Fix MMR range for Toss and Zerg

[llllllllllll-12402]

@Eliwan-1354

@llllllllllll-12402

You ever hear of “arguing semantics?” You knew what I meant and you know I was being hyperbolic.

Half the issue is that you’re arguing that A is B when they are not actually equivalent statements!

So, if this isn’t what’s occurring (you haven’t stated, specifically, whether or not you think it is), then give an example of what may be occurring instead.

[thrasher-1486]

@Sentry-2358

Im just saying it was not enough. I dont know many guys who said the same but i know many guys that basically said that the nerf wasnt that visible.

I’m feeling the DT enjoyers are talking for the colossus enjoyers let’s be real. IM used to be 50 energy, now it’s 75 with an upgrade and it has counter play. I’ve seen Cure and Clem open up ravens without even getting the upgrade despite playing against colossus so the patch has definitely helped.

[Eliwan-1354]

@llllllllllll-12402

(you haven’t stated, specifically, whether or not you think it is)

It’s not. I’m not actually 100% sure what I’m disagreeing with, though I’m fairly confident in this assertion.

@llllllllllll-12402

then give an example of what may be occurring instead.

I distinctly remember 250, 200, and 150 odd posts ago, where just that was explained:

@Eliwan-1354

Specifically, the evidence to the contrary is that the difference in MMR almost vanishes if you look at a subset of the data; and that as you go up in leagues, the representation of Terran players decreases.
That implies that it isn’t that Terran is weaker or that Terran players are worse, it’s that of the players who are bad, more of them play Terran.

The distinction is that it’s not that Terran 2000 MMR is equal in skill to 1800 or 2200 MMR, it’s that there’s a huge population overlap between players whose MMR will end up at 2000 and players who play Terran.

@Miro-11607

2. There are more Terran players in low leagues.
   2b) This can be general similarity to other RTS games, familiarity through the tutorial, familiarity through the free campaign or from various arcade games, because they’re the human faction, or simply because they found that particular race the most fun and fitting for them.
   2C) Ultimately, there are a host of potential reasons that we don’t really have the ability to quantify from the data we have without holding a poll from those players to find out why they chose the race they did.

@Eliwan-1354

Because People Who Are Bad have an increased probability of playing Terran for numerous reasons that have been stated, repeatedly, namely:

They function the way that many RTS have at least one faction function, therefore, it is very likely to be very familiar.

Supporting this, the reason many RTS have a faction that works that way is because it is intuitive for new players.

People like things that are familiar when they are learning a new game.

The Zerg and Protoss campaigns give you a Zerg/Protoss that is very dramatically different from their ladder races, which is not true of Terran’s. So a story player will naturally be most comfortable with Terran, too.

The in-game tutorials available for Terran are better at doing their job because of where the different nuances of the races are (ex. Overlord spreading isn’t something that any tutorial mentions).

Terran, as a race, has more mechanics that are ‘qol’-esque like range and heal beam/SCV repair; and it has a focus on defensive tools (bunkers / tanks), which are things that both new and weak players find attractive while they focus on “uh, how does this game work” like beats.

Terran, as a race, has a very high intrinsic ‘hype’ to its micro due to the explosive amount of additional benefit it gives: This makes people want to play it, because they want that feeling of out-skilling their opponent.

@Eliwan-1354

The issue I have is the one approximately outlined in the thread’s title: The game is balanced, and when players have bad MMR it’s because they are bad, not because their race is bad. It’s not because their race is harder or more unforgiving, it’s because they lack some skill to succeed as their selected race at a tier higher than the one they are currently in.

Terran appeals to new and casual players. These players tend to be bad. Therefore, of the population of bad players, a disproportionate number of them play Terran. If they switched races, they would remain bad.

The specific relevance of that is that if Terran was the hardest race, we would expect to observe switching off Terran increases MMR.

[Sentry-2358]

@thrasher-1486

I’m feeling the DT enjoyers are talking for the colossus enjoyers let’s be real

A little bit yeah…I mean what can i say? Its pretty dumb that raven counters like everything protoss can do early game. It pretty much counters blink stalker allin/pressure and it counters colossus and it counters dt BOOORING!

I want the times back when all terran were screaming on top of their lungs that its never ever ever ever viable to make even ONE raven…only for them to die vs 2 dts at 5:00 lmao. It even didnt matter if my dts did dmg or not, because if they dont have a raven they still cant go across the map, hehe.

[llllllllllll-12402]

@Eliwan-1354

I distinctly remember 250, 200, and 150 odd posts ago, where just that was explained:

@Eliwan-1354

Specifically, the evidence to the contrary is that the difference in MMR almost vanishes if you look at a subset of the data; and that as you go up in leagues, the representation of Terran players decreases.
That implies that it isn’t that Terran is weaker or that Terran players are worse, it’s that of the players who are bad, more of them play Terran.

The distinction is that it’s not that Terran 2000 MMR is equal in skill to 1800 or 2200 MMR, it’s that there’s a huge population overlap between players whose MMR will end up at 2000 and players who play Terran.

This doesn’t explain what you think is happening. What do you think is occurring? That people who are “bad” are just picking Terran randomly? Maybe. But if so, those players are “inferior players.”

[Miro-11607]

@llllllllllll-12402

That people who are “bad” are just picking Terran randomly? Maybe. But if so, those players are “inferior players.”

We have made no secret that players who are in low leagues are not good players, for one reason or another. Again, it doesn’t matter what race they play. You’re the only one who’s trying to apply “inferior players” to the entire race, which is not what was said.

[llllllllllll-12402]

@Miro-11607

We have made no secret that players who are in low leagues are not good players, for one reason or another. Again, it doesn’t matter what race they play. You’re the only one who’s trying to apply “inferior players” to the entire race, which is not what was said.

I said “on average.”

Again, you’re applying an individual cause to an entire population. It’s just as ridiculous as saying that 40% of Protoss in GM just magically sprouted out of thin air. Or that Protoss winning literally zero tournaments last year is all just a skill issue. In fact it’s far more ridiculous because the population sizes are way larger.

But since those players are “bad,” what makes them that way?

They don’t play less and they aren’t “newer,” so what makes them “less good?”

[Miro-11607]

@llllllllllll-12402

I said “on average.”

Which is exactly why we differentiate “Mean” and “mode”. Why the data was stratified in the first place to give us more information.

Let me ask you this; of the players in gold and below, how many of them play more than the players of other races in gold and below? How many of them are newer players or have fewer games?

This is an important distinction because otherwise data from the higher end can conflate and corrupt data from the lower end. Stratification here is all the more important.

[Eliwan-1354]

@llllllllllll-12402

This doesn’t explain what you think is happening.

Then I need more information on what question you are wanting me to answer.

@llllllllllll-12402

That people who are “bad” are just picking Terran randomly? Maybe. But if so, those players are “inferior players.”

… Yes?

Both of these are close enough to what I have been saying that I don’t mind that as a response.

Players who are bad, compared to the average player, are more likely to pick Terran; and they are, indeed, worse than the average player.

@llllllllllll-12402

I said “on average.”

And the thing that I’ve been trying to say, for the last feels like a dozen posts, is that one really should never try to apply “The average X is worse than the average Y” because mathematically and colloquially these are different sentences.

Mathematically speaking, the mean average Terran might be worse than their equivalent from Protoss. But, by itself that actually tells us nothing.

If you look at race distribution graphs, you will also note that the absolute number of Terrans in Bronze is really high. Of course when you take a mean average of a bottom heavy population the result will be lower than a well distributed population’s.

So it’s not that “I pick a random Terran out of a hat and they’re usually worse than a random Zerg out of a hat,” which is how people will use the idea of ‘average performance’.

This is why semantics are important, because if you say

The average Terran is an inferior player

We don’t know what meaning of ‘average’ you’re using. Or what’s being argued for with this claim.

@llllllllllll-12402

Again, you’re applying an individual cause to an entire population.

No, you’re the one doing this.

@llllllllllll-12402

It’s just as ridiculous as saying that 40% of Protoss in GM just magically sprouted out of thin air. Or that Protoss winning literally zero tournaments last year is all just a skill issue. In fact it’s far more ridiculous because the population sizes are way larger.

The first is, again, weird. But doesn’t it indicate the game is meaningfully imbalanced, nor that one race is easier or harder than another.

The second, while I agree is absolutely bizarre, I think stems from a design choice - the Disruptor is a skill check unit.

On Disruptors

Due to the unit’s nature, at the top level, Protoss has a major - And I would argue its most powerful - tool crippled to extreme degrees. The mechanics of the Disruptor means that as players become more and more adroit with manipulating their positions and splitting their armies in fractions of a second, the value of the Disruptor will decay rapidly into nothing, because the player skill of the splitting player will eventually be capable of far eclipsing any fancy maneuvering that could allow the shots to make good connections. The Protoss ‘cut enemies off’ ability - Force Field - is not as good at forcing units to stay actually still without an actually absurd quantity of them, which means pro players can often just dodge the shot despite the impediment. And this is ignoring the other few answers like picking it off or constantly dropping armies around the map.

@llllllllllll-12402

But since those players are “bad,” what makes them that way?

They don’t play less and they aren’t “newer,” so what makes them “less good?”

I cannot believe you asked this - Why the player is bad is insignificant to the arguments compared to that they are.

But if you insist, an easy way is to invert “How do you improve” - What makes bad players bad is that …
-they miss macro cycles and bank resources or stay supply capped,
-they have poor micro possibly to the point of killing their own units,
-they don’t have logically sound build orders,
-they don’t make rounded unit compositions,
-they don’t have timings in mind,
-they don’t predate upon enemy weaknesses,
-they don’t adapt when an enemy counters their composition,
-they don’t build counters
-they don’t use spellcasters,
-they don’t expand,
-they don’t scout,
-they don’t get forward vision

At this point, can I just say “the fact that they picked Terran” since that would at leaast cause someone to get upset and that has a chance of being entertaining?

[Sentry-2358]

@llllllllllll-12402

They don’t play less and they aren’t “newer,” so what makes them “less good?”

I feel like you only care about “the other variables” as long as you can find them anywhere on nephest or nonapa. You neglect any other variables/influence as they cant be proven or seen directly and therefore they dont exist to you. Or rather: they dont exist for you since we cannot use them to prove something.

In your mind there currently are only 3 things: Activity, seniority and MMR. Since you can only see a deviation in MMR you conclude that the only possible explanation is because they picked terran. Is this really the way you think? Do you actually think this is the way to look at things? You ignore every possible of those thousands of influences just because they cannot be proven by our tools at hand?

Im just asking how you view the world and how to change your mind.

[llllllllllll-12402]

@Eliwan-1354

The first is, again, weird. But doesn’t it indicate the game is meaningfully imbalanced, nor that one race is easier or harder than another.

The second, while I agree is absolutely bizarre, I think stems from a design choice - the Disruptor is a skill check unit.

It’s poor design. Protoss players should be able to win top level tournaments.

Protoss shouldn’t be a shoe in for GM, either.

@Sentry-2358

I feel like you only care about “the other variables” as long as you can find them anywhere on nephest or nonapa.

More or less accurate. I don’t care about them if they can’t be demonstrated. They might exist, but I assume people are all more or less even, unless there’s proof otherwise. You’re assuming Terran players are inferior without said proof.

[Sentry-2358]

@llllllllllll-12402

You’re assuming Terran players are inferior without said proof.

Again. I am not. YOU are saying this. Basically everyone in this discussion knows that WE dont do that. YOU do that.

@llllllllllll-12402

They might exist, but I assume people are all more or less even, unless there’s proof otherwise.

The proof that people are not more or less even is a) people have chosen terran much much much more often, to a degree that you cant say its because of random decision and b) there is a much much much higher population for terran in very low leagues.

The statements a) and b) let nearly every sane people conclude what is believed to me as pretty obvious: People who arent that familiar to the game will pick terran as its the go to race and because they are not that familiar to the game they will be placed at the bottom in very low leagues. Because usually people would acknowlegde the difference in mmr for too big to be simply by imbalance or “too hard” to understand/to play. Also the difference of population is pretty much nonexistant in the highest leagues. Why should terran be so freaking much harder that it becomes apparent to everyone on lower leagues but its suddenly not visible in the highest leagues? That really doesnt make any sense to me.

@llllllllllll-12402

They might exist, but I assume people are all more or less even, unless there’s proof otherwise.

The problem is that you believe something without a proof (terran is harder) but wont believe us without a definite demonstratable proof (terran is not harder based on the data). Also you ASSUME that they are all equal. But you dont know it. And you cant demonstrate that.

You even say that there MIGHT be something that exists. So you are giving in that there are more than 3 variables (mmr, activity and seniority) but you still wont acknowledge other variables to be important just because we cant demonstrate it because we lack data. Do you really actually think you can explain everything that happens in sc2 with just 3 variables? Even tho the 3 races are so much different?

In short: Thats pretty unscientific and not well thought and you pretty much contradict yourself by believing something without proof but you wont believe other people because they cant demonstrate something even tho they have proof of concept (for example mere-exposure/familiarity).

Your basic believe is: I cant see it, therefore it doesnt exist.

[thrasher-1486]

@Sentry-2358

raven counters like everything

I mean when else would you see ravens if it was nerfed further? If they want to keep cracked cyclones and make the early game dangerous for terran, I think the fun meta shifting change would be to have siege mode be an upgrade again. That would make it so Dark can open roaches again and punish people that just yolo their cyclones.

[llllllllllll-12402]

@Sentry-2358

Again. I am not. YOU are saying this. Basically everyone in this discussion knows that WE dont do that. YOU do that.

You are though. You’re just refusing to admit that you are. You say again and again “Terran players are more likely to be bad” or “Bad players are more likely to pick Terran.” Source? Trust me, bro.

[Miro-11607]

@llllllllllll-12402

“Terran players are more likely to be bad” or “Bad players are more likely to pick Terran.”

And again; these are literally two different things!

You are, and have been the only person who has said:

@llllllllllll-12402

“Terran players are more likely to be bad”

I will again reiterate that this is fundamentally different than:

@llllllllllll-12402

“Bad players are more likely to pick Terran.”

[Sentry-2358]

@llllllllllll-12402

You say again and again “Terran players are more likely to be bad” or “Bad players are more likely to pick Terran.”

This is not what i have stated at all. YOU try to do that. YOU try to make it seem like ALL terrans are doing worse just by picking terran. YOU try to strawman us into saying ALL terrans are worse. Only that you can then say they arent worse, the race is just harder.

Since the beginning of the argument i have always said its not about ALL terrans. Its specifically about very low leagues, aka players below average. You try to use them for your balancewhine. Which is straight up laughable.

Since the beginning of the argument i have said to you that you dont understand the difference between correlation and causation and you would think that people who smoke are less intelligent (or earn less money) than people who dont smoke because of causation. This is nearly the exact same debate. Because its not about the thing itself (playing terran; smoking) which causes the difference, its mostly about all the other variables/influences there are.

[llllllllllll-12402]

@Miro-11607

And again; these are literally two different things!

And again you avoid the topic. The point is that maybe Terran players are, in fact, “just worse” on average. Whether it’s because bad people are more likely to play them or not. What you haven’t done, any of you, is submit evidence for such a claim. All evidence we DO HAVE point to them being more or less the same as other players.

[Sentry-2358]

@llllllllllll-12402

The point is that maybe Terran players are, in fact, “just worse” on average

Fallacy of composition.

We can see a huge deviation between the behaviour of terran in lower leagues compared to higher leagues.

To average them out and mixing those 2 very different behaviours to just one behaviour would be pretty fallacious.

@llllllllllll-12402

What you haven’t done, any of you, is submit evidence for such a claim.

You havent submit any evidence for your claim at all.

“Terran is harder”

• Why?
  “They have on avg lower mmr”
• Where is the proof that the only deviating variable is avg lower mmr and means terran being harder?
  “IT IS THE ONLY POSSIBLE EXPLANATION OK??”

…yeah no.

@llllllllllll-12402

What you haven’t done, any of you, is submit evidence for such a claim.

Actually we have submitted proof that most of the arguments we used have proof of concept and are generally considered to be true or important (in the sense that they can be true). Even your chatgpt prompt told you that there are multiple things to consider, like players preference. To this day you didnt have any answer to the problem that terran has by far much more players other than “you dont know WHY players choose their races”. Exactly. You dont know WHY they do it. We only know that they are doing it and there are reasons for it and it cant be random. I mean even a 12 year old could figure out why the most familiar and similiar race that gets played in campaign and tutorial would have more players than an unfamiliar slimey, buggy, difficult to grasp alien race. Who would have thought…

[Eliwan-1354]

@llllllllllll-12402

You say again and again “Terran players are more likely to be bad” or “Bad players are more likely to pick Terran.” Source? Trust me, bro.

The source is the data on the ladder - let me show you. The reason we observe:

@llllllllllll-12402

The point is that maybe Terran players are, in fact, “just worse” on average.

is because – and only because – the data are heavily bottom weighted. When we bracket the players, we notice more Terran players in Bronze than the population ratios would suggest we should. That doesn’t, initially, make sense, because:

@llllllllllll-12402

All evidence we DO HAVE point to them being more or less the same as other players.

Yes. This is the exact sentence everyone has been saying the whole time. Each player - at a given range - is equal. A Gold Terran’s equal to Gold Zergs, a Platinum Terran is equal to a Platinum Protoss; for any rank with any race pair:

A player’s rank/MMR is based on the player’s skill, and has nothing to do with what race they selected.

So now armed with that, we go back up a step. We observe too many low rank Terrans compared to the population, but we have no reason to believe that the Silver Terrans are actually equal to Gold Zergs.

Therefore, the reason why there are so many Terrans in this MMR is because bad players are more likely to pick Terran. We find this a reasonable conclusion because Terran, mechanically, has multiple things that appeal to new and-or bad players.

Therefore, because more bad players pick Terran, their average MMR is only lower than everyone else’s because they’re bottom heavy, therefore, if we separate the top and bottom before averaging, the averages should be equal.

And they are, so the case is closed: We have come to an explanation that lines up with the data and that makes sense given our understanding of the world. It’s a weird explanation, and not an ‘ideal’ one, but it is logically sound.

[llllllllllll-12402]

@Eliwan-1354

is because – and only because – the data are heavily bottom weighted.

You can’t draw conclusions based on “statistics contain clusters.” Nobody does that.

@Eliwan-1354

Yes. This is the exact sentence everyone has been saying the whole time. Each player - at a given range - is equal.

You add this part:

@Eliwan-1354

at a given range

And then act like it’s the same sentence. It means two totally different things.