To all the APM=Skill deniers

General Discussion
122

Shouldn’t it go something like this: If the perfect player always knew the right strategy and what to do, across the board, he would need, say, 350 APM for Toss, 450 APM for Zerg and 500 APM for Terran to actually be able to do/execute it.

Thus, if a Toss player doesn’t have 350 APM, in this instance, then he can’t be the absolute best (room for improvement at least) and so on and so forth. The closer one’s APM is to these numbers, or if it surpasses them, then, in theory, they could be better than someone with less APM. At that point, it just comes down to whether you know what you’re doing or not and can you micro well enough, etc.

But, if you don’t have the required APM, then you can’t do what you want to do. You’re just not fast enough to accomplish everything that needs to be done, whether you have all the right answers or not.

So, in a sense APM is unlocking the ability to “be the best.”

But, at the end of the day… none of you guys create anything. You just copy w/e the meta is. Not many cognizant players. So, sure… in SC 2… maybe most games are being decided by simple multitask, thus more APM kinda = more skill.

However, one of the absolute greatest players of all time in BW is Iloveoov. An absolute legend. Way better of a player than Boxer ever dreamed of being (had to be said). He played Terran. Do you know his APM? It was around 220. Terran was considered the most APM intensive race. Far more actions to do/needed in BW than SC 2 with Terran.

Despite this, Iloveoov was the best player in the entire world. I played games vs him where he was 212 APM, not far off his avg in pro games. The guy simply had a better understanding of how to play than anyone.

If you don’t have a strategical advantage, you need a multitask advantage. Supreme strategy/tactics can still win over sheer APM. It just comes down to how big is the gap/edge in strategy.

Ideally, ofc, you could make the correct decisions while playing at a fast pace. But you’d rather make good decisions than just see how hasty/fast you can be.

123

The fact that the correlation strengthens is exactly how the proof works. So you have a very big misunderstanding there. The correlation between non-APM variables shrinks to 0 while APM grows to 1 as the sample size goes up. Translation: non-APM factors have a small effect on winrate when all other factors are roughly equal. The effect that apm has is 500x larger than non APM factors. Here is summary of the proof:

Sorry pal a correlation this strong is proof. Here you go:

  1. All factors that affect MMR must be conjoined (they combine to affect MMR) & must correlate with MMR (anything that affects MMR will correlate with it).
  2. Any variable that is dependent is indistinguishable from APM since APM’s correlation is 1, thus a variable that affects MMR must be independent AND conjoined. If the correlation between APM&MMR were lower, this would not be true.
  3. A variable cannot be both conjoined and independent given the correlation with APM is near 1. A conjoined variable that has any effect on the outcome would, by definition, reduce the correlation of APM to <1 given that they are conjoined. APM has a correlation very close to 1 and thus such a variable does not exist.
  4. Any variable MUST either A) have no effect on MMR or B) be indistinguishable from APM. APM is therefore the only variable that affects MMR.
1 Like
124

Here is the proof in more layman terms: since factors MUST combine to affect MMR, if one has a correlation with MMR it comes at the cost of the other losing a correlation with MMR. Given that APM has a correlation near 1, there are no other factors that affect MMR. Since their effect on MMR is conjoined (combined), if there were any other variable then APM would have a correlation of <1. Apm has a correlation of very close to one.

Now this is assuming that the variables that affect MMR are independent (e.g. that they don’t correlate with one another). It is possible for a variable to correlate with both MMR and APM. However, since APM has a correlation of near 1 with MMR it means these other variables would aggregate into APM’s correlation and are thus indistinguishable from APM in a statistical sense. So in other words you’re talking about the same thing but with a different name.

125

Setting aside opinions about whether OP’s claim is “pretty similar” (whatever that subjectively means to different people) to another claim - As long as we agree on your above quote, “APM and skill strongly correlate for obvious reasons” and the fact that protoss, as group of players, has significantly lower APM than the other two races, then I don’t think we disagree here.

2 Likes
126

Asamu says the fact that each race has a slightly different correlation between APM & MMR is proof that APM can’t measure balance. He says each race has different APM requirements. This is an assumption that APM requirements don’t affect MMR which is absurd given that APM has such a high correlation with MMR.

If balance is perfectly equal, race selection has absolutely no effect on MMR. That means there will be ZERO correlation between race & mmr. The fact that MMR & APM correlations change depending on race is proof that balance correlates with MMR. If MMR didn’t correlate with race, then grouping by race should not affect the correlations between MMR and APM.

This is because the factors are conjoined (combined). So if you have one variable driving MMR, it comes at the cost of another variable. So if you group by race and APM/MMR correlations change, then it means race & MMR has a correlation that is taking away from APM/MMR.

1 Like
127

For different reasons.
Go play all 3 races. Tell me your APM with each. A protoss player that plays Zerg for a few games will see an increase of 20-50 in their APM while playing Zerg without any increase in their mechanical ability/overall skill, and without putting in “more” effort. That doesn’t mean they suddenly became a better player for having more APM while playing Zerg.

There’s a reason - separate from skill, that causes Protoss to have significantly lower APM at the same level of play.

IE: that correlation of skill and APM is dependent on race, not independent of it.

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

This graph, courtesy of Batz himself, is a pretty clear indicator of that. Both the top end and bottom end of APM for Protoss is significantly lower.

Saying something along the lines of “well Protoss are just 25%-30% worse on average”, because that’s the difference in APM, when those averages are taken across the entire playerbase, would be absurd.
Granted, that graph isn’t 100% accurate or anything; we know that GM representation for Protoss is a bit inflated right now, which will drag down the APM average somewhat for that league.

The average skill of players of each race across the entire playerbase should be assumed to be similar; there’s no logical reason to assume otherwise.

It’s the same sort of issue with a some of the other typical metrics, like unspent resources - Protoss intentionally hold resources for emergency warp ins. Neither of the races do such a thing, or have a good reason to bank resources when not maxed for extended periods. Obviously, that’s going to make resource management look worse for Protoss, if the measure you’re going by is “average unspent resources”. Zerg has the easiest time keeping resources down because of their production mechanic; missing a production cycle doesn’t set you behind and require you to build more production facilities to catch back up in spending.

128

These statements were literally right next to one another. In one he says it’s absurd to say that lower APM equates to lower skill, then he does a complete 180 and says that the lower APM in GM is due to lower skill Protoss dragging down the average. Lmao. He is so close to realizing how this concept works.

A race is overpowered → their MMR goes up → their average APM goes down because the APM/MMR relationship is skewed by the higher MMR.

That’s exactly what is happening in GM. You’re right, Asamu! But guess what? That’s what’s happening across the whole ladder, too!

1 Like
129

I think you’re missing the point.

1 is pointing out that GM in particular has deflated APM for protoss as a result of increased representation, due to the correlation between APM and skill, though probably not much. I doubt it’s changing that average by any more than ~10.

The other is that, even without that. In a period of perfect balance, Protoss APM would still be significantly lower, but that would not mean Protoss players are bad, or that the race is OP. You literally cannot have as high of APM/EPM as Protoss as you can as Zerg/Terran (assuming you’re playing at an equal level as well as you can).

Protoss has ALWAYS had lower average APM in every league than the other races, and for most of the leagues on your stats, the numbers are pretty much the same as ever. an APM of 77 for Protoss vs 110 as Zerg is the normal state of gold league. That 77 APM Protoss will likely pull ~110 APM as Zerg if/when they play it.

IE: this isn’t some phenomena caused by race balance, but by how the races are designed.

130

These are contradictory statements. Imbalance causes a player’s MMR to go up. His APM stays the same. So people with lower APM are pushed into higher leagues, which drags the average APM down. So balance and APM are negatively correlated. If a race is favored by balance, their APM goes down.

You already admitted the Protoss GM representation is inflated and that that is dragging down the average APM in GM. Now apply that same logic to the rest of the ladder and you finally understand the argument. There is a very strong correlation between APM & MMR and Protoss messes that up because they get a higher MMR which breaks the typical APM/MMR correlation.

131

The point.
Your head.

Go play 10 games with each race vs some bots or something doing meta builds and look at your APM/EPM after each game with each race.

132

Nope. Balance and APM are negatively correlated so if a race is disfavored their average APM goes up because balance bumps the bottom-end of players to a lower league, thus raising the average of both the upper league and the lower league.

If a race is favored by balance, the average APM goes down because people at the top of a lower league, who have lower APM, are bumped up a league, thus dragging down the average.

Again, there is a relationship between APM/MMR and if grouping by race messes up that relationship, then it means race affects MMR.

It would be lower for my Protoss, thus proving my point. Protoss is easier to play as a result of requiring less APM, and this makes it a stronger race for people with low APM.

133

Probably so, but your results also showed (if being reasonable) that Terran and Zerg’s APMs are inflated. You’re basically showing an APM gap starting at bronze that is maintained through GM, where the obvious thing to glean from it is… your APM is naturally boosted by playing Zerg or Terran. It’s not that you’re playing faster, necessarily, it’s just extra “fairy dust” APM.

At the end of the day, what’s the actual point? What is the obsession with GM? I just came back after not playing in 4 months, and you guys are making me look like a world beater, as I struggle to even properly wall off on a map.

There’s probably a lot of parallels between BW Toss and SC 2 Toss. If you want to, say, reach 2k MMR (BW) the quickest, pick Toss. You want to be the most successful pro gamer ever and have lots of fan girls? History would say… pick any race but Toss.

Toss units are truly trash (SC 2). So many of them are horrible. There’s not many good synergies. It’s really not a good race at the pro level. There’s only 1 way to skin the macro cat with Toss. Even if you play optimally, it’s so predictable that it’s losing strength based on that alone. The other races? Plenty of options. Hard to go wrong.

It really just feels like bullying and trying to put down Toss players and find a way to spin Toss into needing nerfs. The race isn’t too strong. It’s probably the weakest race in the game, if the perfect player were to play each race, who had 500+ APM…

You’re just salty that Toss is easier to get a certain MMR with. I’ve been there before. But then I just played Toss and enjoyed it. At some point, if a race is as easy as you think, you just switch and you’re able to prove it, instead of just trying to find obscure stats.

If your race is going to be the worst, overall, it kinda sounds like a nice/needed consolation prize of being the easiest to play optimally. May as well get something out of it. Annoying Batz isn’t exactly the best reward…

134

Are you the guy in Avilo’s new video? He posted a video yesterday and there’s a guy collegewing in it.

135

facepalm
No, this does not mean it’s easier to play as a result of requiring less APM. You are still the same person, and still doing things just as fast, yet you have less APM - IE: there is another factor in play, like race mechanics/what types of actions are being done. You seem not to be able to comprehend that not all actions contribute to APM equally, which should be obvious.

While Protoss is OP on ladder right now, your argument is just fundamentally flawed. That’s not because of APM. Protoss APM has ALWAYS been lower. Do you think all protoss players should drop ~500 MMR, resulting in almost no Protoss should be in GM, and Bronze being 90% Protoss? Those would be much stronger indicators of something being wrong than the current metric of APM that you’re trying to use, which shows a fairly consistent difference in APM, regardless of the level of play.

136

That’s exactly my opinion, I see threads with endless rants…about apm, and the question is, who cares?, and why should we care about it?. It’s just a game, one race requires less apm, that’s all. I’ts not the end of the world, you play the races you like, if you want to play slower, play protoss or learn to play slower with other race. And the GM talk is nearly useless, the game is balanced around the top players, not the average GM player, here you guys love to talk about Lowko, well, the other day he got GM on NA.
So the point is that the rants are usless because not only the game is not balanced around ladder, the game maybe won’t be patched again, so any meaningless discussion is a waste of time.

2 Likes
137

Not all APM is equal. I think it is only a fair gauge of skill at the very highest level, and even then you shouldn’t use it to compare each race. People seem to greatly underestimate how difficult it is to eliminate wasted actions. Zerg has always required a higher apm and protoss a lower apm, that is just the way the game plays. Does that mean protoss is easier to play? Yes it does. Does that mean it is overpowered? No it doesn’t. If you wanted perfect equality there would be no point in having different options. This topic itself has been beaten into the ground over the years and is nothing more that hurt egos whining.

138

Nope. If you have less APM, you are by definition doing things slower.

I have a challenge for you Asamu. I am going to make a mod. In this mod, pylons will be half cost and provide half supply. You’ll need to make twice as many pylons. Then I will halve the stats/cost of all Protoss units. So now you need to make twice as many Protoss units.

According to you, having to have more APM for better production won’t make Protoss harder. Are you ready to prove your point by playing on this mod?

1 Like
139

Dont forget to halve the cost and size of gateways as well, so that we can still spend money as fast. And halve the size of our units, so that we can still have the same amount of power in a fight.

Oh, and halve the size of pylons too, so we actually have room to put them all.

1 Like
140

Nope. APM has a correlation of near 1 with MMR. It is a near perfect indicator of skill for the whole ladder and that’s a fact.

This is a severe misunderstanding of how the game works. The reason why APM has an extremely high correlation with MMR is because APM is an easy way for players to win games. Simply put, a player with more APM can handle a larger workload without making an error. As the game goes longer, the lower APM player makes more mistakes that accumulate eventually into a game-losing position. If a race is easier and requires less APM, a player is less likely to make errors and will have their rank inflated above what they could achieve playing a different race. Getting a higher rank with one race and a lower rank with another is the definition of imbalanced. Easier = imba. Imba = easier.

Also completely wrong. Terran and Zerg are near-equal in APM (and other skill metrics). So it’s obvious that you can have different races be the same difficulty.

Also completely false. It has nothing to do with ego and absolutely nobody has covered this topic as in depth as I have. So you’re flat out wrong across the board it would seem. Do you think it’s a coincidence that at the same time Protoss is dominating, the game’s popularity is tanking? This isn’t about ego. This is about making the game fun and an integral part of being fun is being fair.

When ZvP is so ludicrously Protoss favored that Protoss with Masters-2 level skill are in GM, while Zergs with GM level skill are in Masters, ZvP is a fustercluck that is polar opposite of fun to play. It’s a nightmare. There is a reason Zerg is the least played race on the ladder right now.

141

Lmao. PPP trolls were losing the minds that Zerg was winning 1.7x as many tournaments as Protoss. Protoss are currently winning 4x more but now the PPP is saying “Who cares about balance!”. It’s it funny how that works?

This is so delusional it’s hard to know where to being. First, the majority of GM accounts ARE FROM PRO PLAYERS. Second you can’t balance the game around the top, it is mathematically impossible, because balance affects a win-rate maybe 5% at most while skill affects it 0%-100% and the top of the professional scene is driven ENTIRELY by skill - balance has no say. Balance has MAYBE a twentieth of the impact that skill does. Third, Blizzard has stated numerous times that they do consider not just GM, but ladder balance as a whole. Fourth, it would be patently insane for a profit-driven company to not consider the needs of their consumers. Fifth, balance by definition has to affect all of the people who play Zerg/Terran/Protoss or what you are talking about isn’t balance. You can’t exclude the majority of the Zerg set and then say you’re talking about Zerg.

This is pure delusion in like 50 different ways. My advice to you is to go to school and get an education. Because this is just very bad. This level of pure nonsense is just off the charts.