some micro can be fun. sure.
Here’s what makes Terran not fun: Terrain requires significantly more micro than Protoss. So while I’m pushing my body to the extreme to micro Terran, my Protoss opponents doesn’t even have to look at his army.
That’s the NOT fun part.
When the other guy has 200 apm but the same macro.
200 APM in plat? The only way to achieve it is apm spam. APM spam means instead of doing something useful, they waste time on that thing, better for you, or it is a Smurf, but there is nothing you can do about it then.
Why are keep bringing terran? I understand that you are biased terran, I have understood that a few years ago, just relax and continue playing protoss.
Terran has stimpack, which gives unlimited potential for micro, which results in the problem that some players are so good with it, that it impacts the balance and as a result forces all Terran to be good with it, but you can always stick with protoss.
There are games with no micro and no macro, it is called chess. Poor strategy, nothing else. And you can still blame the opponent, because he played white.
Your argument is based on faulty assumptions built upon a false sense of superiority.
My assumption are build on self-experience, because I was once in silver, gold, plat, diamond too.
Indeed.
This equivocation is absurd. We know, for a fact, that doing actions in a game will be causal with winning the game. How quickly you can do the actions is a measure of how well you know the game because you don’t have to stop and think about what the right answers are. Because the game is real time, having to stop and think is a catastrophic outcome because the events progress without your input. We know, for a fact, that the rate at which actions are done is also causal with skill and it is absurd to suggest otherwise.
To suggest otherwise is a complete and total rejection of the entire game genre at the most fundamental level.
Based on your hypothesis, we could win games without doing actions, or we could gimp Serral’s reaction speed by 10 seconds and he’d still win WCS. If you honestly believe that, you are delusional. Actions & the rate of actions is obviously and irrefutably causally linked with skill. The correlation between APM gives us a measure of how much it is, and it’s the majority of skill (p=0.65). There is no way way anyone can believe what you are saying, and that means you are trolling. This stuff is simply far too absurd that anyone could believe it. We’re talking “lizard people control the government” levels of nonsense.
Setting aside the above absurdities, I will now address the second major absurdity in your argument which is that skill is more complex than APM. APM is a lot like IQ, believe it or not. IQ can have many various sub-traits. If you have two people with the same IQ score, do the underlying mechanics of their brains work in exactly the same manner? No. IQ is an aggregate score that measures any and all traits that could impact cognitive performance. It’s an amalgamation. Saying that there could be more to APM is a fundamental rejection of what APM is, because APM is an amalgamation of traits by definition. APM is a measure of any and all traits which allow you to react to game events faster than the events progress in real time (whatever those traits may be). Literally every trait of skill in the entire game has to funnel through APM because you can’t express any trait of skill without doing actions in the game.
You could have x-ray vision. You could have maphacks built into your brain. You could have alphastar wired into your brain through a chip. You could be able to peer through space time and see your opponent’s screen and read his mind. None of these traits would be worth anything without the ability to enter actions into the game more rapidly than game events progress.
Furthermore, how rapidly game events progress is a function of how rapidly your opponent plays. Being a great strategist does not matter. Being a fast strategist matters. The speed at which you can express those traits is integrally linked with how powerful those traits are, and it is absurd to say otherwise.
Everyone who switches to Zerg will instantly have more APM. It’s easy to get high APM as Zerg.
I’ve proven to your face that Zerg isn’t the hardest, but you always ignore it for some reason.
Yes, the game forces players to spend more time on various rapid fire mechanics (creep, inject, etc) which is equivalent to having less time to focus on other aspects of the game. That’s why APM is preferable over EPM – APM properly represents the time-cost of rapid-fire mechanics. Having less time to focus on the other aspects of the game is synonymous with meaning those aspects of the game are harder. Translation, Zerg makes the game harder and the increased APM cost reflects that.
If you disagree, we can easily create a mod that increases Terran’s APM-costs and see how it affects your win-rate. To maximize the stress, lets increase the number of clicks to make a marine from 1 to 10. Now you have to spend ten times more time to make each marine. Anyone with a brain knows that that will hurt your ability to win. A higher APM cost is equivalent to playing a harder game.
So, if APM = Skill, Is my Zerg more skillful than my Terran?
For the same workload, you will make more errors in other domains. Due to more focus on rapid-fire mechanics (like creep and inject), you have less time to focus on the other tasks like micro. With more errors, you must offset with either a lower performance rating or better play.
It’s not rocket science. Unit micro + inject + creep is much harder than unit micro alone.
Terran is provably the most difficult. PROVABLY.
I’m going to blow your mind. You saying your race is harder doesn’t actually mean it’s harder
“The correlation between APM gives us a measure of how much it is, and it’s the majority of skill (p=0.65). There is no way way anyone can believe what you are saying, and that means you are trolling. This stuff is simply far too absurd that anyone could believe it.”
I do believe it is you who is the one trolling at this point. Not sure what your correlation coefficient even means, because you don’t explain it.
It would imply a 1 unit increase in skill corresponds to a 0.65 unit increase in APM. Now, what measure of skill is even used here? Skill can’t be measured, and this is not winrate as proxy because then the correlation makes no sense. Correlation measures variance and covariance between two variables, thus you need data points of both the mean value and each individual value in the sample, and you need this for both variables.
Unless you have access to all data you can’t do this, or if it’s reported somewhere it needs clarification on how the skill variable was measured.
APM and win-rate are correlated at p=0.65. Given APM is race-stratified, a multivariate correlation would be much higher, ergo the majority of the correlation not accounted for by APM can be accounted for by T/P/Z. Translation, APM is the majority of the skill in the game, and different races require differing amounts of skill to play.
That’s a straight up lie. All the statistics prove otherwise. These trends are so blatant they even showed up in the AlphaStar paper. Their Zerg agent had the most APM and the lowest performance. Are you going to accuse Deepmind of rigging their AlphaStar agent to be biased for Zerg? Lmao. The arguments on these forums are 
arguments.
By the way, it has been rather interesting to see how the PvZ meta has evolved in exactly the manner that AlphaStar predicted. All AlphaStar did was run around with stalkers and disruptors, and that’s exactly what top Protoss are doing now. I saw MaxPax beat Dark the other day using exactly that. This build has been so oppressive they had to nerf the Disruptor for crying out loud and Protoss are still doing it.
But, you are telling me that AlphaStar’s performance isn’t correlated to balance even though its performance mirrors ladder/tournament performance almost perfectly and its builds mirror the meta? That’s legitimately insane. AlphaStar honed in on the strongest builds it could use, and the strongest were with Protoss and the weakest were from Zerg.
It’s also worth noting that the Zerg agents simply allined off of hatchery tech. Yep, Zerg is so powerful that AlphaStar decided the best build was to allin off of hatchery tech with ravagers and queens. The bot that could perfectly micro 100 blink stalkers decided lair tech is literal garbage 
. That’s obvious to anyone with a brain. I was listening to a PiG cast the other day. He said pro players aren’t scared of the new hydra. The infestor and swarm hosts are trash, and mutalisks are a waste of money. They are thinking the new patch is going to be a Protoss patch. I don’t agree with that part, but he’s absolutely right that Zerg lair tech units aren’t worth making. Zerg makes mass tier 1 in all matchups. In PvZ it’s roach ravager ling bane, in TvZ it’s the same or just ling bane, and in ZvZ it’s just roach ravager. The only two reasons you go to lair tech is to get a couple upgrades that make your tier 1 units a little better and to get to tier 3. Zerg tier 2 is literal garbage.
AlphaStar is right in so many ways it’s frankly uncanny.
Another interesting thing to note is that ling/bane is the most resource inefficient unit comp in the game. They trade at a drastic negative efficiency. So, zerg players prefer a unit comp that is literal garbage over lair tech units which are even worse by comparison. That tells you everything you need to know about just how aweful lair tech units are.
The statistics that show that Terran players have the lowest average MMR? Despite playing more games and having more veterancy?
Keep chirping away with your stupid APM argument like a mentally stunted parrot, though.
You are absolutely delusional
8 world championships in a row. Soon to be 9 at Katowice and Zvz final guaranteed.
Ramsey theorem: with enough data, you can find any pattern to any level of precision. For example, you can draw a star with Woolsworth stores in the UK. Is it a CONSPIRACY? No, it’s basic math.
The only statistic that matters is the average, and it only matters when certain conditions are met. Those conditions are not met for stars drawn with Woolsworth stores, nor are they met for the finals of premier tournaments:
You’ve violated #1. The sample is not collected at random. You’ve picked Premier finals when they encompass 0.000000000001% of games played. You violated #2 because Premier finals are drastically over-represented by at least many tens of thousands of times above their actual incidence. You’ve violated #3 since you only have 9 events. You’ve violated #4 as well because there is a 0.0% chance of reaching this criteria if you can’t first meet #3.
It’s a clown argument to claim that premier event finals are correlated with balance. The fact that none of the other forum members have pilloried these exact arguments is proof that the people here are statistically illiterate.
Because i’m bored I will do an example to see.
If we look at a single game (correlation is almost always single observations), winrate is a binary variable, you either win or lose, so 1 or 0 percent winrate. Correlation requires variance and covariance, in order to calculate variance for winrate you take each observation, compare it to the mean and the square it to get the rid of negative values, then you do this for every observation in the sample and sum it up.
If n=3 and first game I lose, second game I win and 3rd game I win, and my average winrate in the sample is 67%.
Then the variance for winrate is [(0-0.67)^2 + (1-0.67)^2 + (1-0.67)^2]/3 = .22
If my APM was 180, 240, 215 respectively and my average APM was 212, then the covariance is the following:
[(0-0.67) * (180-212) + (1-0.67) * (240-212) + (1-0.67) * (215-212) ]/3 = 10.6
APM variance
[(180-212)^2 + (240-212)^2 + (215 -212)^2]/3 = 605.7
Correlation= 10.6/squared root(.22 * 605.7) = 10.6/11.5 = 0.92
So yeah, in this case a very strong correlation because of my inputs.
Now if we compare with a GM player with an average APM of 300 with the same winrate, winrate spread and same APM spread, lets say 268,328,303 respectively. Then we get the same correlation input, which is fair. So correlation does not depend on the players APM, but more about variation within the samples between the winrate and APM.
The main problem here is that as you mentioned, each race have a different average APM. The only way to do this make sense of it is to measure the correlation for 1 individual, for many games, then repeat this for every individual and then take the average of it. So essentially, you need to be able to measure the same individual over time, otherwise this won’t work. It is also very likely the races have different correlation between APM and winrate, but the 0.65 mentioned is some average between them.
The other problem is that correlation does not take into account the opponents APM. It would make more sense to calculate correlation based on the relative difference of APM between players and winrate.
Correlation between APM and winrate is a valid way too look skill overall, I agree with that. However, it does not imply zerg are more skilled due to having higher APM average, just look at the calculations. It only measures variation within each individual, withing each race, not between the races. If we assume everyone that switched to zerg would gain 50 APM, correlation would not explain or indicate that at all, but APM sure does correlate with skill. In your world, that would imply everyone that switched to zerg would instantly be more skilled, by the mere fact of switching the race.
Now, is is true that switching to zerg will instantly give you 50 more APM? No, I made that up, but if we pretend that is true, what I said is correct and correlation does not say anything about that. Though there are many reasons to expect that simply the mere fact of playing zerg gives you higher APM, due to the way the race functions, and this does not result in the player playing zerg to be more skilled for the sake of playing zerg compared to the other races.
Correction on the last post. You can also measure correlation without looking at a specific individual over time, the average APM and winrate would be for everyone in the total sample, not for this specific individual. The calcuations will be the same though, and it won’t change my general point.