EPTBlizzard recently posted a community feedback which can be found here
In it they state that they have been receiving feedback from specific korean players that Zerg stands no chance against Terran. They also mention that they are currently trying to get more feedback from a larger pool of players which I think is a reasonable approach. I´ll skip past the discussion of being critical when evaluating sources and credentials this time around. Instead I will focus this article on how TvZ plays out these days and some of the issues there might be for both sides. There have been many heated discussions brought up in the wake of this piece of feedback and it is my honest opinion that a fairly large percentage of the debaters are focusing on the wrong aspects of the game. So I hope this article can help shed some light on some of the underlying issues in the matchup.
Balancing aimed at the correct level of players.
Before having a closer look at the ingame nitty gritty stuff I thought it would be appropriate to discuss the fact that balance can be very different dependent on what level of player you are dealing with. Now there have been many discussions about which level of players blizzard should aim for and it certainly is a tricky issue. There is a huge chasm in skill between world beating all time great players like Maru and the currently highest WCS point ranking non korean terran, Masa. (And I certainly hope Masa takes no offence from this statement as it was obviously not the intention).
So what should the target group be?
It is my personal opinion that since they have developed a large globalized tournament structure in WCS their goal should be to have a balanced game for the players in the highest tier of this system leading up to blizzcon. This includes the premiere circuit events along the with korean leagues. So what that ultimately means is that balance should not be solely restricted to the best 30 players in the world but rather something in the realm of the top 300. We are still talking professional or extremely skilled semi professional level players in all regions but it does put some requirements on the balancing process of the game since you have to be mindful of a reasonably broad spectrum of players.
Assessing the level of meta games across regions.
This is a very delicate topic and I will attempt to tread as lightly as possible on the matter. We have multiple separate regions in Starcraft II with relatively low interaction right now and these regions all have their own meta in which ideas are nurtured, trialed and tested. Korea has always been the best, I mean its a country that is considered a region on the same premises as Europe and North America. If you were to make a list of the top 10 players of all time they would all be korean and they would all be better than any foreigner by many orders of magnitude. But with all that said there have been times when foreign players have been on the same level as koreans, and most notably the Zergs. It is my personal opinion that even right now if we exclude the category of the tier 1.0 koreans with players such as Maru, Zest and Dark, the elite of the foreigner player base is able to play ball in the same league as the koreans. The image below shows the aligulac ELO standings and while I perfectly understand something like this is only partial information it´s still a part of the bigger picture and thus interesting to include because it shows that the top of the foreigner scene is still a competitive entity.
![[image loading]](http://i.imgur.com/Bedp16H.png)
A few foreigners sneaking their way into the statistics.
Even if you don´t quite share this sentiment it´s still worth remembering that in Starcraft II you can win with a multitude of skill sets. We have had many champions over the years who won with different strengths. MvP was a king of all trades and planning series. MC was a boss of controlling finely tuned aggression. Life was the epitome of small unit skirmishing and counter attacks. Taeja was the sage of brilliant decision making in all stages of the game. Squirtle and his execution bot Parting showed what the beautiful combination of extremely clever builds and real top3 control looked like and innovation was basically AI level APM mechanics put into a human. So with the understanding that there are different ways to be a successful player we can take a closer look at the foreign Zergs in the middle of 2016. Their primary strength is their understanding of strategy and their knowledge of what exactly Zerg is extremely good at and how to abuse it to the maximum level. Even historically it´s no secret that foreign Zergs developed a much stronger grasp on how to play the lategame situations even if their mechanics, micro and crisis management sometimes fell short of their korean counterparts. Koreans might come out on top in the end, but Starcraft II is a game where you can win doing the wrong thing if you´re just a better player overall. So I do believe that you certainly should not discredit the lessons that come out of the metas from other regions than korea and particuarly the Zerg ones as they certainly have the performances, achievements and history to support their legitimacy.
Seperating player skill from balance
A minor point in the discussion of balance is what the current roster of players for each race looks like compared to how they perform. Is a race over- or under- performing due to player or balance reasons? My general philosophy on this topic has always been that racial balance is best reflected at the preliminary and early stages of tournaments. That would mean something like the old Code B, Code A except for finals, Code S ro32, Dreamhack and IEM group stages. During this time many players of somewhat even skill are competing across a reasonably large sample size of games and should provide a decent base of statistics to determine how the balance looks fairly disconnected from which players are playing. The images below shows the racial distribution from the initial stages the offline premiere tournaments over the last few months.
![[image loading]](http://i.imgur.com/u8wZZq6.png)
SSL S2 Challenge
![[image loading]](http://i.imgur.com/QU63Vx6.png)
GSL S2 Code A
![[image loading]](http://i.imgur.com/yCokvlN.png)
Dreamhack Tours (WCS Spring Circuit Championship)
![[image loading]](http://i.imgur.com/qVG4vz2.png)
Dreamhack Austin
I think it´s evenly enough distributed that it´s not a serious cause of concern. Zerg is slightly underrepresented in SSL but on par in GSL and fairly overrepresented at dreamhacks while Terran is slightly underrepresented at dreamhacks and protoss on par across the board. It is a small sample size and you could easily include things like pro league and online cups in your statistics to get a more complete picture but doing so brings in other variables such as map selection systems, match formats, player ping and so on which adds other uncertainty factors to the equation. Cherry picking or nitpicking in statistics is a dangerous act and these images are not meant to support the claim that "everything is fine guys" nor are they meant to incite the opposite. The point I want to make with them is that if there are claims that a significant imbalance exists one would need a much more comprehensive analysis and statistical work than what is most often presented.
The players matter
Now in the later stages of tournaments it´s much more about the individual players and we have continuously seen a very small number of players massively skew statistics throughout the games history. Take MVP out of your stats for the later half of WoL and the championship record for terran is abysmal. Discount MC and protoss championships in the first 14 seasons of GSL doesn´t even exist. Look at zergs premeire wins without life during the middle and later parts of HoTS and that also becomes quite the lackluster record. Champions are exactly that, champions. And they will always fuck with statistics. So whenever balance discussions are based off of the top 8 of Code S or SSL it is a ridiculous enterprise from the very beginning. You are trying to gauge which races performs the best but you are dealing with literally 8 of the best players on planet earth and it just so happens that one or two more of them happens to play one particular race rather than the others. As an illustration of such a scenario the picture below is the previous code S top 8. The better player won in pretty much every match regardless of race with the only debatable one being herO vs Cure.
![[image loading]](http://i.imgur.com/FA1YrfU.png)
The previous Code S top 8.
So understanding the fact that individual players can dominate results we can now make a rough guess at how each region look in terms of it´s players isolated from the wrong types of statistics. I will preface this segment with the statement that I´m mentioning players by name to give credit and certainly not to take away credit from anyone not mentioned and I´m very much open to the fact that I could be mistaken on some accounts.
Korea
Starting from the top my impression is that korea is primarily driven by Terran and Protoss at the very top right now. You can easily ramble off a few names like Maru, TY, Zest, Classic, $o$ and Stats for the two races while Zerg really only has one household name in the Tier 1.0 fraternity and that would be Dark. From there it´s downhill pretty fast. As it turns out having your best player for the race lose his way in life has a fairly detrimental effect on the statistics.
Europe
In EU it´s a little different. Zerg is the clear forerunner in terms of its players with Protoss somewhere in the middle and Terran a remote pipe dream. Nerchio has a pretty legit claim to being the best foreigner right now with other players like Snute, Elazer and perhaps Bly closing up the ranks. Showtime has been the best EU Protoss for what seems to be ages and I think there is a small gap down to the rest. Terran is limbing far behind and I´m actually fairly unsure about who I would rank first. I don´t think there is any player on the same level as players like Nerchio or Showtime but I think if someone put a gun to my head I would say uThermal, but hes just a streamer right. I will say that some of the other Millenium players would certainly be on my short list normally but they have been a little MIA recently so it´s hard to tell where they are at.
North America
For NA Neeb is kinda driving the bus and everyone else is just along for the ride. I don´t think there is any other player on the level of Neeb right now, despite there being a pretty good selection of players in the tier below. Now I will say Neeb looks extremely good even in an international context and he is certainly someone I would recommend to study.
I didn´t include the Chinese region simply because I am fairly uninformed and I don´t want to write about something I really don´t know much about.
The overall picture
The general representation between the 3 races seems to be fairly even at the initial stages of premiere tournaments and we haven´t seen a massive skew in one direction or the other for a prolonged period of time. Ultimately it comes down to which players are currently in hot form when deciding how the top of each region looks. Any non-revolutionary patch is unlikely to change this and I strongly believe being critical of statistics in this particular instance is important when deciding on applying balance patches to the game. Next we´ll have a look at what TvZ looks like and keep in mind that the metas are different between the regions and this time around foreigners might have it better figured out.
TvZ, The Matchup
So let´s dive into having a look at TvZ. A very small and simple topic that is easily covered in a few written words without any inaccuracies by one person. I think the easiest way to start out when trying to understand a matchup is to look at what the common perception is and then put it into words. I have tried to phrase my understanding of TvZ as the following:
"TvZ is a matchup in which Zerg has the faster growing economy if left unchecked but Terran has a wide variety of harassment options. Both races have a good array of variations for their strategies and army compositions but the Terran early and midgame armies tend to scale better than the Zerg ones. However Zerg holds a significant advantage in the lategame with the exception of large straight up army battles fought in small controlled spaces."
So here is when it gets difficult. A game of Starcraft II is a web of cause and effect and discussing any one point in isolation is extremely difficult and I urge readers to keep the phrase above in mind because whenever a point is brought up you have to look at how it fits into the overarching way the matchup functions in its current form. I wrote and rewrote this segment a couple of times because the ordering of things was difficult and in the end I realized I might as well start with the most defining factor in the matchup right now rather than following my definition of balance in the matchup chronologically as I had intended originally.
The lategame
Zerg lategame holds the advantage because its core units in cracklings and ultralisk combined with high map vision from creep and overlords is extremly mobile, easy to split on multiple fronts, easy to control and suffers minimal negative effects if the player is caught off guard. That core is backed up by a good array of siege units in brood lords and supporting spellcasters in vipers and infestors to make a very strong deathball if required to do so. On top of that the production system that supports it also has the ability to remax at a rate way beyond the Terran capabilities. So how does it compare to Terran and why is it so favoured? The major advantage Terran has in the lategame is that if you can get into a fight with a supermaxed lategame army in a single position facing a limited number of approaching angles, it will beat pretty much anything Zerg has in the first wave. Something along the lines of 20 ghost, 12 libs, 6 vikings, 2 ravens, 6 medviacs and decent ball of supporting bio with a few mines or tanks. The fact that Terran can operate on a lower number of workers with mules certainly help but with the pace of LoTV having a large number of orbitals isn´t nearly as commonplace as in the previous expansions. The problem with this strength of Terran is that the army is a composition based deathball and it splits Extremely poorly to fight on multiple fronts both in terms of the actual units as well as player ability and attention. Cracklings and ultralisk have reasonably low attention requirements relative to bio ghost and if no attention or micro is involved it is often better for the Zerg than the Terran. This is a large contribution factor as to why fighting on multiple fronts is not a very prosperous enterprise for Terran at this stage in the game. As a Terran player you can get into scenarios where a nydus starts in the main, there is a ling runby into the third while a decent group of ultralisk and corruptors are waiting outside the 4th but can go to the 5th base in a fairly short amount of time as well. If you were defending with the previous mentioned army, what do you do? I want you to imagine having to unsiege half your liberators, unsiege half your tanks or mines, split your bio and ghost control groups in two maybe even three, send everything to different locations, deploy both libs mines and tanks while making sure your ghost are not exposed anywhere. All this in time spans often limited to seconds and what you really end up with is an impossible scenario. On top of that the Zerg core units in the lategame in ultras cracklings and corruptors are all really good at killing command centers so the Terran bases will die very fast if attacked and sometimes even with a considerable amount of defense present.
"Liberators are a bit like trenches in World War II while Ultralisk are the Panzers. If they break through anywhere all your stuff in the rear is gone."
The way Zerg production works ensures that if the Zerg had the necessary money banked they can rebuild their army to almost full strength in the time that it takes the Terran to regroup the army after mopping up the attack and then push across the map presumably with a significantly weakened economy. Terran does have some options to do similar moves on a smaller scale with drops and liberator harass but a very important thing to understand about the Terran army at this point is that it is a deathball and it needs all its components present to function. Terran can´t just send away large portions of bio forces to drop or liberators to harass because then the main army wont have the right mix to fight the Zerg. So to sum it up. If a Terran can get into a scenario where they can grind huge lategame army battles with a healthy economy they are probably going to win. On maps like dusk towers where you can hold a ridiculous amount of bases from one or two locations I believe Terran is favoured in the lategame, but on any map where you have to extend much further out to defend your later bases it is almost impossible to win against any competent Zerg. Yes Terran can make a very strong army but the army they are building can´t realistically defend the high number of bases that is required to support it.
A final note on this topic is the strength of mass liberators. They were potentially broken before they got hit hard with the nerf bat but these days I have not seen a single game where it looked remotely too strong. I have seen games of people trying to mass liberators rather than adding ghost to support the bio army and most notably from korean Terrans. But honestly I have yet to see it work out favorably with credit to the liberators. Ravagers and corruptors with potential vipers do a tremendous job at thinning out the liberator ranks and if they are your only answer to ultralisk it won´t end too well for you. As a funny reference I encourage you to watch game 5 of INnoVation vs Dark from the GSL Code A Season 2 2016 where INoVation dies while maxed with a very high number of liberators through the tiniest choke imaginable to really support the notion that liberators alone are not the solution.
![[image loading]](http://i.imgur.com/g1wSOh1.png)
This isn´t going to end well.
The game plans
So with this understanding of why Terran wants to avoid playing the lategame on most maps we can easily see why the Terran game plan in the current state of the game is to try and win before the late game and why Zergs are trying to get there. The major strategies for both sides can probably be simplified into a few major branches. I´ll talk about them briefly but keep in mind we are examining the unit compositions used to carry out the mentioned game plans and not what kind of army we´re going to deploy in the lategame.
Zerg
For Zerg they have the option to either play a minimal midgame army into fast hive strategy or large midgame army into slow hive strategy. It´s entirely possible to tweak the supporting economy that goes along with these strategies but an important factor to keep in mind is that once hive is unlocked it is very likely that Zerg can rebuild their economy to a decent state if they didn´t take critical damage. The reason for this is that the Terran army needs a long time to transition into something that can deal with hive and during this time is not a composition that can attack onto creep. When it comes to unit compositions Zergs can play with either ling bling muta or roach ravager. I´ll add that there are some styles which plays both ling bling and roach ravager combined but it leans heavily towards the heavy unit slow hive type of strategy. I´ll address the unit compositions a little later on in the post and especially ling bling muta since there have been a lot of discussions about it in relation to the talk about buffing larva injection.
Terran
On the Terran side of things the two main branches of strategies are heavy 3 base all ins and 4 base macro with perhaps some honorable mention to the experiments with sky Terran. However due to the recent nerf to liberator anti-air-armor capabilities I consider this a dead end. The terran army compositions can be categorized as bio mine and bio tank. None of the terran midgame armies should include liberators for anything other than harrasment.
The circular balance of compositions
The different game plans and composition for each race matches up differently against each other and I have tried to write them in a list form with the understanding that there are many factors subject to change during a game and these are only overall guidelines.
- Bio mine 3 base is good against both slow and fast hive ling bling muta.
- Bio mine is bad against both versions of roach ravager.
- Bio tank is good against slow hive roach ravager.
- Bio tank is bad against fast hive roach ravager and both versions of ling bling muta.
There is some consideration for 3 base pure bio being able to out muscle roach ravager but I consider it a little more of a fringe scenario so we´ll just leave it here as an after thought. So what a clever reader might notice is that I didn´t list any good Terran compositions against a fast hive roach ravager and that is precisely what a lot of Zergs are trying to exploit. Now often mistakes happens and people lose but it certainly has been one of the better strategies as of late. Understanding that each unit composition has it´s good and bad matchups is a large part about playing TvZ at the moment. The hard part then becomes figuring out which set of units your opponent is building and designing builds that can adjust in time accordingly to give you the upper hand. What your opponent is doing matters and I hope readers will keep this in mind when they next discuss which unit composition to go for. In the scenario that the Terran has chosen a favorable unit composition it will almost universally outperform the Zerg army over time. The map being played often has a huge influence on this relation of power and on some maps Terran aggression is certainly very powerful and perhaps too good while on other maps Zerg sometimes defends without too much trouble.
The muta ling bling issue
The viability of this unit compositions seems to have been a hot topic on the forums recently and it mostly came about when blizzard talked about buffing the larva inject from 3 to 4 larva per cycle. I want to talk a little about why muta ling bling isn´t as powerful as it once was because I think people either forgot or simply don´t understand. I was asked earlier today. "Do you think Dark is making roaches because of the larva issue to go with his ling bling?". If you consider the fact that the price of a roach warren and speed upgrade is very similar to simply getting another hatch and a queen to put you on parity with previous versions of Zerg larva it suddenly seems very hard to argue that larva was his concern. It´s not about the larva in LoTV when it comes to ling bling muta. Now buffing larva might change all sorts of timings and the overall strength of Zerg that much is true, but it is not the driving factor in bringing back muta ling bling. So why isn´t muta ling bling that good anymore if it´s not about the larva? There are two main reasons which both deals with what the primary win condition in HoTS was for Zerg using that composition. The two primary ways to win was:
- Overwhelming creep spread.
- 25-30+ mutas with upgrades.
The LoTV creep was massively nerfed compared to it´s HoTS counterpart and it´s extremely rare you ever see it go out of control the way it could in HoTS. It gave the Zerg a tremendous amount of map control, time to react to attack and the ability to take favorable engagements. The LoTV version of creep means that Zerg often has to fight sooner in worse positions and that they lose their iron grasp on the game much easier than before. On top of that Terran now has the option to build liberators to demolish the muta cloud. I had someone ask me if thors didn´t already fulfill the counter role that liberators do and they partially did except liberators are much more mobile and can act as a chasing entity rather than a static deterrent like the thor. If you combine these two factors it´s really not that hard to imagine why muta ling bling struggles as a stand alone strategy It is now merely a stepping stone towards hive tech in a very different way from how it was previously. Obviously having more larva and thus more units would make it stronger but inherently the fundamental win conditions of the strategy are no longer there and that is why it will never fully return to glory unless there is a major overhaul of the game.
![[image loading]](http://i.imgur.com/kVhB6if.png)
Fucking Namshar in HoTS, I never won.
Economy and harrasment
I wrote in my phrase to define the balance for the matchup that the growth rate of an unchallenged Zerg economy was better than the Terran one. Changing the properties of either the Zerg economy or the harassment tools to keep it honest is a very dangerous way to balance the current TvZ. The reason for this is that TvZ is dangerously close to a binary matchup where you either get to ultralisk or you don´t. Tweaking the relative power of the early game or midgame for either race can easily sway the scales in the favor of one of the sides. If you buff Zerg during these stages of the game it might remove the timing windows Terran has to close out the game and lead to a very lopsided affair but on the other hand if Zerg is truly too weak early on and not enough players can survive long enough perhaps something does have to change. I will say from a personal point of view I find the "Don´t let them get there" situations to be poor game design when we aren´t talking about supermaxed capital fleets and things like that. I wish we didn´t have those situations since it´s never particularly fun for either side.
The ultralisk moment
It really is no secret at this point that the appearance of ultralisk on the field has abruptly ended it´s fair share of games. The changes LoTV brought about turned ultralisks into ULTRALISKS and it´s kinda hilarious that the giant hybrids in the campaign die faster than ultras do in 1v1. The core issue with them when talking balance is that they sway the scale of power in the middle of a game by such tremendous amounts so fast that it´s very difficult to deal with. Every single unit in the core of the early and midgame Terran army is downright garbage against ultralisk. But at the same time the units that deals well with ultralisk (ghosts and liberators) are next to useless against the early and midgame armies that Zergs field. Liberators have minor applications as harassment units but they take away from medivac production which is a significant trade off. It puts Terran players in a bind because they have to choose whether they want to attempt to kill the Zerg before ultras and be very screwed if they ever show up, or be able to fight the ultras when they arrive but have very slim to no chances to end the game early and thus end up playing an unfavorable lategame. Most players choose to try and end the game before ultras and that is where we are today.The pictures posted below are more funny than actual material for balance arguments since there certainly were mistakes on both sides.
![[image loading]](http://i.imgur.com/5zvfUG5.jpg)
Just when you think you have them cornered.
![[image loading]](http://i.imgur.com/JDYRBJZ.jpg)
Nope.
The Zerg approach
From the Zerg side of things the challenge then becomes learning how to survive the different types of Terran aggression while getting ultras out in a timely fashion. However if we generalize a fair amount there seems to be a strategical difference in how koreans approach the matter compared to foreigners. The gameplan for koreans could be formulated as:
"Create and maintain a large midgame focussed army and transition into hive and supporting ultralisk when it is relatively safe to do so".
It is very much the way Zerg was played during most of HoTS. Hive was the closer, the deathblow but never the main swing of the hammer. The balance in this scenario is focused on how the midgame armies compare and has little concern about how hive units play since the Zerg should already be in a winning position by that point. It is a strategy that puts emphasis on mechanics and is willing to try and go muscle to muscle with the Terran forces. For foreigners the gameplan is very different and I would phrase it as:
"Invest the minimal amount possible in the early and midgame army and instead put the resources into attempting to get to hive before the main Terran offense deals critical damage."
Fundamentally it is a different way of playing because the hive units suddenly becomes the core of the army with support from the early and midgame units. The question of balance in this scenario revolves around whether or not it´s possible to get out ultralisk before the Terran can kill you in a reliable and consistent fashion. Which gameplan is better is hard to tell and is still being explored. I´m of the personal opinion that trying to get out ultralisk before the Terran can kill you is the better approach. It attempts to circumvent a clear weakness of Zerg which is the efficiency of the lower tier armies relative to the Terran and instead aims to exploit the areas in which Zerg holds the advantage. An interesting thought is whether or not the korean gameplan was developed and refined as a result of not being able to successfully get out ultralisk before dying against very top tier Terran execution. There are examples of fast hive being pulled off in korea but not as often as in the foreign scene.
Final thoughts
At the end of the day people will ask "Is TvZ balanced?". If I had to make an educated guess I would lean towards yes. The player representation and winrates for the most important tournament system in SC2 appears to support this. And while Zergs are slightly underperforming in korea I think in this particular case it´s more of a player issue rather than a balance one. Zergs from other regions are performing quite well in contrast to korea and as I wrote earlier I believe the top of the scene is driven by individual players rather than their race. The korean Zergs are certainly not weaker than the foreigners in terms of mechanics, micro or ingame decision making but perhaps this time around their choices of strategy are not always what they could have been. Now it is entirely possible that korean Terrans are able to execute on a level that invalidates some of the strategies that foreign Zergs are using but it´s not something we can be sure about until we have more source material of korean Zergs trying to do it. To close out the discussion I will also echo my previous sentiment by saying that I believe the balancing of the game should encompass every region involved in WCS and breaking the balance of some regions should be considered a reasonable no-go. The binary nature of TvZ means that you have to be careful when discussing balance that targets either the strength of the Terran ability to close out the game or the power of Zerg once hive units are deployed. If you address these concerns in isolation from one another you´re probably going to be very hard pressed to find the right solution. Balance changes should address both of these areas in tandem and hopefully lead to a better experience for everyone involved.
So in the end I will conclude this post with the statement that I believe TvZ is pretty balanced but it sure as hell is not a very fun matchup as it currently stands.
