally I find there is something irrevocably silly about this type of combat, but that is not the substantive objection to this type of game design. What ubiquitous minimum range affords is the necessity to zone melee units away from your ranged units for them to be truly effective. If melee units get up close and personal with your archers, they have suddenly become useless, or at least substantially less useful. Ranged units that are generally slower than melee units, have a minimum range, and fire actual projectiles put a premium on positional play, careful and precise engagements, wise usage of terrain, and proper abuse of the ranged units’ strengths. Ranged units stop being primarily a matter of “kite, kite, kite!” or “spread, spread!” and revolve more around abusing their range, strong AOE damage, unique abilities, and so forth. These are qualities which can be increased greatly when these units aren’t a viable stand-alone, go-ahead-and-mass-me units such as the marine, roach, or stalker.
#4 – The Attack Ground Command
Units in Myth can be commanded to “attack ground” once by control + clicking on a given location on the map. This will make an archer shoot an arrow at that spot or a warlock a fireball. This is actually an extremely useful mechanic to have when the ranged units are firing projectiles, especially where AOE is concerned. Anybody that played Myth online will likely have engaged in dwarf/warlock/mortar duels where the deciding factor in who came out on top was how well you predicted your opponent’s movements and aimed your shots on the terrain. Simply right clicking on an enemy warlock with your own warlock would result in the fireball being centered on the enemy warlock. Now, if your opponent is at all savvy, he won’t just right click on your unit; he will, rather, attack the ground approximately the radius of a fireball’s AOE in front of your warlock, because he will then fire first and hit you as you continue to saunter forward. This is because he is using the maximum range of the fireball in combination with the area of effect, much as an EMP can hit beyond its listed range on account of its effect radius.
This dynamic is very interesting and engaging, and adds another skill-based dimension that can sway the outcome of a battle. Being able to aim your projectile where you want it to go rewards those who can predict enemy movements intelligently, allows for hitherto nonexistent artillery duels, and generally provides a whole new avenue of skilled micromanagement that is absent in a game like Starcraft 2, where you must target an enemy unit (or your own unit), even with an AOE-based unit such as the tank.
To summarize: The one exception to the general character of ranged units in Starcraft, the siege tank (the reaver--now long dead--was another such exception!), offers us a glimpse of what I think would be an extremely satisfying resolution to deathball tactics and A-moving for RTS games generally, one that is inherent to the basic design of the Myth franchise:
Utilizing a combat system wherein:
(1) every ranged unit fires an actual projectile that behaves according to a physics engine and has the potential to damage both enemy and allied units regardless of targeting
(2) every ranged unit has a minimum range on their projectile attacks
(3) most ranged units are slow or medium movement speed
(4) most ranged units are relatively fragile and easily killed if caught up to
(5) every ranged unit can be commanded to “attack ground” and fire once at a specific point on the map
(6) the aforementioned changes are balanced out by appropriately strong buffs to range, damage, special abilities, and AOE beyond what is the norm in a game like Starcraft 2
What I’m NOT Suggesting
I am not suggesting that future RTS games have the same game speed as Myth, the same 3D camera as Myth, or that the units be balanced just as they are in Myth in terms of range, speed, damage, area of effect, etc. I am only pointing to what I see as fundamental features of a game engine that would discourage several common problems encountered in a modern, competitive RTS like SC2: notably deathball tactics and A-moving with units that sometimes require little to no micromanagement in order to be very effective. These changes primarily revolve around changing how ranged units operate, making them more similar to siege tanks and reavers, less like marines and stalkers, and placing them in the context of a physics engine.
Finally, I am not suggesting that the auto lock system used in Starcraft is inherently inferior. There is certainly a good deal of entertaining, difficult, and enjoyable micromanagement in this system. I have played SC2 more than any other RTS I’ve ever bought, and I’ve played a ton of them. It’s a great game. Personally, however, I would very much like to see refinement and experimentation with a combat system more akin to that used in the Myth franchise.
Concluding Remarks
As I stated in the beginning I do not expect or even desire that SC2 be altered to this sort of combat system, as I highly doubt that would ever occur. My intent is to bring to the attention of other RTS fans a particular sort of combat system that, it seems to me, would have many advantages over more popular ones for a competitive game. Micromanagement generally would be far more important and complex in an interesting, skill-based way with changes along these lines. Deathballs and A-moving armies would be hugely discouraged as a result of friendly fire and the woeful inefficiency of ranged units clumped together. This would be accomplished without having to make the game artificially hard by imposing unit selection limits or making units unwieldy with frustrating pathing. This strikes me as a notably more sustainable and elegant solution than the “harken back to BW” solution often proffered. The potential for catastrophe is much higher in this sort of system, as is the potential for top-notch control making all the difference and turning around what might have seemed a hopeless situation to a lesser player. Combining this sort of combat with a macro component a la SC2 would be extremely cool. I think this general direction would be great for competitive RTS games, and I’d like to know what you all think about this or related design changes.
In parting, I would also claim that there are further advantages to this sort of system above and beyond those affecting the competitive aspects of the game. I would also argue that this sort of system makes for vastly more interesting and dynamic battles (mainly as a result of the physics engine) and that it can produce uniquely exciting scenarios in every game, rather than recycling the same tired animations over and over, the net result of which would be a much improved spectator experience. But if I were to delve into that it would be another post entirely.
Thanks for reading!
TL;DR
![[image loading]](http://i.imgur.com/HkY0R.jpg)
