And that’s dedicated terran players not silver or bronze who bring down the winrate. Deductive reasoning is a gift wasted on you I take it
Where’s your data on this ? Oh that’s right you don’t have any
The only retard here is you unfortunately as you have shown
It’s EVERYONE. EVERYONE. ALL TERRAN PLAYERS COMBINED. Not “dedicated Terran players.” Not “higher level Terran players.” It’s ALL of them.
This isn’t relevant.
What I was clarifying is more instinctive – If I put Starcraft in front of a child particularly, but just anyone who hasn’t played or seen it before, my personal experience with all sorts of people indicates that they’re most likely to “get” how Terran works.
The familiarity, or rather the real-life analogy of “you have to do work to get tasks done” - makes the build pattern of Terran structures intuitive.
The simplicity and basic-ness of “you pay money, wait, and get unit” being the only production method in Terran, also, makes it easy to intuit.
I’ve seen actually dozens of people get confused on how to handle unit warp-ins, because it’s different and that adds a significant learning curve.
Where’s your data to refute it?
Nowhere? Okay, that’s fine.
You can say that it doesn’t exist - but you don’t have to be a dick about it and it’s not like your data actually refutes the point that was said.
We can draw a reasonable conclusion from the following:
Given that there are 79150 Terrans, but 64700 Protoss and 58575 Zerg - 202,425 players -
Therefore, for a given vX match, there’s a 39.10% of vs T, 31.96% vs P, and 28.94% vs Z. This is a 1% discrepancy from the data I can find; after discounting them having Random.
Therefore, let’s grab win rate stats from sc2pulse, aligulac, and nonapa; chosen because of the initial results they’re the only ones I could find these data in less than 10 minutes.
Horizontal rows indicate the match winner.
Aligulac sc2pulse
| P | Z | T | | P | Z | T |
P | 50.0 | 50.8 | 51.4 | P | 50.0 | 48.5 | 52.1 |
Z | 49.2 | 50.0 | 48.0 | Z | 51.5 | 50.0 | 49.8 |
T | 48.6 | 52.0 | 50.0 | T | 47.9 | 50.2 | 50.0 |
nonapa only has region divided statistics
EU P Z T # KR P Z T # AM P Z T
P 50.0 47.9 52.7 # P 50.0 49.0 52.7 # P 50.0 48.7 53.2
Z 52.1 50.0 50.5 # Z 51.0 50.0 50.3 # Z 51.3 50.0 50.4
T 47.3 49.5 50.0 # T 47.4 49.7 50.0 # T 46.8 49.6 50.0
Well, those all agree with one another pretty closely. The worst outlier (American Terran vs Protoss) is not too outstandingly different.
So since they’re close, for simplicity, I’m just going to use sc2pulse’s numbers because aligulac’s excludes low ranks and averaging nonapa’s stats is not just (x+y+z)/3.
If each player played one game a day (obviously incorrect); we’d see the following match up frequencies.
64700 protoss; 58575 zerg; and 79150 terran.
Picking a random match,
39.10% vs T
31.96% vs P
28.94% vs Z
20,675 P v P. 18,725 P v Z. 25,300 P v T.
18,720 Z v P. 16,950 Z v Z. 22,900 Z v T.
25,295 T v P. 22,905 T v Z. 30,950 T v T.
Those do line up nicely - despite lots of accrued rounding errors we still see a maximum discrepancy under five (I rounded for cleaner reading), so we can use it to judge how many matches would get played if each player played once a day.
It’s TvT+PvP+ZvZ+TvP+TvZ+PvZ; or 30950+20675+16950+18725+25300+22900, which is 135,500.
The actual number of games played per day is around 55% that. So the ‘average player’ plays one game every two days (or, more likely, plays 3-4 games on one of their days off).
If you look at “How many XvY matches got played” statistics?
| P | Z | T | R |
P | 931344 | 875951 | 1203017 | 252165 | = 3,262,477
Z | 875951 | 905042 | 1107328 | 242143 | = 3,130,464
T | 1203018 | 1107329 | 1677088 | 346181 | = 4,333,616
R | 252165 | 242143 | 346181 | 81342 | = 921,831
Terrans play way more games when they play. Way more. In 90 days, our statistics earlier would suggest that we would get game counts of -
| P | Z | T |
P | 1,860,750 | 1,685,250 | 2,277,000 | = 5,823,000
Z | 1,684,800 | 1,525,500 | 2,061,000 | = 5,271,300
T | 2,276,550 | 2,061,450 | 2,785,500 | = 7,123,500
Dividing the matches by the player counts said earlier,
P: 50.42468
Z: 53.44369
T: 54.75194
So Terran and Zerg players just play more games on average; and since there are dramatically more Terrans than anyone else this effect is huge.
Also, if you go pull up the “how many games got played” statistics? Terran plays the most by a significant margin up to Diamond.
You can’t quantify this in any way whatsoever. It’s just another attempt to make this entire process as subjective as humanly possible.
ARE Terran players significantly “harder working” than the other 2? No. Not really. In would go so far as to say that the disparity is almost negligible.
What I AM trying to demonstrate, is the idea that Terran players are not somehow inferior. The common explanations of “Terran players are newer” or “Terran players are more casual” are refuted by the fact that:
- They’ve actually been playing slightly longer on average (the % has gone up whilst the player base has gone down).
- They play slightly more
Man you guys will never change. Keep being annoying forum whiners while misconstruing statistics, empirical data and basic reasoning. Eliwan and even batz have both given some basic breakdowns. Terran may indeed be more try hard and they are actually rewarded by being higher up on the ladder overall after gold, both in absolute nunbers and percentages. The fact that there are thousands of less than 10 game players playing terran in bronze and silver shouldn’t matter. Or even worse, a thousand game terrans stuck there.
You don’t accuse a car manufacturer if a crazy guy gets drunk and goes a murder spree with that company’s car do you?
Tldr: less qq, more ggwp
average
ăv′ər-ĭj, ăv′rĭj
noun
- A number that typifies a set of numbers of which it is a function.
- An intermediate level or degree.
- The usual or ordinary kind or quality.
The average is meaningless is the point. If two people have 100 and 0 dollars respectively, their average is 50 but it doesn’t represent either guy’s wealth range.
Median is far better when there is a large spread especially when there is good reason to discard the bottom feeders
Yes, yes, there’s always a reason to discard data that doesn’t conform to your desired conclusions.
And of course, a sample size of 2 is just a relevant as a sample size of 220,000.
1 Like
It literally doesn’t matter, even id they were 6 gorillion, when they can’t play the game.
Plat 5 should be the minimum benchmark
I like how you get to determine which stats are and are not significant. Neat.
Not to mention you arbitrarily chose the point at which the point at which the trend reverses, lol. Nice. Glad you didn’t apply any bias.
2 Likes
Yeah, it’s the truth buddy. Sucks to suck
Kinda hard to take a 100m runner seriously when he says he is competitive and does the distance in 20 seconds you know.
Same applies here. L2p
It’s the truth that you just arbitrarily decided to deem half the statistics as irrelevant with source : “trust me, bro”
If someone is seriously trying to argue this I really hope they catch the drift that it doesn’t hold weight. Like… a person is a person. Damn.
I’m sorry, I’m just getting caught up on this - Have you actually encountered people who should have purging fire rain upon them like this? I’m so sorry.
This is completely irrelevant to the point that 12402 is making. The median average is a good floor if we want to make statements about how the game suffers a significant balance issue, because at the bottom we’ll see people who barely know half the game and at the top we’ve already culled to “the people who can play at that level despite imbalance”, in a sense.
(Except in that case; balance issues for sc2 are usually catered to the upper side of play because, unsurprisingly, higher end players are usually the most affected by changes. But generically, it’d be sound.)
However, we’re talking about something where cutting off the ends doesn’t make sense, because we’re not looking at the bell curve distribution where the edges are outliers, we specifically want to look as close to the outliers as possible because they’re relevant to our discussion topic.
Additionally, while the mean average constantly gets data wrong, that doesn’t mean that the mean average is useless. All statistics might be useful, and actually looking at them in good faith is…just necessary.
Well, it’s the concept that all Terran players are lower average MMR is explicable by something so individual as “skill.” That would mean that inferior players all just randomly chose Terran.
There’s tons of theories that fly in the face of what every pro player says (that Terran is the hardest):
-
“Terran players are newer.” I’ve demonstrated multiple times that Terran players are roughly 6% more senior than the mean. Players like Sentry simply ignore these numbers and repeat the same arguments verbatim.
-
“Terran players are more casual.” This bring back my previous demonstration that Terran players play slightly more. Then of course they select the idea that you can play more and be more casual, which just makes it incredibly subjective yet again.
2 Likes
(This is a generic reply, not to anything in particular; because whee, statistics.)
The data that OP posted and basically all data I’m able to find indicates something not dissimilar to what it contains - That Gold, Silver, Bronze more Terran representation than is justifiable as “natural variance” - Blizzard’s stated LotV goal was to make the population distribution of the ladder be around 4% for master and bronze, with the remaining 92% taking a rough split across Diamond through Silver.
In an ideal world, ie. the game is perfectly balanced and we don’t have giga-outliers with 80% win rates skewing ladder stats, we should see - using the OP’s player count of 202,445 and a range of 5%:
8098 ± 405 players in Bronze/Master (7,693 ~ 8,503)
46,562 ± 2328 players in each other tier (44,234 ~ 48,890)
And instead, we see
Master: 8,808 (+300 from upper bound, 108.77% of middle)
Diamond: 48,864
Platinum: 43,856 (-500 from lower bound, 94.19%)
Gold: 47,000
Silver: 42,223 (-2000, 90.68%)
Bronze: 11,155 (+3000, 137.75%)
And if we look at that same data:
Of 11,155 Bronze players, 2434 are Zerg, 3228 are Protoss, and 5493 are Terran.
Of 202,445 StarCraft 2 players, 58,577 are Zerg, 64,715 are Protoss, and 79,153 are Terran.
4% of those player counts (Bronze/Masters) would be 2,343 Zerg, 2,588 Protoss, and 3,166 Terrans.
23% of those counts (otherwise) would be 13,472 Zerg, 14,881 Protoss, and 18,204 Terrans.
So there are, for each race, the following imbalances in the leaderboard:
| Zerg | Protoss | Terran | Sum
Master | +233 | +486 | -8 | +711
Diamnd | +3092 | +532 | -1317 | +2307
Platin | +1103 | -966 | -2838 | -2701
Gold | -568 | +492 | +519 | +443
Silver | -4095 | -1379 | +1140 | -4334
Bronze | -91 | +640 | +2327 | +2876
Sum | -326 | -195 | -177 | -698
Which outline that, for some reason, if you pick a random Terran, they’re significantly more likely to be below Gold than if you pick a random Protoss or Zerg.
The reason I wanted to share the data in this form is because it shows that this “cliff” effect does exist for the other two races - just something neat to know.
While anything could be a significant reason, because there are obviously dozens of explanations and there’s no way any one of them is the only one, and it’s even more likely that all of them are partially true and contribute to the reason the observed data are the way they are; which is why I’m so embittered about this affair. There’s not really a demonstrable way to say, one way or another, why the data are that way.
For example, my personal experiences of getting people to try playing RTS lead me to suggest that ‘People who struggle with RTS’ has a high correlation with ‘People who stick with Terran’, because Terran’s defensive tools let them not be overburdened mentally, and because the campaign was hard enough so they don’t want to learn a new race.
But that holds no water, for rather obvious reasons.
I do think this is a true idea (that playing a lot does not automatically make you not a casual player); but it is definitely not a rebuttal to the statement you’ve made.
1 Like
that holds no water even though it’s the observable reality I experienced
Oof
Here, I’m gonna help you out:
anecdotal
ăn″ĭk-dōt′l
adjective
- Of, characterized by, or full of anecdotes.
- Based on casual observations or indications rather than rigorous or scientific analysis.
- Pertaining to, or abounding with, anecdotes.
Statistically, it is fairly obvious that a sample size of 100 does not make good data.
So, yeah, while it’s the trend I observed personally and that I got information about; if I extrapolate the implication from that data; I get something that is hilariously distant from reality.
So it holds no water, because even though I’ve personally interacted with a hundred people who found playing SC2 hard, and something like 50 of them picked Terran to ladder with; that’s statistically irrelevant and could just as easily be explained with something like “Twenty of them played Age of Empires and Terran works like Age of Empires does”.
Which feeds back into the earlier stated familiarity, because most RTS have a faction that works, approximately, like Terran or AoE does. It’s just one of the intuitive ways for building things to work.
But that should not cause a dramatic effect as players learn to play the game. The effect should vanish as players play the game for more hours, as the player’s need for intuitiveness will decrease as they become familiar with the RTS systems overall; and the effect should also be diminished as players become more familiar with the specific race or “how long have you played that game”, both of which are something that 12402 has pointed out are true.
If i might add few words to the discussion i would say:
I was hard-stuck in Diamond 3 for the last 2 years. Last week finally i got promotion to Diamond 2 which isn’t any accomplishment at all in competitive 1v1.
In order to improve i am practicing:
-apm trainers
-multitask trainers
-build orders
-analyzing my own replays.
I feel like terran mechanics are a joke. You have to manage so many things. Produce untis from 4 facilites - CC, rax, fact, ports. Micro two armies at the same time. And on top of that you need keep an eye on for such bull crap like raising/lowering depos so that your base doesn’t get overrun by lings or zealots, repair turrets in case of muta fly-by and so on. It is literally the most unforgiving race. One mistake snowballs into a situation where you cannot win.
I am working hard to improve my MMR, but my progess is so slow it’s laughable. Do i suck or is this race really hardest to improve (at least in diamond league) ?