TvZ Broken Beyond Belief Zerg OP to insane levels

General Discussion
1

TvZ is broken beyond belief in favor of zerg.

I have played SC2 since it came out and SC for a decade before that. I have never been a great, high ranking player. In fact I have always been below average. This game is not just an esport, it’s not just for GM and master players, and that is part of it’s success. The game should be designed for almost everyone at all skill levels to have fun with it and find success.

Over the years I slowly but surely progressed up from Bronze, to Silver, to Gold, and in the current game bounce back and forth between platinum and gold. I do fine in TvT, and fine in TvP. Often I will have stretches where my win rates for these matches are 60% or higher. Against zerg my win rate is near 0.

The match ups at their core are uneven. Even watching the best of the best playing via youtube, you constantly hear people making comments like how much work terran has to do in comparison to zerg just to survive.
In order for me to beat a zerg as a platinum player, the zerg has to be silver or lower. A gold T vs a gold Z the zerg will win every time.

Nothing helps. I watch youtube videos. I watch replays. I know exactly what is required of me to even have a chance of winning, and it is next to impossible. The skill levels required for the races and the MMR match ups are wildly uneven. Even stupid lame cheeses, which I prefer not to do, like 3 rax all ins with bunkers in the natural, have a low chance of success since pulling all their workers can almost always stop it. And everything after the first 2 minutes requires terran to do more than their rival and to do it perfectly just to even survive. I’m not a pro level player. I cannot perfectly split marines in a second to avoid countless swarms of banelings. I cannot perfectly micro hellions to harass worker lines and deny all creep spread in order just to barely stay in the game long enough to be attacked another way. I understand the game, I understand the counters, and terran is constantly behind and unable to match zerg. Every response is weaker, every pre emptive attack has a stronger counter.

After playing this game for decades I am at my wits end. I’m in my 30s and my reaction times are going to start dropping off a cliff if they haven’t already. The past few weeks I seem to face 5-10 zerg players in a row for every 1 game of T or Z. And it’s just loss after loss after loss, no matter what I do, no matter what I throw out, no matter how much I dig in. I can’t stress enough before people just say “get good” or “watch replays” or “learn to play” that I try my hardest, I do everything that I can. The game, this match up, is fundamentally broken at most levels of play, certainly at everything in the bottom 50 percent. And it makes me very depressed, as this is my favorite game of all time, and it feels broken beyond repair.

2

lmao, you been playing this game for how long and your only gold, and you cry about your race. How about this before you start crying, go play the other race and see how far you get. you terrans, whine and cry about everything, yet all you gotta do to win a game is turtle defence and outlast every one of your opponents. every one of you terrans say zerg op, yet probably never actually played zerg.

3

Well I feel completely opposite.

Terran has way more options then Zerg this undeniable.

First of all since Terran control tempo of game and forces Zerg to react.

Maybe play bio less counters

4

TvZ Broken Beyond Belief

Correct. Consider the following picture:

https://i.imgur.com/QPKzvcx.png

344 APM vs 257. That’s 87 more clicks per minute or 1.5 per second. Over a 22 minute game, that’s roughly 2,000 more clicks to win.

Terran can just YEET a BC across the map off of 1 base and win. Zerg has to grind out a win over the course of 7,400 actions. One of these is not like the other.

Something needs to be done about the desperate APM requirements to play each race. Zerg’s mechanics are way too hard by comparison to the other races. The ghost nerfs, ultra/hydra buffs absolutely decided this game, by the way. Normally, the terran would’ve won easily as soon as I reached ultras because he’d rapid fire snipe and the game would end. Balance is headed in the right direction, but it will never be fair until the APM requirements are roughly equal. I wouldn’t mind 20 more APM, but 90? That’s just plain insane.

For comparison, the difference between Grandmaster and Masters is only 46 apm. It’s like, give me a break. This is open-and-shut proof that there are huge differences in the effort required to play each race. Why don’t we rework the APM requirements and keep the difference under 1/4th the difference between GM/Masters. That means roughly a 12 apm variation between each race. That sounds like a reasonable goal.

5

One of biggest problem is bc became Zerg mutas. They had buff spores to massive bio damage so Zerg could move out ground army.

Bc is doing same thing plus they have huge advatage they telport snipe key tech and there is nothing Zerg can do about it.

6

At lower levels just play mech va Zerg. Forget bio as splitting against banelings and clearing creep is far too difficult without sacrificing your macro.

Just sit there and make hellions tanks and thors if Zerg goes mutas.

Zerg is OP to the moon, the race needs severe nerfs to queens and creep spread, but they are very bearable at lower levels.

7

Post EAPM or no balls.

8

APM is superior to EPM because it measures the time-cost of rapid-fire actions, while EPM does not. If you think a longer rapid-fire isn’t a higher cost in a real-time game, then you won’t mind if we double the time it takes to rapid fire ghost snipes and barracks production, right?

This is a testable theory. I will make a mod that doubles the rapid fire cost of terran abilities and, according to you, this should not impact your win-rate. Obviously, you are wrong. If you have to spend more time rapid-firing these abilities, it’s less time doing other things. That means less marine splits, fewer mule drops, more supply blocks, etc. A Zerg who has the same MMR as a Terran, but higher APM, is a superior player because he achieved the same performance level despite a higher skill and effort cost. He has less time to do useful tasks in the game because he is stuck spending more time rapid-firing abilities like inject, creep and unit production.

APM is a superior skill metric because it measures this cost, period, end of story. Also, Zergs have higher EPM on average. Zerg is simply a more mechanically difficult race. It’s just a fact. It’s a very unpopular fact that makes a lot of people very irate, but it’s still a fact. A fact doesn’t stop being a fact just because you are mad.

EPM correlates more with decision making. Because it filters out repeat actions, it measures how often you change what action you are doing. It’s not a measure of how many actions you do, but how often you switch from one action to another action. Zergs do more of that, too, but it’s a different metric that I think correlates more with a deliberate decision to change your focus, which is, essentially, decision making.

Protoss are lower in APM and EPM, which means fewer mechanics and less decision making. Don’t even get me started about Protoss. Let’s just say that MMR is half-off with the Protoss coupon.

9

The Tardcode is delusional, do not feed him with attention.

10

You need to have better unit compositions and maintain your upgrade path. WM and BFH should be constantly chipping away at Zergs expansions and drawing attention away from the main engagements which you should be dictating. At these levels (Diamond and below) just a few run-bys is enough to typically discombobulate your opponent. They don’t even have to be serious they just need to be incredibly distracting, thus the Blueflame.

11

I get your point that you should include rapid fired apm but the difference between APM and EAPM from Zerg players is usually extreme. Here’s an example:
h ttps://imgur.com/a/EyvRIu2

12

Exceptions don’t define trends. That’s like saying everyone on Earth has 160 iq because Einstein did.

Division Games Avg APM Avg APM Protoss Avg APM Terran Avg APM Zerg
unranked 133,740 139.52 116.77 135.21 164.56
bronze 69,058 100.15 84.44 95.07 123.71
silver 98,487 82.15 70.24 73.84 101.27
gold 93,523 92.35 77.85 87.08 111.99
platinum 191,058 115.47 96.33 111.63 137.48
diamond 509,306 158.02 131.16 158.35 183.75
master 220,657 206.36 185.85 209.59 244.10
grandmaster 35,091 278.41 251.14 275.69 329.72

The difference between Terran and Zerg in GM for the past year, according to sc2replaystats (n=35,091), is 54. So, when somebody says that inject inflates Zerg’s APM to 500, that’s called a misrepresentation.

Time spent rapid-firing inject, creep, etc, is a real cost that does reduce how much time you have to do other tasks in the game like micro. EPM does not properly represent those costs. People like to point out the times Zergs had 500 APM from a good production wave, but never point out the times when Zergs did not have the APM to micro some banelings which then blow up on a single marauder.

Every unit that Zerg makes in a typical TvZ trades at a negative efficiency, Zerglings and Banelings included. If that’s not cut and dry proof that Zergs don’t have the APM to micro their army, then nothing is. They do have the APM. They just have to spend it elsewhere because Zerg’s mechanics are drastically more difficult than the other two races.

Losing banelings to a marauder sucks, but it sucks less than floating money and not being able to make banelings in the first place.

13

APM is not a good proxy for skill, even though it obviously correlates with skill. Producing a wave of zerglings will burst your short run APM to outrageous levels, something that could never be matched by actual precise baneling micro for example, which highlights the flaws with APM as proxy for skill or speed. The sum of each of those short run APM bursts throught the game will obviously impact average APM.

It doesn’t measure precision in the movements, it measures quantity instead of quality and is very dependant on faction as well as style of play for that faction.

I get much higher APM playing bio compared to mech as Terran, not because bio is harder to play or requires faster actions, but because production itself boosts the APM by about 30-50 from just spamming A and D on the keyboard over the course of the game.

The reason the average APM between the races differ is due to the nature in how they function. Zerg is more about quantity over quality in the APM, you don’t need to be as precise in your clicks as Terran or Protoss, unless you use a specific unit composition like lots of casters.

Certain actions don’t even generate APM, like scrolling/moving the camera around which is clearly an action. This is pretty obvious stuff, but the APM debate has always been about ego boosting and bragging rights.

14

APM has a correlation with win-rate of 0.65, which is the highest of any variable that is separate and distinct from win-rate. Generally, a correlation over 0.05 is proof of a relationship. When you get to correlations as large as 0.65, you can start talking about if the two variables are actually the same thing. APM is the largest aspect of skill, period. I didn’t read the rest of your post.

15

As I said, APM correlates with skill but also correlates with other things. If you don’t take those other things into account you are biased by definition. This is pretty basic stuff, if the same player, from one game to the next, can have over 50 APM fluctuations there are obviously many factors that influence APM. Your skill don’t magically drop from one game to another from my example of playing Mech vs Bio which typically results in a 25-50 APM decrease.

1 Like
16

That’s called sampling variance. The variability of one coin flip is 50%. The variability of the average of 100 coin flips is np(1-p)^0.5 which is (100 * 0.5^2)^0.5 = 5. That means in >99% of cases, the number of heads/tails will vary by no more than 15%. If you continue this trend, the sampling variance goes to 0 as n goes to infinity. This happens very rapidly – generally only n=30 or so is needed. This is called the central limit theorem.

The number of replays in the sample is roughly 35,091. There will be virtually no variability in the averages of these statistics. If you set the sample size to 1, then of course there will be variability because you are measuring not average APM, but the APM of that specific game which are two totally different concepts.

17

Thresher in the other thread already told you to look at the EPM. Maybe some gold guys will listen to you, but players from higher leagues know that when you aclick, you do not get +1 apm, but +1 for each unit. So if terran a-moves 50 marines, he will get 50 apm, but if Zerg with the same supply of lings and banelings a-moves, his apm will be twice as high. that does not mean, that Zerg is harder or doing more, that means the more units you give the command to, the higher is your apm. And Zerg will have the highest as lings and banelings are the only units in the game, which have <1 supply.

18

If you are trying to say APM is an exact proxy of skill, and by skill I would assume we mean winrate, then if you are to model this would have to take into account every other factor that determine APM, unless you assume APM and winrate are perfectly correlated.

This is clelarly false because any idot can just spam his keyboard and get close to 300 APM and still get completely destroyed by someone with 100 APM.

What you are trying to imply is that Zerg players on average higher APM, and thus higher skill. The only way to prove that would be to take a large completely random sample of players, have them play random for a large sample of games, and then measure the variation not just between the individuals but also within the individuals. Even then there are problems, the type of strat being used needs to be taken into account, if every zerg player who plays protoss uses canon rushes to get a high winrate then APM might even be negatively correlated with winrate because this strat doesn’t require much APM

You would also have to assume skill is non-changing during the sample, which may not be true if this is studied for a longer period of time, because your skill changes depending on how much you play in a specific period.

Increasing the sample size will not do anything if your sample is biased. If we consider an alternative universe where every terran player in your sample in GM plays mech instead of bio, and we take exactly the same sample size, there would no doubt be changes in the APM averages for the terran GM players, despite having the same average skill as previously, they are the same players.

Unlike what reality often is, coin flips are an entirely random process with 0 external factors influencing it, thats why this analogy doesn’t work here.

1 Like
19

You’ll likely find zero sympathy on the forums about this.

And for years I’ve mentioned the same issue, that the game isn’t just for the few dozen Korean pros that dominate the professional game.

But nobody cares. Everyone is just a Gold league future GM pro endlessly pushing at the ladder to “git gud”. Few realize that the the game should be fun at all levels or that it should be fun at all because it’s actually serious business to endlessly grind the ladder for the skillz.

Good luck and have fun though.

1 Like
20

And I for years will repeat, that you will never be able to balance the game for all levels. Simple example from team games: Every time I get matched against team with some gold-plat protoss, I always laugh seeing voidray being made, because I already know, that the guy lost a game at the moment he started building one. And after he lost, he will not make any conclusion, he will simply play voidray again. That’s why players stay in gold-plat for years. They play bu#!@hit again and again, even if it doesn’t work, and claim that balance is the problem for them. You can’t make the balance for someone who make planetary on his main, spam marines only and so on.

2 Likes