Because they have lower MMR/are more likely to have low MMR despite no data containing possible explanations for why.
Also, because I’ve asked like 30 pros on stream and they all said Terran was the hardest (actually a lot of them avoided the question, but the ones that answered).
It’s not a strawman. Characterizing my statements as “Bronze players aren’t bad” is a strawman. You can’t use the stratification of data to draw conclusions (well, you can, but not individualized conclusions of an entire population).
I know exactly what you were saying. It was stupid.
Here, I split them by MMR. There’s a statistical cluster (Literally find any statistic that doesn’t contain clusters). Ergo the disparity is due to ___________ (insert some individualized characteristic that you can’t measure or even define a measure for).
Okay. There are a few particular things to point out here.
We need to compensate for population sizes - how population pertains to match frequencies, how that affects player ability;
We need to account for how league stratification works, that in order to ‘get’ to a rank you have to ‘accomplish’ a certain list of feats / acquire a certain set of abilities, and those are different per-race, ie. does a race’s learning curve as that pertains to skill acquisition affect it;
We need to make sure of other variables, particularly, skill distribution compared to population distribution - there’s numerous arguments in this and sister threads about how that’s not the case, for instance.
But all around, thank you. I appreciate the clarity and straightforwardness here.
To be clear I have your point correct: You think that Terran is the hardest race because the median and mode average Terran has a statistically significantly lower MMR.
But equally critically, none of the data suggest foul play.
I find it very doubtful that these were not all Terran players who just want to cultivate their audience that way (:
And, as an actual point, a lot of pros that I’ve seen interactions with over months don’t think any race is the hardest as some objective statement, but that each race has the hardest time with a specific task.
It’s a good thing I didn’t do that, then.
Seriously, did you even read my post? I said that I think that isn’t a sentence that requires proof. At no point did I even imply that you had said it or something like it.
You just want the people you’re talking to to be wrong, and are misreading or misrepresenting what they’re relaying to you - to a degree that I swear it is on purpose at this point - instead of actually engaging with content.
This is the definition of extrapolation - Which does hold true in both statistics and general mathematics. Using a subset of the population, we can find probably true information without having to analyze the whole population.
The important thing is that this cuts both ways: If, when looking at the totality or a subset, a particular trend appears or disappears, that is a vital piece of information about the ‘truth value’ of that trend.
If you knew exactly what was being said, then you would also know it wasn’t stupid. Something that everyone has been telling you.
Secondarily; a note:
One might ask “where did that come from”, well; see here:
This being a reply - when I have the courtesy to read every one of these tinyposts even though they, by sheer volume, have just as much content as my sprawlposts, is frustrating, because it means that even if I explain something at length, I don’t know if you even bothered to read it.
I’ve had to type out Terran appeals to casuals is not the same as Casuals play exclusively Terran is not the same as Terran players are bad is not the same as Casuals are bad - multiple times.
I’ve explained multiple times how When you divide the playerbase by MMR then average their MMR they have the same MMR is logical because it controls for the variables of population size, casual player appeal, and bad player appeal.
And, of course, the data science behind the idea that stratifying data changes the trend line, and looking at data from multiple angles will reveal interesting things or demonstrate that the whole of the data tells you something critically different from each section of the data and that’s extremely important? Covered that three times. Well, four, now.
And Here you Showcase that you dont understand what it means. I would agree with you If i Split Them in many small groups and say: “hey in about 3/5 the mmr is nearly exactly the Same”…then you can come Up with Clusters.
But i didnt. I Just devided the Data into 2 parts (which are still very very very big so they cant be called Clusters)
When we Look at the big Picture (Just using whole Population) we See that terran has on average less mmr. You concluded that ALL terrans therefore achieve less.
By dividing the group into 2 parts and looking for average mmr again we See that the mmr disparity diminishes.
What does that Tell us? It tells us that the mmr disparity from the whole Data Set comes from the Population disparity in very Low leagues since there are Sometimes Like 50% silver/Bronze terran Players.
Meaning terran does Not achieve less because their mmr is the Same as Other races when you Look at inexperienced Players (Bronze to gold) and experienced Players (Plat to Masters)…there are Just way more inexperienced Players for terran and thus dragging the mmr down for whole Population.
Btw you Said multiple Times that terran is in General too hard because of those Guys in silver/Bronze. Which is in your own words a) a Huge generalization and b) fallacy of composition.
Finally: you came Up with 2 examples to falsify my approach and both were hilariously wrong:
If we Put men and women in 2 categories of Shorter and taller we would still See that men are on average taller than women. While you Said the difference would diminish.
If we Put people with similar BMI together we would See that the conclusion we can Draw from the bmi are pretty much Limited: we could See many different Body shapes; even If someone has a bmi of 30 it doesnt mean they are fat. You told me you were a Boxer or mma Fighter. You should actually know that o.O but No you told me that Like people with similar BMI would always Look similiar.
So you either dont want to understand Things or you really dont understand Things.
For clarification:
In cluster sampling, researchers divide a [population](scribbr .com/methodology/population-vs-sample/) into smaller groups known as clusters. They then randomly select among these clusters to form a [sample](scribbr. com/methodology/population-vs-sample/).
Cluster sampling is a method of [probability sampling](scribbr. com/methodology/probability-sampling/) that is often used to study large populations, particularly those that are widely geographically dispersed. Researchers usually use pre-existing units such as schools or cities as their clusters.
This is not foul play. It’s by design. David Kim even talked about it back in 2011 or something (he mentioned in a separate interview that Terran was the most difficult). Not all races are equal in difficulty, but they’re meant to be balanced for the highest skill level play. “Terran is the hardest race” is a controversial statement maybe here and reddit. It’s been in the Starcraft vernacular since 1998.
They were not. Many of them were (I am a Terran main), but other include Crank (P), TLO (Z), BeastyQT (he was Random, but I’ll give you he was a T main), Showtime (P), even Parting.
The issue I take with the “casuals” comment is you have no measure to define someone as such, other than their MMR. Sentry tried to characterize them both as newer players then as player who play less frequently, failed, then got angry at me lol.
Again “casual” is in incredibly subjective term. I have no idea what you even mean by it. REALLY mean. I know it means “someone not focused on improving,” but by what measure?
You people all keep saying “the evidence doesn’t point to Terran being more difficult,” but at the exact same time, if we simply ran the counterfactual, and found that Terran players did play fewer games, that Terran players were newer players. What conclusion would you have drawn from that? That Terran players were casual? I think you would have (and I would have been hard pressed to disagree).
Literally quantifying the effects thereof in front of you. That’s why you proposed those same metrics… That’s also why you got angry with me when those metrics cut you.
If this was your actual point, you would have brought it up in the beginning. You have already shot your credibility, and thus this comes off as appeal to authority.
I remember this line significantly differently: That Kim found personally Terran the hardest to play, as hugely distinct from being designed that way.
And … the game is significantly different now than then. So many units have received so many changes.
Here we see literally everyone offering their relevant definitions of casual.
Three of four: “Casual players are people who are not dedicating themselves strongly to improving game skill.”
Which, indeed, is something we cannot ‘measure’, because we don’t have access to data about it. And this is where Kelthar’s earlier-dated quote factors in: The main way that you can know a player is casual as opposed to actually just horrendous is by asking them.
But absent of that, we only really have Kelthar’s second quote, which lines up with your assertion - the only way we can currently identify a player as casual or not, due to lacking access to data, is to assume so based on their MMR.
You’re allowed to have a complaint with it, and my complaint is that you say this in response to a quote that has nothing to do with that, and second,
It’s important to remember that this didn’t happen.
My conclusion would be approximately unchanged. The race appeals to new and/or casual players, so players who are newer and/or play fewer games pick it heavily.
In this hypothetical, because they’re on average less senior, we would see a higher turn-over in the number of Terran mains, which would - from my understanding - be likely to lead to a weird bubble where Terran would become population-relative hugely under-represented at middle skill levels, because new-likely-bad players are very likely to pick it and then change races or leave the game.
Critically, at no point, either in this reality nor in that one, would I conclude Terran players are on average casual / Terrans are more casual or whatever variant you please; because I don’t have a stake in that issue.
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.
In summary, I disagree with the idea that switching off Terran increases MMR - which feeds into the same ‘things’:
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.
Not “by design” in that it was on purpose. He wanted them to be similar in difficulty levels, but it was impossible to do so.
It’s exactly an appeal to authority. That’s why I didn’t make it. It’s just a reality. I even started this thread by telling Weazy not to complain about it.
I’ll have to cry myself to sleep over you not believing me.
Right. So what measure would you use, if you had the data.
Oh, it absolutely did. Hell, ask Sentry. I bet he’ll cop to it (I give it about a 55% chance he’s willing to be honest).
It’s important to understand that while you might think that, you’re believing it 100% without evidence. In fact, in the case of “newer players,” you’re believing it in spite of direct evidence to the contrary.
It was 2/3 false. I was not angry at you and you didnt prove that terrans arent newer (again the keyword here is variance; if they are newer or not (in low level leagues only) is hard/impossible to show and the only indication would be the variance).
The only correct statement from above is that we were looking at the data and could conclude that on avg terrans (also in low league) dont play fewer games.
Again: I gave you a number of POSSIBLE reasons why terrans have lower average mmr (like 1.5 years ago). That included that they might play fewer games, are newer to the game, are more casual (in the sense of only picking the most familiar tutorial race, terran, and never learn anything else in that game, thus sticking at low leagues) and gave you reasons why i think they could play fewer games, are newer to the game or are more casual. Again: this has nothing to do with the TERRAN POPULATION at a whole. Its only about those guys in the bottom 2 leagues (bronze and silver) and thus skewing the result for terran.
i dont believe for a second that someone can be stuck in bronze being stuck means you first needed to go out and try.
The correct data to use would be a player survey asking how casual a player is.
ie. do both a slide scale question “On a scale of 1 to 5, how casual would you rate yourself?” and a “When you play ladder, what are your aims?” with answers like to climb, to get better, to have fun, to play with friends, etc.
You know, we’ve been down this road before.
The difference between “the race appeals to new players” and “the race’s population consists of newer players” is one you refuse to acknowledge.
And until you do, there’s no hope for you understanding what everyone has said to you.
Fine. Let’s say it does. You know that finding out the player base was newer would be damning evidence that you were right. You know that finding out the player base played fewer games would be damning evidence that they’re casual.
And the fact that the opposite of these two means what?
There’s simply no possible way they can be a “newer population” The percentage of Terran players has grown, while the overall player base has shrunk.
So, then they’d be joining the game new, getting promoted out at roughly the same time they’re switching races to a new race, then quitting the game (remember the entire player base shrunk, so there’s play turnover across the entire player not just the Terran players).
So, IF newer players pick Terran and the player base of Terran players are newer, that means they just randomly got promoted as they switched from Terran? Sure sounds like it’s a more difficult race to me if that’s the case.
Maybe it appeals to new players, but those players haven’t been “new” for a decade.
But they dont play more games (just as much) and if there is a portion that is newer (you know people starting the game and picking terran as their starting race) is yet to be proven/disproven; could be shown by variance, but i dont care since it COULD BE another factor. But basically the assumption that there are more new players is synonym with terran being the starter race. Its an assumption, but a very likely one because of the reasons we discussed like 1000 times which are also like confirmed by lead sc2 designers. And its not like a controversial statement.
This is basically what its all about:
You come up with a claim that cannot be proven: “Terran is harder because of lower avg mmr”
And we come up with dozens and dozens of explanations why that can be. Somehow they are all wrong/cant be proven therefore “obviously” also wrong. But your point still stands in your eyes, despite you have 0 proof or actual reasons to believe that lower mmr means harder.
I apologize that this treads such similar grounds to two of my posts ago,
But no. I would not take this as confident evidence on its own. I wouldn’t even take it in context as particularly evident of my claims overall, because -
As I have said numerous times, there are multiple weird and interesting things about the data that we have. The MMR bubbles don’t make sense with what we know because they deviate ridiculously from standard bell curves, oddities that are very unlikely to be linked to a single cause.
Which, again, I’ve spent multiple posts writing out ways in which multiple effects uniquely more influence Terran in this regard; but also I have pointed out more distant in the past how that doesn’t really add up to the levels of deviation we see.
What that means in turn is that even if one more factor tips to the other side, there’s still too many outliers for me to be ‘happy’ with that as a conclusion - because of the stated base assumption that I have, that the game is mostly balanced at most play levels.
The reason I hold this belief against what you say is evidence to the contrary is because when you look at that same evidence from other angles, the data just as much say other weird statistical things – see earlier comments about There's just a ridiculous over-population of Bronze Terrans and how that doesn’t correlate to the effect that wheasy asserted of If a player switches off of Terran they'll gain 300 MMR very fast.
This isn’t how that works.
What the fact you outline says has multiple possibilities. For example, suppose the following race distributions between seasons:
+ Ter | Zer | Pro
Pop | 100 | 100 | 100 | 300 = players in S20
Pop | 105 | 95 | 95 | 295 = players in S21
Now, the player base has shrunk by about 1.7%; and Terran’s representation went up from 33.33% to 35.59%. The other races shrunk from 33.33% to 32.20%.
There are a ton of ways for these numbers to be reached.
For example, 10 players of each race quit, with an influx of 15 new Terrans and 5 new Protoss and Zergs.
Or only Z/P players quit or swapped races.
Or each race got 10 new players, but only 5 Terrans quit compared to 10 Protoss and Zergs.
And these are all visually going to result in the same data.
This is why I keep harping on that the data do not say what you assert they say and why I kept implying the distinctions that I was drawing.
Of the limited number of new players, a huge number of them pick Terran.
The number of players who are new to the game who are bad are very high.
The number of players who stick out the game’s learning curve are very likely to switch races as they learn information, which means there’s a decent chance that the mid ranks will have a disproportionately small number of Terrans - while as they continue to learn and excel they are also likely to switch back to maining Terran, creating a drop-off for the other races.
That entire ladder of events leads to the same conclusion you drew! The average Terran - due to the declining general population - has played longer. It fits the data of the MMR brackets that we have looked at.
But it also shows an explanation for the oddity in Bronze to Silver ranks, as story-based as it is; that lines up with all the utterances from the data.
And, most importantly of all;
I’m not saying this, have not said this, and probably will not say this!! As Miro said;
You are continuing to not acknowledge the bullet that’s being said. It’s not that they’re newer! It’s that, to new players, they have high appeal. Because new players are likely to be bad players, a wildly disproportionate number of new-and-bad players, the players who come and go after just a bit, are Terrans.
The players who stick around and improve are likely to switch races - but these players are also plenty likely to switch into Terran or return to Terran.
I’m not saying “Terran’s population is newer”, I’m saying “The new population is significantly more likely to pick Terran over the other options”.
Do you see how this is frustrating; that I am having to clarify something that Miro made very explicit, in this second instance ten days or eighty posts prior?
Re-ordering the words does change the meaning, and I pick my words carefully precisely because I’m trying to not be misunderstood.
And everyone else in the thread has understood that meaning! Please stop putting words in a poor girl’s mouth.
Not at all. Sounds to me like players who are bad are less willing to experiment with all the races, because they want to maximize their familiarity with their tools, get a feel for the things they can do, instead of learning a whole new set of strengths and weaknesses - at least, until their game knowledge is shored up.
Harstem in particular has made this comment many times, that switching races - for him - makes him feel like the other race plays itself, because of his strong mechanical foundation already being present. He then goes on to make pretty simple mistakes, but is exceptionally mirthful and relaxed - and frequently makes observations of things that he can bring back to his Protoss.
That’s the phenomena to which I refer to. It’s not harder - it’s that switching off what you started with changes your paradigm and makes you think differently, realize new things and that process requires a certain degree of skill to begin, and some people don’t ever end up bringing it back, but keeping with their switch.
I’ve decided to go with stories not because facts clearly don’t work but because I am just tired of trying facts.
If we’re going to be at an impasse I will happily remind you we are at one.
You equate two things that are not equal; and the answer to your sting is not only present in my post but is the part of my post you quoted.
‘The average Terran is not a new player’ and ‘Terran appeals to new players’ are not mutually exclusive statements, just like how ‘the race favored by tutorials’ and ‘veterans can play the tutorial race’ are not.
You don’t have any facts. You have stratification of old facts. Stratification that only proves anything because you want it to.
If you have another fact, please list it.
I’m more mocking than I was ever mad. That’s why you get angry and call me names, lol.
Yes. I know. You’ve come up with explanations. Then when those explanations were proven wrong, you came up with other explanations. It’s called “goal posting.”
being stuck means you first needed to go out and try.