The synergy of an army depends on numerous factors, including:

- Complementary armor categories (e.g., marines and marauders)
- Complementary targeting (i.e., ability to hit both air and ground)
- Synergistic abilities (e.g., forcefield and zealot charge; stimpack and heal)
- Complementary attack ranges vs. tanking abilities (e.g., roaches and hydras; zealots and colossi)
- Relative attack speeds
- Relative movement speeds

Some of these factors are obvious and often explicitly discussed in the SC2 community (such as forcefield/zealot charge), whereas others are seldom even considered. Here, I'd like to discuss the role of unit movement speed in determining army synergy, and differences between races in this regard.

[image loading]

Please note, this is not intended to be an imbalance thread. I am simply trying to share some of my own thoughts on game mechanics with the community. I am not suggesting there there is some sort of game imbalance resulting from different speed synergies. However, I do believe this is an important aspect of race mechanics which is not often explicitly discussed.

First, I'll list all unit speeds:

+ Show Spoiler +

Protoss

Probe 2.81

Zealot 2.25/2.75
Stalker 2.95
Sentry 2.25
Immortal 2.25
Colossus 2.25
Dark Templar 2.81
High Templar 1.88
Archon 2.81

Phoenix 4.25
Void Ray 2.25/2.95
Carrier 1.88
Mothership 1.41

Observer 1.88/2.81
Warp Prism 2.50/3.38

Terran

SCV 2.81

Marine 2.25/3.38
Marauder 2.25/3.38
Reaper 2.95/3.84
Ghost 2.25
Hellion 4.25
Siege Tank 2.25
Thor 1.88

Medivac 2.50
Raven 2.25
Viking 2.75/2.25 (air/ground)
Banshee 2.75
Battlecruiser 1.41

Zerg

Drone 2.81 (2.81)
Queen 0.94/2.50

Zergling 2.95/4.70 (3.84/6.11)
Baneling 2.50/2.95 (3.25/3.84)
Roach 2.25/3.00 (2.93/3.90)
Hydralisk 2.25 (3.38)
Infestor 2.50 (3.25)
Ultralisk 2.95 (3.84)

Mutalisk 3.75
Corruptor 2.95
Brood lord 1.41
Overlord 0.47/1.87
Overseer 1.88/2.75

N.B. All speeds taken form Liquipedia, and given to 2 decimal places. In cases where a unit has a speed upgrade or ability that modifies speed, both speeds are listed, separated by a slash. Zerg unit speeds on creep are given in brackets. I have not included unit accelerations in this analysis; these can play an important role, but are of secondary importance to speed in most cases.

What you may notice is that for the most commonly used terran and protoss army compositions, unit movement speeds are very similar, enhancing the synergy of these units.

Marine, Marauder, Medivac: 2.25/3.38, 2.25/3.38, 2.50
+ Siege Tank: 2.25

Zealot, Stalker, Sentry, Colossus: 2.25/2.75, 2.95, 2.25, 2.25
+ Immortal: 2.25

On the other hand, zerg ground army compositions have relatively different movement speeds.

Generally speaking, uniform unit speed helps simplify army movement, since the units move and attack cohesively, even when placed in the same control group. I believe the speed differences within zerg armies help to account for the perceived fragility of the zerg ball when facing a terran or protoss ball of similar or smaller supply (separate from the question of which army is more supply effective). Poor micro and/or control group assignment can easily result in fragmentation of zerg armies. Zerg armies are thus especially sensitive to the timing of engagements.

As a counterpoint, different unit speeds can improve synergy under certain circumstances. For instance, speed upgraded roaches move faster than hydras (both on creep and off), and are thus automatically more likely to take the front in battle, which is beneficial given both their shorter range and better tanking abilities. Similarly, speedlings can run ahead of banelings to trap enemy units. In general, it is also beneficial in engagements if melee units are faster than ranged units. In this respect, prior to the research of zealot charge, zealot/stalker compositions have poor speed synergy because stalkers have a tendency to run ahead of zealots.

While I don't claim that speed synergy is the dominant factor in determining overall army synergy, I think it is an important factor that is often overlooked, and I hope this post gets people thinking more about it. As an example, consider how differently terran armies would behave if the movement speeds of marines and marauders slightly differed.

GreenMachine

Canada5 Posts

January 08 2011 22:17 GMT

I would argue that due to the speed differences with Zerg armies, the need for flanking and proper setup (and ideally, choice of arena) for your battles is extremely important and uses the speed differences between units to not make a difference.

bobucles

410 Posts

January 08 2011 23:32 GMT

If you want to get into bad synergies, try Ultralisk and Zergling. Those units trip over each other ALL the time, so much so that there's no real reason to have both on the field. Add Broodlings in to completely screw up your melee line.

It's a bit strange that players run Zerglings up the front, when in reality you don't have to. Roaches have a long weapon cooldown and a VERY quick burrow. It's not disadvantageous to run the Roaches in front, let them take the first shots both ways, and burrow for the Zerglings to flood in. With some practice you don't have to lose much, if any potential Roach damage at all. The alternative is to lose half your Zerglings before firing a single shot from your Roaches.

It lets you hide your true numbers, encouraging your enemy to rush in before bringing the full forces to bear. In this respect, Zerglings act like an auxiliary or cavalry force for your Roaches.

Black Gun

Germany4482 Posts

January 09 2011 02:37 GMT

while i appreciate this thread and agree with most of it, i think there is one fundamental flaw in it: you assume that similar movement speed makes a diverse army stronger. the truth however is that differing unit speeds just increase the amount of control/micro needed, but once the necessary micro is available (pros), the faster a unit is the better.

a good example would be the medivac speed nerf. now they have a much lower tendency to run ahead of the bioball, so mmm balls are easier to move across the map. the unit synergy in terms of movement speed was increased. but it still was clearly a nerf, not only because it made drops much less effective but also because medivacs now have a much harder time fleeing from a lost battle.

the faster the unit, the better. even though stuff like lings + other melee or stalker zealot require more micro to keep formation than e.g. marine marauder.

whatthefat

United States918 Posts

January 09 2011 03:28 GMT

Thanks for the thoughtful responses.

@Greenmachine

I agree, flanking is an excellent way of making sure different components of the army arrive with the right timings, and it's one of the reasons flanking is most used by Zerg.

@bobucles

I think you raise a great point. In many cases, especially against terran, it's beneficial to have roaches in front as the army enters the firing radius of the enemy army, so that they auto-target the roaches. With ideal timing, lings then flood by to reach the enemy army as the roaches attack. A similar technique can be used with banelings busts, by sending the army in the order: ling (or 2), banes, lings. That way, the first ling or 2 will soak up enemy fire, allowing the banelings to hit. A lot of this comes back to control groups, and it surprises me to see that even at the pro level, a lot of zerg players have mixed unit compositions on a single hot key - sometimes even the entire army! This makes it very hard to ensure that roaches tank the first round of enemy fire.

@Black Gun

That's an interesting perspective. I don't think I've said that similar movement speed is just better - as I pointed out, having different unit speeds can in fact be helpful. As you say, a perfect player would simply want all units to have infinite (or at least high) speed so that they could always micro them to exactly where they wanted when they wanted. However, in practice, I don't think even the best pro players are good enough to take advantage of that.

I agree that the medivac speed nerf is overall still a nerf, but it has undoubtedly made 1a-ing with MMM even more viable. In this sense, it's more of a nerf at the pro level (where drops are more common) than the amateur level.