Gosu.com is currently inactive. All content on the site will remain for archiving purposes, but no new content will be added for the foreseeable future. For the weekly podcast and new material from many of our old contributors, check out TiSBcast.com.
So long, and thanks for all the fish!
I recently caught an excellent GSL Super Tournament match between the ever threatening TSL_aLive and the undeservingly Code B LeenockfOu. Alive opens with some scary looking 2rax aggression which Leenock thwarts with expert composure. From there, they game normalizes to your every day marine/tank vs muta/ling/bane. Leenock, however, utilizes a few subtle curve balls that differentiate him from your typical Zerg. Chief among these, he elected to research +1 flyer armor for his mutalisks very quickly instead of the more standard +1 attack.
Leenock’s choice was more than a snap decision. It factored in to the long term army composition and resource distribution he’d determined would be most effective against what he felt he’d be facing. This got me thinking about the various factors that contribute to deciding when to upgrade what and why. Knowledge of these factors isn’t quite as foundational as say, understanding good macro, but it can provide the slight edge you need against an otherwise equally skilled opponent.
Answering the above question is far more complex that it initially seems. It depends largely on the answer to a series of smaller questions, and weighing their overall importance against one and other.
On a supremely basic level, knowing which upgrades are most effective for which units feels like it should be simple. Units with higher attack speed or double attacks, for instance, reap the benefit of an attack upgrade considerably more than units for focused on slower burst damage. It’s a percentage yield of DPS.
Let’s do some math. A marine has an initial damage of 6 and fires at a rate of .86 seconds, making its initial DPS roughly 6.97 ((6 * 100)/86). The +1 upgrade raises that damage from 6 to 7 (an increase of ~16.67%), changing DPS to roughly 8.14. That’s a pretty substantial jump for just 100 min and 100 gas. Of course, the % yields decrease from 1 to 2 and 2 to 3, as is the case for all units.
Forgetting about bonuses against armored, let’s compare that to the slower, heavier marauder. Marauders do an initial 10 dmg and fire at rate of once per 1.5 seconds, making their initial DPS a mere 6.67. Their +1 wep bonus yields a mere 10% bonus, increasing their DPS to 7.34.
Based on this, it’s apparent that unit compositions with higher attack speed and lower initial damage will gain far more benefit from upgrading attack quickly than their more burst-damage oriented counterparts. In a vacuum, if you know you’re going to be building lots of Marines, Hydras, Phoenixes, Zealots, Zerglings, etc. then you want to consider upgrading attack first and fast.
Of course, things in SC2 are often far from that simple. A multitude of situational factors can modify that seemingly straightforward rule. The next questions will address some of those.
Let’s say your opponent is Terran and you scout 4 freshly build barracks, each equipped with a reactor. You know he/she is about to run a metric ton of marines down your throat. As effective as the attack upgrades discussed above are, armor upgrades cancel them out just as well. Likewise, Marines have a considerably low amount of health and a propensity to clump, making them more subsceptable to large, bursting AoE damage (siege tanks, fungals, storms, or banelings) instead of lower level potshots (zerglings or other marines). This is compounded by the presence of medivacs, which are far less effective against large explosive damage and far more effective at mitigating consistent damage over time.
As such, upgrading armor against marines is almost guaranteed to yield more of a benefit than attack. Each point of armor is initially a ~16.67% reduction in overall marine DPS, and even if the marines are at +3 it’s still a reduction of over 10%. This is the logic that likely lead to Leenock’s decision vs Alive.
Of course, there are also situations where what your opponent is producing should prompt you to quickly grab the +1 attack. For instance, large quantities of zerglings. Many units are built specifically so that an initial attack upgrade is absolutely devastating against lings, primarily the zealot and the roach.
A zealot has an attack damage of (2x)8 every 1.2 seconds, while a roach has an attack damage of 16 every 2 seconds. Initially, it takes 3 full attacks (48 damage) for each of these units to kill 0/0 zerglings (35 total hp), which takes a zealot 3.6 seconds and a roach a full 6 seconds. Grabbing the +1 attack upgrade allows zealots and roaches to kill lings in 2 attacks (36 damage each), effectively improving their combat efficiency by a whopping 33%. Keeping these units 1 ahead of zerglings in upgrades yields one of the bigger advantages you can obtain in the game.The reverse is true for the Zerg player, keeping your armor ahead of your opponent’s attack upgrades is downright essential if you want to effectively use lings against roaches or zealots.
Here are a few other useful tidbits on upgrade matchups:
If anyone else discovers any other good ones, please feel free to post them in the comments.
This final question is a bit more abstract, but is perhaps the most important of all:
There are plenty of different ways to play SC2: active vs passive, harassment vs deathball, offensive vs defensive, all in vs macro, etc. Each one of these styles can be enhanced by specific usage of upgrades. For example, let’s say you like to drop and drop a lot, doing quick damage and withdrawing before your opponent can suitably crush your hit squad. In that case, attack upgrades are your best friend. You want to kill off important tech structures and workers as quickly as humanly possible and, ideally, not do much of any direct fighting until you’ve already won yourself a substantial advantage.
Conversely, let’s say you are a passive macro player who likes to slowly take the map while building up a larger and larger force. You may want to focus more on armor to give your units maximum lasting power until you’ve reached critical mass. Then you throw in a quick +1 attack before you roll your opponent with a maxed army.
The overall point is that while a lot of what’s outlined in questions 1 and 2 seems clear cut, your personal play style and overall game plan creates a grey area which may influence your upgrading priorities. Before you do the math, make sure the math will WORK given what you want to accomplish. A pack of attack upgraded roaches may be great against zerglings, but it isn’t going to benefit you much if your opponent is using their lings in a series of small hit and runs and your roaches can’t keep up. You have to know you’re going to play in such a way that FORCES your opponent to submit to your better upgrades. That’s what a “plus one timing push” is all about.
Breaking down the full extent of how upgrades affect a given game scenario is a pretty epic process. There are a number of factors, including shields, bonus damage, AoE, etc. which I didn’t even begin to touch upon but are all valid and important. Please take the time to consider them when you start looking at ways in which your play could be enhanced by upgrade prioritization and optimization. If you discover anything I haven’t been able to touch upon, I do encourage you to post a comment.
EDIT: A Redditor pointed me to this great TL thread as an additional resource.