Average APM correlates with average MMR with an R^2 value of 0.983, meaning MMR can be predicted from APM with 98.3% accuracy. APM = Skill. This is undeniable fact when talking about large groups of people.
Now, given the fact that Protoss in GM have 25% less APM than Zergs and 23% less APM than Terran, where the heck are the Protoss nerfs! There are a bunch of Master-level Protoss players in Grandmaster! How on Earth is that fair?!
I don’t know champ. I usually hover around 130 apm yet when I watch streamers (top GM only) I can guarantee I’m macroing better than them. I also play minesweeper at a semi professional level so you can tell I’m an epic gamer.
You do understand how an average works, right? You do know that you need a large sample of players and games to get a reliable average, right? You do realize that 1 player winning a couple of games meets absolutely zero of the criteria for statistical significance, right?
We can’t have a conversation about balance if an average is a difficult concept.
I’ve sent out the memo to all my Toss brethren to start spamming and just hold a hotkey down. Now that Batz has figured out the most solid metric for evaluating skill, it’s time to make him admit we’re under performing, need lots of buffs and we’re the race with the most skilled players.
I look forward to your apology next week. Don’t let me down Toss bros. I don’t want to hear any excuses about how Zerg’s APM is massively inflated due to being able to make 200 lings at once by holding down a button or how Terran has mindless micro APM spam.
If you have spare APM to do nonsense spam while being in GM, then you’re proving my point. An APM budget for a Terran or Zerg is stretched so thin you’re frantically scrambling every second of the game and you don’t have spare APM to spam with.
Or Toss players are super geniuses who have all figured out the game and have perfectly optimized everything; meanwhile peanut brain is busy spamming the keyboard to do 1 action.
Imagine having no idea how to play the game, yet playing with high APM. Talk about arrogance and ignorance. Like bruh… you’re clueless, realize it… take time to think through what you’re doing.
Given I’ve played both races before… be it a different game or not, Terran’s APM is inflated by stimming and bio micro. It’s really mindless and not a big deal, but yeah… sure you end up with a higher APM, but… the units also own everything if you micro them. Pretty cool trade off…
While I already feel sick for giving Toss players credit, whether trolling or not, I actually have optimized hotkeys. I treated hotkeys as the fourth matchup in the game, and I would make adjustments after every game until it was obvious that… nothing could be improved. Thus, my needed APM is lower than it would be otherwise, by a good bit.
People even miss the obvious one. Like it makes no sense to have all army on anything but tilde. Select all workers, building hotkeys, etc… key for any building… all optimized. So, really… I just take it as a compliment. I have a dream to one day be GM with 10 APM, solely because of how well I optimized everything. True skill happens outside the game.
Spamming is not skill. All it takes is any experience in BW to realize you can’t go off APM. On Asia server, every Silver league equivalent has 300+ APM. Not even joking. It’s out of this world crazy. So, they’ve already proven that it’s possible to have a lot of APM yet have no idea how to play and… no skills outside APM.
Nobody has APM to spam in GM. APM budgets are stretched razor-thin until you ultimately fall short and make a mistake. That’s how macro games work. Oh, you don’t know that? Golly, it’s almost as if APM isn’t a big deal to Protoss.
That’s silly. Again, all one needs is any experience in BW, where people have experience at looking at effective APM verus APM. Many pros have ton of spam. There’s a method/reason to the madness, but it’s still spam.
I am glad you brought that up. EPM is higher for Zerg as well and EPM filters “spam”. Thus it is proven that Zerg’s APM is not higher due to spam.
It’s actually pretty bananas what the average % of APM is considered spam. I can’t speak for SC 2, but in BW, Mondragon (just didn’t spam) and Incontrol (low apm) were basically the only Zergs who didn’t spam. It’s definitely the norm for a high % of actions to be treated as spam, at the pro level.
Before I take your argument apart piece by piece let me start by saying you have chosen a terrible visualization to represent a polynomial regression. You should be using scatter plots with a line of best fit.
Now let me ask, where exactly did you get some of those numbers from? They don’t exactly relate to the single source link you posted.
Now going back to the point on your terrible visualization - I bet you chose this because you don’t have a proper scatter plot of data points. Once again you are working with average values whose distributions you do not know of. This puts your entire model under question. What makes this worse is that you essentially have 8 data points and are making a generalization on an entire population for that. Of course, these 8 data points are averages, but once again, you don’t have a single clue about their distribution or any other assumptions for that matter. You need SIGNIFICANTLY more information than what you are outputting right now to draw any remotely objective conclusions.
Once again, you are using what appears to be 3 DATA POINTS and then applying a line of fit on them. This is so laughable that the only people you could convince are those who have 0 background in statistics. As someone who KNOWS statistics, let me tell you: you are TERRIBLE at using statistics.
You know, let’s say you ACTUALLY had a valid model, large working data set with assumptions filled out, and proper statistical analysis. THEN I would be more on board with what you have to say. However, not only are you absolutely terrible at using statistics, once again you fail to address all the other surrounding factors that lead to the definition of skill.
So I’ll say it once again: CORRELATION DOES NOT IMPLY CAUSATION.
Your entire thread is thus invalid. Come back with a better paper.
It doesn’t have to. The correlation only has to be a strong predictor for skill that way it can be used as a metric to measure skill. Since Protoss have drastically lower APM in GM, and APM is a strong predictor for skill, it is definitive proof Protoss is overpowered. They are almost twice as common in GM as Zerg, despite being Master level in skill according to their APM. How do you get higher representation in GM from lower skill? It’s called balance favors Protoss.
This is the last period: h ttp://aligulac.com/periods/291/
People have to admit that GM is an arbitrary number of people, and… it’s not like some pinnacle of skill. In all honesty, it’s pretty bad. I mean, I’m not even a tournament level player and I can go a year without playing a game, come back and instantly be mid GM if not higher. I mean, are those really the numbers you want to focus on? It’s about as relevant as any other league.
If a person wants to be like, hey… it takes less skill, granted you know what builds to use, to get GM with Toss. I’d have no issue with that. Could very well be true. But, at the pro level? If you’re interested in winning the most games or making the most money… play another race.
Balance by definition has to affect the whole group of people who play a race. Grouping by skill and not game design means your metric is of and about skill and not game design. If a correlation exists within a group, it’s a correlation with that grouping criteria. So as soon as your criteria includes skill, it’s not about balance anymore.
Just look at BW. Toss is 10000000000000000000000000000x easier to play at a relatively high level. Yet, at the very top level of play… absolutely no one wants to admit/think Terran isn’t the best race. Historically, Terran is the best race. Heck, even 90 out of the top 100 were Terran at one point, on ladder.
Yet, if you’re wanting to own the non Korean scene, for instance… you pick Toss. Anything but Terran. If a certain sect of people are only going to reach a certain skill level, then no one is going to reach a high enough skill level to reap the benefits of Terran, for instance. Meanwhile, with the same skill, you could play Toss closer to optimization and thus end up far better off.