I think a simple quality of life buff that could help terran would be to allow orbitals to be able to garrison 5 scvs. It ain’t much but it could help. Main reason terran feels too easy to get stretched thin is in general the lower movement speed units that they rely on to mass or defend with, many of which are ranged or are high value targets like siege tanks and they need a more composite and harder to use properly army, even in smaller groups.
Compare this to say roaches queens or ling bane or stalkers zealots with blink/charge and shield batteries.
They are difficult, but keep in mind it’s not really a strategic race. It’s a mechanical race. If you removed the mechanical difficulty, you’d be left with by far the easiest race in the game.
Couple of things they did that peeved me:
- Making the SCV more easily selectable (really? you nerf macro for every Terran build because Zerg players had to deal with a proxy once out of every 5 or 6 TVZs?)
- Not allowing players to be able to select unit priority (reaction times for stim are super important. Having to click TAB because your bio ball prioritizes Ghosts is a HUGE disadvantage.
- Making mech unviable (or 67% unviable) you really can’t make it work in TVT any more after the Raven nerf.
You’re creating your own averages buddy just stop
You’ve shown nothing , your math doesnt math buddy and you’re twisting the numbers to fit your narrative. Sad
… No?
±5% is a very standard statistical interval.
In an ideal world, when there are X choices, each choice gets 1/Xth of the population.
These are both extremely basic statements.
Blizzard outlined, directly, in an article - the Patch 3.4 ladder revamp - the statistical bounds that I pasted.
Everything else is your own statistics.
So, point-blank, you’re saying that people aren’t allowed to do what you did, except actually type the math out for people?
Because I have done no twisting; you did that and then insisted you were right about it (:
I’m glad to know that you agree with everything I’ve said. That statistics themselves can lie, and that showing numbers is not proof.
Except the cliff effect doesn’t exist in the magnitude as terran.
Look at what blizzard lays out already - they want the distribution to be this way
| Masters | 4% |
|---|---|
| Diamond | 23% |
| Platinum | 23% |
| Gold | 23% |
| Silver | 23% |
| Bronze | 4% |
But in reality it’s this way
| Zerg | % amongst rank | Protoss | % amongst rank | Terran | % amongst rank | |
|---|---|---|---|---|---|---|
| Grandmaster | 147 | 0.25% | 210 | 0.32% | 182 | 0.23% |
| Master | 2576 | 4.40% | 3074 | 4.75% | 3158 | 3.99% |
| Diamond | 16564 | 28.28% | 15413 | 23.82% | 16887 | 21.33% |
| Platinum | 14575 | 24.88% | 13915 | 21.50% | 15366 | 19.41% |
| Gold | 12904 | 22.03% | 15373 | 23.75% | 18723 | 23.65% |
| Silver | 9377 | 16.01% | 13502 | 20.86% | 19344 | 24.44% |
| Bronze | 2434 | 4.16% | 3228 | 4.99% | 5493 | 6.94% |
| Total | 58577 | 64715 | 79153 |
What this shows is that Terran has it significantly harder than the other races
Okay, that’s fair. I was not specific, because I was referring to that, not where, it exists.
As purely an example sentence, “Protoss have a hard time making it to gold, but once there climb fast.”
This describes the same thing - What I meant with “this cliff effect” was that that there’s a point at which there’s a weirdly high number of representatives for each race at some specific league.
Which I found interesting; so I wanted to leave the table I made in that form because it made that “neat” point easier to see.
It shows that something is weird and probably suspicious; since it shows a very large population-proportion relative underpopulation of Terrans in Masters and Diamond.
It also says similarly that something’s weird about Zerg that causes it to pull aberrant numbers in Diamond, Platinum, and Silver.
1 Like
It’s stupid i agree, but for me this one actually isn’t big of a deal. Just boxing over the building selects constructing SCVs.
Yes this one is very annoying. Sometimes you can lose the game just because of that especially if you miss that tab button (pressing CapsLock accidentally)
I don’t play mech so this isn’t any problem for me, but Raven’s actually were buffed (cost and production time) and mech pretty much works well in TvT all the time.
What is problem for me is that i hate building placement in my main base - every structure needs an add-on. This is stupid - not only your production is crippled if you forget to add one, but also later units get stuck in between add ons if you mess building placement.
Another problem which i struggle with is simultaneously micro your army WHILE managing stuff at home - putting extra barracks, adding depos, transferring workers form one base to another etc. this in my opinion is the hardest skill to learn in the entire StarCraft 2 - no matter what race you play.
Also i believe i am losing most of my games in between 5-8 minutes. My first 5 mins are almost flawless - never get supply blocked, put every structure i need, scout with my reaper etc, but then i make a push between 5- 7 min and this is where i forget to do a lot of stuff at home and when my push is over then i end up having supply block, not enough rax to produce units and i lose to counter attack - so this is something i really need to work on.
1 Like
It’s not for selecting the SCV, it’s for selecting the buildings and accidentally hitting your SCVS. It’s just a mechanical nerf that I’ve never seen displayed with the other two races, with the Race that’s already by far the most mechanically challenging.
Yeah it’s even better in TVT, now (not that I’m sure that’s a good thing). But you have to have bulk to your army, whereas the turret provided that before, you now need marines. It’s just a stylistic issue.
Sad someoen should do a data collection on paths to victory for toss and zerg
Zerg + Toss has many more paths to victory as their unit tree is more robust and flexible. Blizzard is a joke
Bumping this thread with stats
Bumping this thread again for friends what a crap game
1 Like
Why so desperate?
Stop bumping a thread no one cares about.
2 Likes
Bro, this people don’t understand about reasons. The last patch just have proved that too… Whit Terrans is imposible to win vs a little more average P or Z player, and it’s incredible how much advantages should P or Z players should have justo to have a chance to win vs real pro terran players as Maru.
Game design still being lame, while in SC1 the maps are designed to stop payable creep… in this absurd game there’s non way to keep in site free creep… while toss should be able to warp only inside a nexus radious.
this guy should a be a terraplanist… A game based in pure math to works… just he is saying that statistics don’t even have something to do whit a pure math game encore.
I have absolutely no idea what this post is supposed to mean.
I have attempted to respond to explain why the post I made is relevant, but I don’t… actually know what you’re saying here.
The thread’s original post is about using statistics as though it is proof of a thing.
A long-winded and not-that-useful example of massaging statistics
For example, the breeding ratio of humans is around 105 males per 100 females; and yet in eg. China, the ratio from 1990 to 2017 was significantly higher (ranging from 110 to 118).
Specifically, for every 100 female births, there were this many male babies - source ourworldindata.org/gender-ratio -
1990 | 1991 | 1992 | 1993 | 1994 | 1995 | 1996 | 1997 |
111.87 | 112.50 | 113.14 | 113.75 | 114.35 | 114.91 | 115.47 | 115.97 |
1998 | 1999 | 2000 | 2001 | 2002 | 2003 | 2004 | 2005 |
116.40 | 116.78 | 117.10 | 117.37 | 117.60 | 117.77 | 117.88 | 117.94 |
2006 | 2007 | 2008 | 2009 | 2010 | 2011 | 2012 | 2013 |
117.94 | 117.88 | 117.78 | 117.62 | 117.41 | 117.14 | 116.79 | 116.41 |
2014 | 2015 | 2016 | 2017 |
115.97 | 115.47 | 114.91 | 114.35 |
Now, we can express this same information in a very, very large number of ways, which will seem to say different things.
We could take this and massage it positively:
In 2005, across China, an astounding 12.3% more male children than average were born.
Sympathetically:
In 2005, across China, a devastating low number of children born were female: just 45.88%.
Or damningly:
In 2005, across China, parents selected male children with such strong preference that around 2 million female children were denied life.
Each of these does something weird to the numbers -
Uses positive language; value is from 117.94 / 105; a basic “actual / expected.”
Contextualizes as negative, value is from 100 / (117.94 + 100); proportion of whole population.
Assigns blame; value is (117.94 / (105 / 100)) - 100; how much more would be needed to match the original proportions… made to a number by dividing by 100 and multiplying in a new statistic, - that there were around 16,170,000 births in china in 2005; source en.wikipedia.org/wiki/Demographics_of_China#Vital_statistics; then rounding to a convenient number (from 1,992,760; exaggerating the statistic by ~0.36%).
Ergo the expression, “Lies, damned lies, and statistics.”
The reason for that being relevant; though, is that like any time one pulls statistics out, they give you information about something being the case.
However, there is often a missing step: Proving the correlation between the hypothesis and the statistics.
Frequently (as demonstrated in this thread), the statistics themselves are taken as proof of the hypothesis. And… they aren’t. There’s frequently a large number of factors that contribute to a reality, not just one or two.
If you want to abstract SC2 to a game of math, though, that’s fine, but also not relevant, because we aren’t talking about the game as it is played at the moment: Despite the thread title basically being Z/P are easy to play, there is no argumentation to that, there is only a collection of numbers about player placement in leagues.
And until demonstrated otherwise, those are tangential; like the X and Y axes on many charts. They might have a correlation, but the correlation is obfuscated by the hundreds of other factors that are not present nor discussed – and it is the fact of their absence that I was pointing to.
Also, hardly relevant, but female, she.
1 Like
People love to use this line when statistics suggest something that they don’t want to admit 
So, we have Terran players in significantly lower MMR brackets on average. This, in and of itself is not damning, but it’s a start. To move from correlation to causation, we then have to eliminate other potential variables.
One of the common explanations I often hear is, “Terran is more popular among newer players.” A fair assumption, given that it’s more visible in the trailers. However, the statistics show that Terran players have actually been playing the game longer, on average, than other players (as evidenced by the fact that the total player base has shrunk, while the Terran player base grew, slightly, as a percentage).
The next argument is that “Terran is more popular among casual players.” Usually this is the one I hear after I’ve debunked the first argument (effectively just moving the goalposts). To this, I point out that Terran players are slightly more active on average (play more games/ any given time period). Is it a perfect measure of what is and isn’t “casual?” No. But it’s the best metric we’ve got.
If you have any other explanations for why the Terran MMR slump is so drastic that amounts to something more than “Terran players are all just trash.” Then I would like to hear it.
All this said. There’s no “easy” race in Starcraft 2. There’s the Hard race (Zerg). A VERY hard race (Protoss). And the most difficult race (Terran).
1 Like
I mean, Eliwan isn’t wrong; statistics can be manipulated to say just about anything in the right context.
4 Likes
People also like to say pithy things like this instead of actually addressing the core of the point, even when I try to say that I don’t even disagree.
Like - specifically - I agree with the refutations you’ve outlined here - they are sound, thought through, line up with evidence, and have no parts that are obviously lacking - But the conclusion “Terran is the hardest” is not actually proven, but the original post is arguing that it is proof-positive from its evidence.
Hence my first two short posts; that the analysis takes the conclusion as true and the argument of ‘the data matches this hypothesis’ does not a proof make.
My point is that if one is winning less, or has an overall lower MMR, that does not actually prove that one’s choice of what to play is harder.
It can mean a very large variety of things, one of which is that it is harder! But “can”, not “does”.
1 Like
hey clown youre still talking? how about you show some data before you talk ? oh yea you have none
Yes we don’t have a silver bullet causational data but we have a good enough correlation. Absence of evidence is not evidence of absence/ Stay ignorant my friend
Apparently, I didn’t do anywhere nearly enough, if this is your reaction.
Please pardon me while I break this down because, you /need/ a lesson:
An ad hominem is when someone attacks the person making an argument, instead of their argument. At no point did I say anything like wheasypeasy is stupid just ignore them* -
I did, however, absolutely make attacks -
But not one of these is addressed to you the person making the argument - It is an attack on the argumentation, that the evidence you offer at no point is demonstrated to have a perfect relation with your initial statement.
This was required because the basis of argumentation is fallacious, specifically: The cause of the data does not have to line up with the conclusion being drawn. Most of these types fall under questionable cause’s umbrella, fallacy-wise, the two I think are the most applicable being jumping to conclusions and/or over-simplification.
* And, note, this example itself is very accusable as a fallacy - I’m not giving what is being said a fair shake, by giving a ridiculous example - strawmanning an argument. You’re not saying that I’m saying that.
There is an excellent expression:
You are not immune to the effects of propaganda.
And it applies to everyone. Nobody is truly immune to these things, and that’s why awareness and paying attention are so important; and why people pay so much money to try all sorts of things to sway public opinion.
But, please be aware that addressing someone as sheeps is going to automatically put them against you, because you’ve elected to assert that their intelligence is lacking. Nobody likes this, nobody responds to it well, and this makes it more likely that they will not listen to you.
At no point is this fact related to any of the data you have offered.
Even if we assume the truth of this, it does not actually inherently prove the sentence that precedes it.
I do not need new data because what I am talking about is your argumentation of the data that you provide.
At no point have you proven anything. That is the only thing I am saying and will continue to be the only thing I say because it’s the part I care about.
Evidence is not proof, and data is not even evidence by itself: The data needs to be demonstrated relevant.
Right now, what you have presented is data and a hypothesis. It’s even verifiable data, it’s even a reasonable hypothesis, it’s even a hypothesis that lines up with the reality presented by the data.
But that is not a proof.
Neither of these apply here. To the point that I was extremely confused and had to go read back up on these to make sure I didn’t just misremember them.
“Absence of evidence” - that is, “we don’t know X isn’t, so X could be”, in its most extreme state - is hopefully obvious why it isn’t relevant
and “Evidence of absence”, or “we expect Y from X and Y didn’t happen so X is false” - is not really what’s happening either. Let me demonstrate with -
We expect the population of a race to be approximately equally distributed across ladder ranks if the game is balanced.
And then we don’t see that, so our conclusion would necessarily be that the game is not balanced.
Except we didn’t actually prove that. We demonstrated only that the idea has merit, that it is a possibility.
We have made just as much proof for the seemingly wholly unrelated claim of “Each race has a different skill floor”, basically none, but both of these claims line up flawlessly with the data and don’t require any massaging to show it.
And that’s ok, that’s fine, that’s good, we’ve demonstrated a plausibility. That’s an excellent starting point to anything.
2 Likes