|
I was curious to see how MMR is distributed on the sc2 ladder. Here is the result (data are from rankedftw.com):
![[image loading]](https://i.imgur.com/2RuRS3H.jpg)
I was expected a similar distribution for all races but it is not the case. The interpretation of this result would be adventurous and I personnaly don't want to speculate on it.
Method:
1. EU ladder only
2. Only players with at least 10 1v1 games in season 44 are taken into account
3. distribution was first computed using 25 regular bins for each race
4. distribution was then interpolated with a regular interval of 10 MMR
5. [edit] The percentage is computed with respect to the race population, not the entire population.
|
Ohhh, love me some statistics! Could you also post a version with a logarithmic y axis? Would be cool to see the tail end behavior of the distributions
|
Interesting stuff! Not surprised to see zerg has the highest on average, because I'm pretty sure it's the least popular race among casual/low level players
edit: ((i say this as a pretty low level player myself))
|
On September 12 2020 00:39 Meadowlark wrote:
Interesting stuff! Not surprised to see zerg has the highest on average, because I'm pretty sure it's the least popular race among casual/low level players
edit: ((i say this as a pretty low level player myself))
where did you get this info from? It seems every young upcoming player on eu is zerg, based on your suggestion it should be terran..except...
|
Where is the general cutoff for masters on here?
|
Well done 99percent, not the least bit surprising to me. It's a straight up descriptive stat, and it's beautiful. I love the succinct no-fluff style of the post. Very brave of you to share.
On September 12 2020 00:39 Meadowlark wrote:
Interesting stuff! Not surprised to see zerg has the highest on average, because I'm pretty sure it's the least popular race among casual/low level players
edit: ((i say this as a pretty low level player myself))
From a stats perspective, the number of players is completely irrelevant to the MMR spread because this is looking at proportions. If it's true that Zerg is the least popular race, which it may or may not be, the argument would then be, "Players who choose the least popular option are more talented as a population." While this appeals to my hipster sensibilities, it's pretty laughable scientifically.
It'll be interesting to see how people spin this to protect their beliefs.
|
Mexico2170 Posts
So this means that on average most of the population is on plat/diamond?
Before it used to be tat if you were in diamond you were top 10%, master top 2% and GM top 200.
Now I feel bad.
|
This has been consistently true for all of LotV (and regardless of whether Protoss or Zerg was the least popular race. Terran has always been the most popular).
The obvious explanation is that when both players are bad and a-move their armies at each other with not much micro, zerg's the better race since it has a bigger macro focus.
On September 12 2020 01:50 [Phantom] wrote:
So this means that on average most of the population is on plat/diamond?
Before it used to be tat if you were in diamond you were top 10%, master top 2% and GM top 200.
Now I feel bad.
The average player in LotV is right on the border between gold and plat. Masters is top 4% and Diamond is the 23% after that.
|
On September 12 2020 01:26 ThunderJunk wrote:Well done 99percent, not the least bit surprising to me. It's a straight up descriptive stat, and it's beautiful. I love the succinct no-fluff style of the post. Very brave of you to share. Show nested quote +On September 12 2020 00:39 Meadowlark wrote:
Interesting stuff! Not surprised to see zerg has the highest on average, because I'm pretty sure it's the least popular race among casual/low level players
edit: ((i say this as a pretty low level player myself)) From a stats perspective, the number of players is completely irrelevant to the MMR spread because this is looking at proportions. If it's true that Zerg is the least popular race, which it may or may not be, the argument would then be, "Players who choose the least popular option are more talented as a population." While this appeals to my hipster sensibilities, it's pretty laughable scientifically.
That's why I specified the low-level/casual demographic. If zerg is less popular than the other two races amongst casual players, but, say, equally popular amongst more hardcore players, then it will have a higher average mmr because the zerg playerbase will skew more towards the hardcore than the casual. The point has nothing to do with talent--mmr has much more to do with your approach the the game and how much time you allocate to play. It really wouldn't be surprising if people who allocate more time and consciously focus on improving in the game are a meaningfully different population with different preferences than people who like to mess around and have fun on ladder every so often (10 ladder games per season is a pretty low floor). I can't remember where I read this, and so maybe I'm full of it, but I do remember reading that zerg was overall less popular in the lower leagues. Obviously this is anecdotal, but my not-very-high mmr laddering experience definitely jives with this--I run into far fewer zergs than terran and protoss. I can't really prove this, but my intuition is that the mmr disparity is mostly caused by how different people who are looking for different experiences with the game may choose their races by different criteria. People more interested in game mechanics might find all three races equally interesting on account of their unique mechanics, whereas people less interested in game mechanics might on average prefer space marines and shiny aliens to bugs. That's my thought, at least. Again, definitely could be factually wrong bc I don't remember where I read this, but I really don't think it's logically implausible in the way you made it out to be.
edit: and to be clear I really don't think there is anything wrong with either approach
|
On September 12 2020 00:07 99percent wrote:I was curious to see how MMR is distributed on the sc2 ladder. Here is the result (data are from rankedftw.com): I was expected a similar distribution for all races but it is not the case. The interpretation of this result would be adventurous and I personnaly don't want to speculate on it. Method: 1. EU ladder only 2. Only players with at least 10 1v1 games in season 44 are taken into account 3. distribution was first computed using 25 regular bins for each race 4. distribution was then interpolated with a regular interval of 10 MMR
If anyone can get the data as sheets (xls, csv, etc...), I can run some statistical analyses in order to see if there are statistically significant differences of MMR by race.  (I am curious to see the result)
|
On September 12 2020 02:19 Meadowlark wrote:Show nested quote +On September 12 2020 01:26 ThunderJunk wrote:Well done 99percent, not the least bit surprising to me. It's a straight up descriptive stat, and it's beautiful. I love the succinct no-fluff style of the post. Very brave of you to share. On September 12 2020 00:39 Meadowlark wrote:
Interesting stuff! Not surprised to see zerg has the highest on average, because I'm pretty sure it's the least popular race among casual/low level players
edit: ((i say this as a pretty low level player myself)) From a stats perspective, the number of players is completely irrelevant to the MMR spread because this is looking at proportions. If it's true that Zerg is the least popular race, which it may or may not be, the argument would then be, "Players who choose the least popular option are more talented as a population." While this appeals to my hipster sensibilities, it's pretty laughable scientifically. That's why I specified the low-level/casual demographic. If zerg is less popular than the other two races amongst casual players, but, say, equally popular amongst more hardcore players, then it will have a higher average mmr because the zerg playerbase will skew more towards the hardcore than the casual. The point has nothing to do with talent--mmr has much more to do with your approach the the game and how much time you allocate to play. It really wouldn't be surprising if people who allocate more time and consciously focus on improving in the game are a meaningfully different population with different preferences than people who like to mess around and have fun on ladder every so often (10 ladder games per season is a pretty low floor). I can't remember where I read this, and so maybe I'm full of it, but I do remember reading that zerg was overall less popular in the lower leagues. Obviously this is anecdotal, but my not-very-high mmr laddering experience definitely jives with this--I run into far fewer zergs than terran and protoss. I can't really prove this, but my intuition is that the mmr disparity is mostly caused by how different people who are looking for different experiences with the game may choose their races by different criteria. People more interested in game mechanics might find all three races equally interesting on account of their unique mechanics, whereas people less interested in game mechanics might on average prefer space marines and shiny aliens to bugs. That's my thought, at least. Again, definitely could be factually wrong bc I don't remember where I read this, but I really don't think it's logically implausible in the way you made it out to be. edit: and to be clear I really don't think there is anything wrong with either approach
You don't have any evidence at all that Zerg players are more focused on improving or anything of the sort, so at some point you just have to use Occam's razor.
|
|
|
On September 12 2020 01:50 [Phantom] wrote:
So this means that on average most of the population is on plat/diamond?
Before it used to be tat if you were in diamond you were top 10%, master top 2% and GM top 200.
Now I feel bad.
Yes, very disappointing  I am apparently right on the mean. I thought I was in the top quartile or something.
|
On September 12 2020 02:19 Meadowlark wrote:Show nested quote +On September 12 2020 01:26 ThunderJunk wrote:Well done 99percent, not the least bit surprising to me. It's a straight up descriptive stat, and it's beautiful. I love the succinct no-fluff style of the post. Very brave of you to share. On September 12 2020 00:39 Meadowlark wrote:
Interesting stuff! Not surprised to see zerg has the highest on average, because I'm pretty sure it's the least popular race among casual/low level players
edit: ((i say this as a pretty low level player myself)) From a stats perspective, the number of players is completely irrelevant to the MMR spread because this is looking at proportions. If it's true that Zerg is the least popular race, which it may or may not be, the argument would then be, "Players who choose the least popular option are more talented as a population." While this appeals to my hipster sensibilities, it's pretty laughable scientifically. That's why I specified the low-level/casual demographic. If zerg is less popular than the other two races amongst casual players, but, say, equally popular amongst more hardcore players, then it will have a higher average mmr because the zerg playerbase will skew more towards the hardcore than the casual. The point has nothing to do with talent--mmr has much more to do with your approach the the game and how much time you allocate to play. It really wouldn't be surprising if people who allocate more time and consciously focus on improving in the game are a meaningfully different population with different preferences than people who like to mess around and have fun on ladder every so often (10 ladder games per season is a pretty low floor). I can't remember where I read this, and so maybe I'm full of it, but I do remember reading that zerg was overall less popular in the lower leagues. Obviously this is anecdotal, but my not-very-high mmr laddering experience definitely jives with this--I run into far fewer zergs than terran and protoss. I can't really prove this, but my intuition is that the mmr disparity is mostly caused by how different people who are looking for different experiences with the game may choose their races by different criteria. People more interested in game mechanics might find all three races equally interesting on account of their unique mechanics, whereas people less interested in game mechanics might on average prefer space marines and shiny aliens to bugs. That's my thought, at least. Again, definitely could be factually wrong bc I don't remember where I read this, but I really don't think it's logically implausible in the way you made it out to be. edit: and to be clear I really don't think there is anything wrong with either approach
The word Talent isn't the point. You can change the description to "Better", or "Harder Working", or "Morally Superior" lol It doesn't matter : The premise is fundamentally unsound because it leads to a limitless rabbit hole. Let's assume what you say is true: Then anyone could just as easily add another layer and say "Hardcore players choose zerg because they recognize it will be the easiest to get more MMR with."
Pretty funny that the mode (highest number in a single category) for both Protoss and Terran is silver league whereas for Zerg it's diamond league.
|
Whew, just above average for toss. Could be worse.
|
It should be pretty clear that Zerg is the more forgiving race at lower levels. I think the main reason is that you need to make fewer production buildings, and you can build an army in bursts rather than having to check if your production is running continuously. Injects and creep spread probably make more of a difference higher up.
|
Czech Republic12129 Posts
On September 12 2020 05:04 Slydie wrote:
It should be pretty clear that Zerg is the more forgiving race at lower levels. I think the main reason is that you need to make fewer production buildings, and you can build an army in bursts rather than having to check if your production is running continuously. Injects and creep spread probably make more of a difference higher up.
You can work around the injects in the lategame, spawn 4 queens, create 4 hatcheries in main, inject once in a while with them. shit ton of larvae.
|
On September 12 2020 01:50 [Phantom] wrote:
So this means that on average most of the population is on plat/diamond?
Before it used to be tat if you were in diamond you were top 10%, master top 2% and GM top 200.
Now I feel bad.
Correct. They changed the percentages around the beginning of LotV to artificially make players feel better about themselves in an effort to keep people playing ladder.
|
![[image loading]](https://cdn.discordapp.com/attachments/382780801164771331/754174529198227456/unknown.png)
It's very interesting that there's an inflection point in the lower side of the zerg curve that's not present on the terran/toss curves. Is this the point where players learn to inject?
I also think this indicates the "just serral" theory is likely untrue. It definitely appears that zerg is the most successful race by a significant margin. We need more info to draw conclusions - would love to see sample size!
Also, is there no random statistics? As a random player, I want to know how my race is doing! 
|
On September 12 2020 12:06 Monochromatic wrote:It's very interesting that there's an inflection point in the lower side of the zerg curve that's not present on the terran/toss curves. Is this the point where players learn to inject? I also think this indicates the "just serral" theory is likely untrue. It definitely appears that zerg is the most successful race by a significant margin. We need more info to draw conclusions - would love to see sample size! Also, is there no random statistics? As a random player, I want to know how my race is doing!
Of course its untrue, Serral can't carry zerg all by himself. As icing on the cake when he runs into fellow good zerg Reynor he loses a lot of series against him, but not versus other good terran/protoss EU players
|
On September 12 2020 12:26 NinjaNight wrote:+ Show Spoiler +On September 12 2020 12:06 Monochromatic wrote:It's very interesting that there's an inflection point in the lower side of the zerg curve that's not present on the terran/toss curves. Is this the point where players learn to inject? I also think this indicates the "just serral" theory is likely untrue. It definitely appears that zerg is the most successful race by a significant margin. We need more info to draw conclusions - would love to see sample size! Also, is there no random statistics? As a random player, I want to know how my race is doing! Of course its untrue, Serral can't carry zerg all by himself. As icing on the cake when he runs into fellow good zerg Reynor he loses a lot of series against him, but not versus other good terran/protoss EU players
He has also lost to Elazer and soO a few times. I think it only shows that zvz is a volatile matchup.
|
I dont know how much work it is but can you provide the same for korea. Zerg was always pretty popular in EU and I want to see if KR looks similar or if better players tend to chose Terran more over there.
|
On September 12 2020 12:06 Monochromatic wrote:I also think this indicates the "just serral" theory is likely untrue. It definitely appears that zerg is the most successful race by a significant margin. We need more info to draw conclusions - would love to see sample size!
What do you mean, sample size ? This is the entire EU population so it's no "sample".
There is no drawing conclusions from this however. First, there can be many reasons why zerg is underrepresented in lower leagues. But more importantly, looking at the bell curve only highlights the middle of the curve where most people are, and I hope we can all agree the 2.5k - 3k mmr range has nothing to do with competitive play and should be disregarded when thinking about balance. This has 0 bearing on the "just serral" debate. What you don't see with this range is that zerg % declines heavily and across all servers from diamond to GM.
https://www.rankedftw.com/stats/races/1v1/#v=2&r=2&l=6
OP, if you're truly unbiased like you claim to be, please post the same curve but with a filter on only diamond+ players. Spoiler alert : it will look like the exact opposite.
|
I think one explanation could be, that casual Players, that Don t have any SC Background, just Don t like playing Zerg.
Their mechanics are quite different from what you know from your usual rts experience.
For example when I introduced my older Brother to sc2, Who had some Basic wc3 and other rts experiences, he said, je doesn t care if he plays T or P (for 2v2 with me) but he doesn t want to Play Z, as he doesn t understand them.
If there is a signifikant amount of people, that shy away from that less typical mechanics without different Produktion buildings, it would explain, why less casual Players choose Zerg
|
Like dbRic1203 says, Zerg plays a bit different than the more traditional rts styles, like T and P. The inflection point might be the point where zerg players "understand" how to play zerg and suddenly get a higher win ratio. Later on the curve it is more and more just good players, so it evens out more, no matter the race.
When I was a T player on gold league many years ago, I struggled a lot versus Zerg players that understand how to take many bases early on versus Zerg players that played on equal bases as T, because the former would just flood me with units and prevent me to take any more bases.
|
I also have to say that it is easy to forget what kind of games are played at 2,6 and 2,9k MMR.
At 2,6, players tend to have their very own and generally very ineffective style, based on playing the campaign or fooling around with units. I doubt many of them have ever seen a pro-game or looked up a build to see what 1v1s are supposed to look like.
Even at 3k+, terrans make planetaries in their mains and go for proxy ghost rushes, while zergs are still one-trick-ponies relying on a single attack to win with no clue about transitioning.
It takes one to know one. I just legitimately brought my offrace Zerg to Plat 3, which would make me contribute to the peak of the curve. Even with just very basic understanding about expanding and making drones, upgrades and roaches, I feel I had a very easy time in most games, even ZvZs.
|
dbRic1203 has mentioned a pretty interesting point there. This reminds me of when new players bought WoL and I'm pretty sure lots of them kind of chose terran by default (as it was the campaign race for WoL). The dominant race in competitive play is probably a big factor as well, what with GomTVT and literal god Mvp dominating the entire scene for the first couple of years. Race selection is probably a combination of these factors and some others. All things considered, the numbers speak for themselves and the ratio of players is approximately:
36.5% Terran
29.5% Protoss
25.5% Zerg
8.5% Random
(stats taken from https://www.rankedftw.com/stats/races/1v1/#v=2&r=2&l=-2)
Had a similar experience at the time myself. T and P was always "easier" to get used to, and it took almost a year before I even thought about playing zerg at all. The mechanics of injecting, spreading creep and making things out of hatcheries is just so different from RTS principles that are present in wc3, CnC and wh40k - Terran and Protoss follow basic principles like "make units one at a time" and "building A spits out unit A, B and C". There's probably a large proportion of new players who default to T and P, and I assume there is some correlation between being new and having low mmr as well. Just food for thought, of course.
|
you guys are missing the important part of the graph. the right side of the graph.
|
On September 12 2020 21:00 Alejandrisha wrote:
you guys are missing the important part of the graph. the right side of the graph.
what's interesting there? Serral + Reynor taking spots 1+2? With Protoss players following them (Showtime, Neeb, etc) and then some Terrans (HM, Clem, Soul)
All in all, this graph basically just shows that until the 3k MMR range, there are a bit more Zergs than Protoss/Terran.
If we take a look at how the amount of Zerg players grows from about 2k MMR, and considering, as someone mentioned, "what games are played" around this region (cheese) it's a pretty good indicator that from 2k MMR going towards 3k MMR, Zergs learn how to allin, while Protoss/Terran players struggle to defend that (e.g. building proper walls).
Going closer to and onward from ~4k MMR, it shows that the races converge and have equal MMR very quickly, because cheese stops working, while people learn to defend it and actually start understanding how to play SC2 better (macro, build-orders, strategy, etc)
|
On September 12 2020 21:38 suriel- wrote:Show nested quote +On September 12 2020 21:00 Alejandrisha wrote:
you guys are missing the important part of the graph. the right side of the graph. what's interesting there? Serral + Reynor taking spots 1+2? With Protoss players following them (Showtime, Neeb, etc) and then some Terrans (HM, Clem, Soul) All in all, this graph basically just shows that until the 3k MMR range, there are a bit more Zergs than Protoss/Terran. If we take a look at how the amount of Zerg players grows from about 2k MMR, and considering, as someone mentioned, "what games are played" around this region (cheese) it's a pretty good indicator that from 2k MMR going towards 3k MMR, Zergs learn how to allin, while Protoss/Terran players struggle to defend that (e.g. building proper walls). Going closer to and onward from ~4k MMR, it shows that the races converge and have equal MMR very quickly, because cheese stops working, while people learn to defend it and actually start understanding how to play SC2 better (macro, build-orders, strategy, etc)
yes! a thousand times yes !
|
I am suspicious of a graph that has negative percentage on its y-axis instead of starting at 0%.
Anyways quick takeaways:
Difference in peak between Terran (the lowest) and Zerg (the highest) is 370. So the difference is Gold 2 and Platinum 2 as the mode of those races. What does balance or skill particularily matter about those ranks?
Master MMR is about 4400 mmr to get into masters, where all the difference seemingly disappears. Possibly because of scaling. From the graph, if you can't get into Masters, this "statistic" isn't the reason why.
Zerg is the least popular race. So presumably coronavirus being a boost to gaming numbers means less proportionately of a large influx of new players are playing Zerg. This large influx of gaming is seen to be affecting the ranking systems of many games.
A graph taken in isolation without thinking behind how this is proper representation is usually biased. This graph is not a statistical representation of "effort" or "skill" needed to attain MMR.
|
On September 12 2020 21:53 Dangermousecatdog wrote:
I am suspicious of a graph that has negative percentage as one of its x-axis instead of starting at 0%.
Anyways quick takeaways:
Difference in peak between Terran (the lowest) and Zerg (the highest) is 370. So the difference is Gold 2 and Platinum 2 as the mode of those races. What does balance or skill particularily matter about those ranks?
Master MMR is about 4400 mmr to get into masters, where all the difference seemingly disappears. Possibly because of scaling. From the graph, if you can't get into Masters, this "statistic" isn't the reason why.
Zerg is the least popular race. So presumably coronavirus being a boost to gaming numbers means less proportionately of a large influx of new players are playing Zerg. This large influx of gaming is seen to be affecting the ranking systems of many games.
A graph taken in isolation without thinking behind how this is proper representation is usually biased. This graph is not a statistical representation of "effort" or "skill" needed to attain MMR.
so, given that a race has a 'skill' factor to get it closer to the mean, we get
zerg = x(jasbean) + 500
protoss = x(jasbeantwins) + 700
terran = x(janbeandoooods) + 800
we can get a graph that matches a player's effort, x, and just add points as we go along.
|
so this graph essentially shows that there are more zerg platinum players right?
|
On September 12 2020 21:59 sparklyresidue wrote:
so this graph essentially shows that there are more zerg platinum players right?
i think that is the conclusion, yes.
|
Very few people begins with Z because of the specific mechanics of the race.
They generally switch to Z after having played a bit with the 2 other races before going for Z which means they are more experienced and stronger.
|
On September 12 2020 21:57 Alejandrisha wrote:Show nested quote +On September 12 2020 21:53 Dangermousecatdog wrote:
I am suspicious of a graph that has negative percentage as one of its x-axis instead of starting at 0%.
Anyways quick takeaways:
Difference in peak between Terran (the lowest) and Zerg (the highest) is 370. So the difference is Gold 2 and Platinum 2 as the mode of those races. What does balance or skill particularily matter about those ranks?
Master MMR is about 4400 mmr to get into masters, where all the difference seemingly disappears. Possibly because of scaling. From the graph, if you can't get into Masters, this "statistic" isn't the reason why.
Zerg is the least popular race. So presumably coronavirus being a boost to gaming numbers means less proportionately of a large influx of new players are playing Zerg. This large influx of gaming is seen to be affecting the ranking systems of many games.
A graph taken in isolation without thinking behind how this is proper representation is usually biased. This graph is not a statistical representation of "effort" or "skill" needed to attain MMR. so, given that a race has a 'skill' factor to get it closer to the mean, we get zerg = x(jasbean) + 500 protoss = x(jasbeantwins) + 700 terran = x(janbeandoooods) + 800 we can get a graph that matches a player's effort, x, and just add points as we go along.
Well, no. That's an assumption that every race uses the same skill to attain their own mode (not mean, mean isn't in the graph, good luck calculating that from lack of data, though I suppose you can just eye it inaccurately, but it seems more liekly from your later reply that you just don't know the difference between mean and mode), or that the mode is particularily meaningful. Afterall, can you tell the difference in skill between gold 2 and plat 2? I can't, they both just look randomly bad. Players of that skill level are inconsistent and prone to wide swings in skill. Which is only to be expected when in those ranks you should be learning and accumulating experience rapidly.
Nevermind that I cannot see the datapoints. Where is the table that shows the data? For all I know there could be only 7 datapoints, 1 for each league, for each race which makes the people talking about inflection points seem particularily naive as it is just an artifact of a smoothing algorithm from Microsoft Excel or whatever.
In anycase, it seems odd to choose EU in particular. I assume NA and Korea would look different.
Edit: Judging from your later reply, you don't have the education or experience in interpreting graphs, so I think you shouldn't try to make any judgements.
|
On September 12 2020 22:00 Alejandrisha wrote:Show nested quote +On September 12 2020 21:59 sparklyresidue wrote:
so this graph essentially shows that there are more zerg platinum players right? i think that is the conclusion, yes.
Wrong. It doesn't show what the ranges of MMR is. You'll have to take the area under the curve for the MMR range to interpret it as a population group. For instance, if you interpret the graph to show that most protoss players are Gold, you would be wrong, as the most protoss are in Diamond according to https://www.rankedftw.com/stats/races/1v1/#v=2&r=0&l=4. You cannot interpret the mode as the most of a league.
See? That's what I am saying about interpretating graphs.
The more I look at the graph the more nonsensical it become in terms of presenting data. Such a graph, judging from the axis should be presented as a column graph.
|
On September 12 2020 22:27 Dangermousecatdog wrote:Show nested quote +On September 12 2020 22:00 Alejandrisha wrote:On September 12 2020 21:59 sparklyresidue wrote:
so this graph essentially shows that there are more zerg platinum players right? i think that is the conclusion, yes. Wrong. It doesn't show what the ranges of MMR is. You'll have to take the area under the curve for the MMR range to interpret it as a population group. For instance, if you interpret the graph to show that most protoss players are Gold, you would be wrong, as the most protoss are in Diamond according to https://www.rankedftw.com/stats/races/1v1/#v=2&r=0&l=4. You cannot interpret the mode as the most of a league. See? That's what I am saying about interpretating graphs. The more I look at the graph the more nonsensical it become in terms of presenting data. Such a graph, judging from the axis should be presented as a column graph.
true
|
On September 12 2020 22:11 Dangermousecatdog wrote:Show nested quote +On September 12 2020 21:57 Alejandrisha wrote:On September 12 2020 21:53 Dangermousecatdog wrote:
I am suspicious of a graph that has negative percentage as one of its x-axis instead of starting at 0%.
Anyways quick takeaways:
Difference in peak between Terran (the lowest) and Zerg (the highest) is 370. So the difference is Gold 2 and Platinum 2 as the mode of those races. What does balance or skill particularily matter about those ranks?
Master MMR is about 4400 mmr to get into masters, where all the difference seemingly disappears. Possibly because of scaling. From the graph, if you can't get into Masters, this "statistic" isn't the reason why.
Zerg is the least popular race. So presumably coronavirus being a boost to gaming numbers means less proportionately of a large influx of new players are playing Zerg. This large influx of gaming is seen to be affecting the ranking systems of many games.
A graph taken in isolation without thinking behind how this is proper representation is usually biased. This graph is not a statistical representation of "effort" or "skill" needed to attain MMR. so, given that a race has a 'skill' factor to get it closer to the mean, we get zerg = x(jasbean) + 500 protoss = x(jasbeantwins) + 700 terran = x(janbeandoooods) + 800 we can get a graph that matches a player's effort, x, and just add points as we go along. Well, no. That's an assumption that every race uses the same skill to attain their own mode (not mean, mean isn't in the graph, good luck calculating that from lack of data, though I suppose you can just eye it inaccurately, but it seems more liekly from your later reply that you just don't know the difference between mean and mode), or that the mode is particularily meaningful. Afterall, can you tell the difference in skill between gold 2 and plat 2? I can't, they both just look randomly bad. Players of that skill level are inconsistent and prone to wide swings in skill. Which is only to be expected when in those ranks you should be learning and accumulating experience rapidly. Nevermind that I cannot see the datapoints. Where is the table that shows the data? For all I know there could be only 7 datapoints, 1 for each league, for each race which makes the people talking about inflection points seem particularily naive as it is just an artifact of a smoothing algorithm from Microsoft Excel or whatever. In anycase, it seems odd to choose EU in particular. I assume NA and Korea would look different. Edit: Judging from your later reply, you don't have the education or experience in interpreting graphs, so I think you shouldn't try to make any judgements.
mean is the mathematical average. mode is the most prevalent data point. what does this have to do with this graph
i don't think figuring out the difference between plat and gold 2 is going to bring you any epiphany
|
Ok, you tell me where the mean of anything is from that graph. Go on.
Hint: it's not at 2610, 2680, 2980.
|
On September 12 2020 22:41 Dangermousecatdog wrote:
Ok, you tell me where the mean of anything is from that graph. Go on.
Hint: it's not at 2610, 2680, 2980.
what if it were? would that bring you to a stunning realization that would enable you to solve sc2? what does this graph have anything to do with the game? anything drawn from the graph is whatever you want it to be. why is this a discussion
|
Hey you asked right? Don't act offended. Ok, so you don't understand the difference between a mode and mean, but you like to talk about mean. It's ok. It just mean that you shouldn't interpret the graph.
By the way those numbers aren't the node either. That can only be shown from the actual data table. Like I said, the more I look at the graph, the more nonsensical it appears. I'll have to see the data myself, but if it's from that same link then what we have is just a bullshit graph. There is only 7 datapoints per race so how could he make such a graph to present the data?
|
On September 12 2020 22:47 Dangermousecatdog wrote:
Hey you asked right? Don't act offended. Ok, so you don't understand the difference between a mode and mean. It's ok. It just mean that you shouldn't interpret the graph.
By the way those numbers aren't the node either. That can only be shown from the actual data table. Like I said, the more I look at the graph, the more nonsensical it appears. I'll have to see the data myself, but if it's from that same link then what we have is just a bullshit graph. There is only 7 datapoints per race so how could he make such a graph to present the data?
did i say they were? why are you agreeing with me aggressively?
|
On September 12 2020 22:47 Dangermousecatdog wrote:
By the way those numbers aren't the node either. That can only be shown from the actual data table. Like I said, the more I look at the graph, the more nonsensical it appears. I'll have to see the data myself, but if it's from that same link then what we have is just a bullshit graph. There is only 7 datapoints per race so how could he make such a graph to present the data?
Wait, why does rankedftw only give 7 datapoints per race? https://www.rankedftw.com/ladder/lotv/1v1/mmr/?f=eu If you go here, there are 160239 datapoints for EU alone, it seems plausible that that OP generated the plot from rankedftw.com
Edit: your argument seems to be that the graph doesn't represent the 7 leagues correctly. Yes, indeed, if you want the % of gold players, you should integrate the curve between min(gold MMR) to max(gold MMR), and you'd get the correct percentage. But that's just b/c the transformation between leagues and MMR isn't linear, which makes sense since they're based on percentages and you have long tails. In other words, the statement "most protoss players are gold" is wrong, but "most protoss players are in the range 2300-3100 MMR" is perfectly correct, one is based on a league PDF and the other is based on an MMR PDF.
|
heres some raw data and a quick graph for each region from the battle.net api. only counting players with at least 10 games.
docs.google.com
|
On September 12 2020 21:00 Alejandrisha wrote:
you guys are missing the important part of the graph. the right side of the graph.
The extreme right side of the graph shows Zerg at the top.
(They also win most major tournaments.)
|
On September 12 2020 20:19 Slydie wrote:
I also have to say that it is easy to forget what kind of games are played at 2,6 and 2,9k MMR.
At 2,6, players tend to have their very own and generally very ineffective style, based on playing the campaign or fooling around with units. I doubt many of them have ever seen a pro-game or looked up a build to see what 1v1s are supposed to look like.
Even at 3k+, terrans make planetaries in their mains and go for proxy ghost rushes, while zergs are still one-trick-ponies relying on a single attack to win with no clue about transitioning.
It takes one to know one. I just legitimately brought my offrace Zerg to Plat 3, which would make me contribute to the peak of the curve. Even with just very basic understanding about expanding and making drones, upgrades and roaches, I feel I had a very easy time in most games, even ZvZs.
As someone who peaks at 4k and drops to 3.5k after not playing for a while, the planetary in the main thing sounds shocking to me. I have seen a planetary in the natural like 1 in 50 games, never in the main.
|
Not surprised to see Terran having the lowest average mmr as Terran seems to be the go to race to play for new players entering ladder.
|
On September 13 2020 00:59 greenturtle23 wrote:Show nested quote +On September 12 2020 20:19 Slydie wrote:
I also have to say that it is easy to forget what kind of games are played at 2,6 and 2,9k MMR.
At 2,6, players tend to have their very own and generally very ineffective style, based on playing the campaign or fooling around with units. I doubt many of them have ever seen a pro-game or looked up a build to see what 1v1s are supposed to look like.
Even at 3k+, terrans make planetaries in their mains and go for proxy ghost rushes, while zergs are still one-trick-ponies relying on a single attack to win with no clue about transitioning.
It takes one to know one. I just legitimately brought my offrace Zerg to Plat 3, which would make me contribute to the peak of the curve. Even with just very basic understanding about expanding and making drones, upgrades and roaches, I feel I had a very easy time in most games, even ZvZs.
As someone who peaks at 4k and drops to 3.5k after not playing for a while, the planetary in the main thing sounds shocking to me. I have seen a planetary in the natural like 1 in 50 games, never in the main.
It stops way before 3,5k, and it is rare even at 3k. Every 100 MMR points in SC2 mark a significant skill difference. 2800-3500 are very highly populated ranks where the leagues are close.
The main point of the post was the dub 3k games, though.
|
I was bored so I built a script to scrape all the player data from the site in the OP and then did a bit of basic pandasing and seaborning so I could see what NA looks like.
Here's what the NA graph looks like:
![[image loading]](https://i.imgur.com/j7RyULI.png)
Here's a few basic tidbits from the data since people were wondering about means and stuff:
Total active accounts as of when I scraped the data:
protoss: 51947
terran: 56176
zerg: 46133
random: 15610
mean MMR:
protoss: 2833.9011492482723
terran: 2768.2159819139847
zerg: 2966.6612836797954
random: 2898.173478539398
median MMR:
protoss: 2740.0
terran: 2659.0
zerg: 2901.0
random: 2845.0
overall mean MMR: 2854.140540190503
median MMR across population: 2779.0
|
+ Show Spoiler +On September 13 2020 08:51 Ben... wrote:I was bored so I built a script to scrape all the player data from the site in the OP and then did a bit of basic pandasing and seaborning so I could see what NA looks like. Here's what the NA graph looks like: Here's a few basic tidbits from the data since people were wondering about means and stuff: Total active accounts as of when I scraped the data:
protoss: 51947
terran: 56176
zerg: 46133
random: 15610
mean MMR:
protoss: 2833.9011492482723
terran: 2768.2159819139847
zerg: 2966.6612836797954
random: 2898.173478539398
median MMR:
protoss: 2740.0
terran: 2659.0
zerg: 2901.0
random: 2845.0
overall mean MMR: 2854.140540190503
median MMR across population: 2779.0
thansk for posting! i think medians are very important in a lot of cases. modes can be skewed by outliers. however, the fact that anyone can sign up regardless of their skill in other races and be data points in this still makes me question whether or not this graph really tells any story at all. still fun to look at stats tho
median MMR:
protoss: 2740.0
terran: 2659.0
zerg: 2901.0
random: 2845.0
seems like my satirical handicap model
zerg = x(jasbean) + 500
protoss = x(jasbeantwins) + 700
terran = x(janbeandoooods) + 800
holds up though
2900 + 500 = 3400
2740 + 700 = 3440
2660 + 800 = 3460
almost haunting
|
I saw OP filtered out people who played fewer than 10 games (I didn't in my previous post) so I did also (it's trivial since the website provides number of games played as a column) just to see the difference. It's not really all that different but for consistency's sake here's the data:
The graph looks basically identical:
![[image loading]](https://i.imgur.com/oeRCStA.png)
data is quite close too:
Number of accounts with 10+ games by race:
protoss:31247
terran: 33855
zerg: 27118
random:8203
mean:
protoss:2848.8236310685825
terran: 2778.8631516762666
zerg: 2985.8191975809427
random:2872.1861514080215
median:
protoss:2759.0
terran: 2674.0
zerg: 2920.0
random:2818.0
overall mean MMR: 2864.1406052398356
population median MMR: 2793.0
Mean number of games played so far this season:
factoring in accounts with fewer than 10 games played: 55.53022382348439
ignoring accounts with fewer than 10 games played: 90.96791571651913
|
well, simply the fact that random has a higher median than 2 races tells you the study is flawed. these random players have played all the races and are just doing a victory lap. doesn't give me much faith in the rest of the data
|
On September 13 2020 09:18 Ben... wrote:I saw OP filtered out people who played fewer than 10 games (I didn't in my previous post) so I did also (it's trivial since the website provides number of games played as a column) just to see the difference. It's not really all that different but for consistency's sake here's the data: The graph looks basically identical: data is quite close too: Number of accounts with 10+ games by race:
protoss:31247
terran: 33855
zerg: 27118
random:8203
mean:
protoss:2848.8236310685825
terran: 2778.8631516762666
zerg: 2985.8191975809427
random:2872.1861514080215
median:
protoss:2759.0
terran: 2674.0
zerg: 2920.0
random:2818.0
overall mean MMR: 2864.1406052398356
population median MMR: 2793.0
Mean number of games played so far this season:
factoring in accounts with fewer than 10 games played: 55.53022382348439
ignoring accounts with fewer than 10 games played: 90.96791571651913
Hi, I asked for a log y axis plot but OP doesn't seem to have responded; since you seem to be checking this topic, would you be able to post one with a log y axis as well? Would be able to show the long tail at higher/lower MMR much better
I definitely agree that using mean/median/mode to draw balance conclusions is extremely dubious, but it should be fun to theorize nonetheless
|
On September 13 2020 09:31 yubo56 wrote:Show nested quote +On September 13 2020 09:18 Ben... wrote:I saw OP filtered out people who played fewer than 10 games (I didn't in my previous post) so I did also (it's trivial since the website provides number of games played as a column) just to see the difference. It's not really all that different but for consistency's sake here's the data: The graph looks basically identical: data is quite close too: Number of accounts with 10+ games by race:
protoss:31247
terran: 33855
zerg: 27118
random:8203
mean:
protoss:2848.8236310685825
terran: 2778.8631516762666
zerg: 2985.8191975809427
random:2872.1861514080215
median:
protoss:2759.0
terran: 2674.0
zerg: 2920.0
random:2818.0
overall mean MMR: 2864.1406052398356
population median MMR: 2793.0
Mean number of games played so far this season:
factoring in accounts with fewer than 10 games played: 55.53022382348439
ignoring accounts with fewer than 10 games played: 90.96791571651913
Hi, I asked for a log y axis plot but OP doesn't seem to have responded; since you seem to be checking this topic, would you be able to post one with a log y axis as well? Would be able to show the long tail at higher/lower MMR much better I definitely agree that using mean/median/mode to draw balance conclusions is extremely dubious, but it should be fun to theorize nonetheless
yes i think that any conclusion drawn are bunk but i do like lookin at numbers xD
|
On September 13 2020 09:34 Alejandrisha wrote:Show nested quote +On September 13 2020 09:31 yubo56 wrote:On September 13 2020 09:18 Ben... wrote:I saw OP filtered out people who played fewer than 10 games (I didn't in my previous post) so I did also (it's trivial since the website provides number of games played as a column) just to see the difference. It's not really all that different but for consistency's sake here's the data: The graph looks basically identical: data is quite close too: Number of accounts with 10+ games by race:
protoss:31247
terran: 33855
zerg: 27118
random:8203
mean:
protoss:2848.8236310685825
terran: 2778.8631516762666
zerg: 2985.8191975809427
random:2872.1861514080215
median:
protoss:2759.0
terran: 2674.0
zerg: 2920.0
random:2818.0
overall mean MMR: 2864.1406052398356
population median MMR: 2793.0
Mean number of games played so far this season:
factoring in accounts with fewer than 10 games played: 55.53022382348439
ignoring accounts with fewer than 10 games played: 90.96791571651913
Hi, I asked for a log y axis plot but OP doesn't seem to have responded; since you seem to be checking this topic, would you be able to post one with a log y axis as well? Would be able to show the long tail at higher/lower MMR much better I definitely agree that using mean/median/mode to draw balance conclusions is extremely dubious, but it should be fun to theorize nonetheless yes i think that any conclusion drawn are bunk but i do like lookin at numbers xD
Ohhh, one fun plot to make would be a scatter plot of (# games played, MMR) with the three races color coded. It might give a slightly more faithful signal.
How are y'all scraping the data, did you just download all of the pages and parse the HTML?
|
On September 13 2020 09:31 yubo56 wrote:Show nested quote +On September 13 2020 09:18 Ben... wrote:I saw OP filtered out people who played fewer than 10 games (I didn't in my previous post) so I did also (it's trivial since the website provides number of games played as a column) just to see the difference. It's not really all that different but for consistency's sake here's the data: The graph looks basically identical: data is quite close too: Number of accounts with 10+ games by race:
protoss:31247
terran: 33855
zerg: 27118
random:8203
mean:
protoss:2848.8236310685825
terran: 2778.8631516762666
zerg: 2985.8191975809427
random:2872.1861514080215
median:
protoss:2759.0
terran: 2674.0
zerg: 2920.0
random:2818.0
overall mean MMR: 2864.1406052398356
population median MMR: 2793.0
Mean number of games played so far this season:
factoring in accounts with fewer than 10 games played: 55.53022382348439
ignoring accounts with fewer than 10 games played: 90.96791571651913
Hi, I asked for a log y axis plot but OP doesn't seem to have responded; since you seem to be checking this topic, would you be able to post one with a log y axis as well? Would be able to show the long tail at higher/lower MMR much better I definitely agree that using mean/median/mode to draw balance conclusions is extremely dubious, but it should be fun to theorize nonetheless
Does this work? It's for accounts with more than 10 games.
![[image loading]](https://i.imgur.com/wpVd6Rm.png)
All I did was add ".set(yscale='log')". Nothing fancy.
Yeah I don't think anything meaningful can actually be drawn from this data. I just like tinkering.
edit: On September 13 2020 09:31 yubo56 wrote:
How are y'all scraping the data, did you just download all of the pages and parse the HTML?
Pretty much. I just used the requests library to download each page (you can use the offset GET parameter they use for pagination to hop between pages), then used BeautifulSoup to parse the HTML and then cleaned stuff up a bit before chucking it in CSVs. Nothing too advanced.
|
On September 13 2020 09:42 Ben... wrote:Show nested quote +On September 13 2020 09:31 yubo56 wrote:On September 13 2020 09:18 Ben... wrote:I saw OP filtered out people who played fewer than 10 games (I didn't in my previous post) so I did also (it's trivial since the website provides number of games played as a column) just to see the difference. It's not really all that different but for consistency's sake here's the data: The graph looks basically identical: data is quite close too: Number of accounts with 10+ games by race:
protoss:31247
terran: 33855
zerg: 27118
random:8203
mean:
protoss:2848.8236310685825
terran: 2778.8631516762666
zerg: 2985.8191975809427
random:2872.1861514080215
median:
protoss:2759.0
terran: 2674.0
zerg: 2920.0
random:2818.0
overall mean MMR: 2864.1406052398356
population median MMR: 2793.0
Mean number of games played so far this season:
factoring in accounts with fewer than 10 games played: 55.53022382348439
ignoring accounts with fewer than 10 games played: 90.96791571651913
Hi, I asked for a log y axis plot but OP doesn't seem to have responded; since you seem to be checking this topic, would you be able to post one with a log y axis as well? Would be able to show the long tail at higher/lower MMR much better I definitely agree that using mean/median/mode to draw balance conclusions is extremely dubious, but it should be fun to theorize nonetheless Does this work? It's for accounts with more than 10 games. All I did was add ".set(yscale='log')". Nothing fancy. Yeah I don't think anything meaningful can actually be drawn from this data. I just like tinkering. edit: Show nested quote +On September 13 2020 09:31 yubo56 wrote:
How are y'all scraping the data, did you just download all of the pages and parse the HTML? Pretty much. I just used the requests library to download each page (you can use the offset GET parameter they use for pagination to hop between pages), then used BeautifulSoup to parse the HTML and then cleaned stuff up a bit before chucking it in CSVs. Nothing too advanced.
Both of your answers are exactly what I was asking about, thanks so much!
Interesting tail on the log scale, I think the extreme high end is basically just Parting, Neeb, and Scarlett holding the distribution up (you said this was NA right). There's a bit of small deviation around 6k, which is kinda interesting, but you're probably already up to like top 20 at that point. That the curves actually track each other so well from 3k-5.5k is really cool; this is the group of people that know how the game works but don't play anywhere near perfectly, and all their balance complaints notwithstanding, balance is incredibly good at their level!
Also, that the tail is linear in semi-log space means the distribution ~ exp(-MMR) right, so not Gaussian? That's also kinda cool to see
Really appreciate your responses and plots!
|
On September 13 2020 09:42 Ben... wrote:Show nested quote +On September 13 2020 09:31 yubo56 wrote:On September 13 2020 09:18 Ben... wrote:I saw OP filtered out people who played fewer than 10 games (I didn't in my previous post) so I did also (it's trivial since the website provides number of games played as a column) just to see the difference. It's not really all that different but for consistency's sake here's the data: The graph looks basically identical: data is quite close too: Number of accounts with 10+ games by race:
protoss:31247
terran: 33855
zerg: 27118
random:8203
mean:
protoss:2848.8236310685825
terran: 2778.8631516762666
zerg: 2985.8191975809427
random:2872.1861514080215
median:
protoss:2759.0
terran: 2674.0
zerg: 2920.0
random:2818.0
overall mean MMR: 2864.1406052398356
population median MMR: 2793.0
Mean number of games played so far this season:
factoring in accounts with fewer than 10 games played: 55.53022382348439
ignoring accounts with fewer than 10 games played: 90.96791571651913
Hi, I asked for a log y axis plot but OP doesn't seem to have responded; since you seem to be checking this topic, would you be able to post one with a log y axis as well? Would be able to show the long tail at higher/lower MMR much better I definitely agree that using mean/median/mode to draw balance conclusions is extremely dubious, but it should be fun to theorize nonetheless Does this work? It's for accounts with more than 10 games. All I did was add ".set(yscale='log')". Nothing fancy. Yeah I don't think anything meaningful can actually be drawn from this data. I just like tinkering. edit: Show nested quote +On September 13 2020 09:31 yubo56 wrote:
How are y'all scraping the data, did you just download all of the pages and parse the HTML? Pretty much. I just used the requests library to download each page (you can use the offset GET parameter they use for pagination to hop between pages), then used BeautifulSoup to parse the HTML and then cleaned stuff up a bit before chucking it in CSVs. Nothing too advanced.
thank you, this is a much better graph. still don't think we can draw real conclusion from it. but this is better in that we can actually see the differences
|
I did the scatterplots for fun also (y being games played, x is MMR). They're kinda neat.
For this first one, I did log scale for y axis since otherwise it's incredibly cramped. I set it ridiculously wide for this because I noticed something (spoilered because this one's kinda biggish download-wise):
+ Show Spoiler +
Those vertical lines are interesting.
A smaller scatterplot reveals the three most prominent lines on the right are around 4000, 4400, and approximately 4700-4800 MMR, which I'm guessing are cutoffs for various tiers of leagues, probably Master or Diamond 1. A bunch of accounts seem to cluster around the MMR cutoff for league tiers:
+ Show Spoiler +
Cool stuff.
edit: Yes, that giant space on the left is populated. There's someone with ~200 MMR.
double edit: Cleaned up the plots a smidge and made the y axis more reasonable. I had to do it manually since 10e2 was too small and 10e3 left a bunch of empty space so I went with a range of 10 to 3500 games. When I went from 3000 to 3500 I saw one data point added so somebody has to have played a lot of SC2 this season.
|
On September 13 2020 10:24 Ben... wrote:I did the scatterplots for fun also (y being games played, x is MMR). They're kinda neat. For this first one, I did log scale for y axis since otherwise it's incredibly cramped. I set it ridiculously wide for this because I noticed something (spoilered because this one's kinda biggish download-wise): + Show Spoiler +Those vertical lines are interesting. A smaller scatterplot reveals the three most prominent lines on the right are around 4000, 4400, and approximately 4700-4800 MMR, which I'm guessing are cutoffs for various tiers of leagues, probably Master or Diamond 1. A bunch of accounts seem to cluster around the MMR cutoff for league tiers: + Show Spoiler +Cool stuff. edit: Yes, that giant space on the left is populated. There's someone with ~200 MMR. double edit: Cleaned up the plots a smidge and made the y axis more reasonable. I had to do it manually since 10e2 was too small and 10e3 left a bunch of empty space so I went with a range of 10 to 3500 games. When I went from 3000 to 3500 I saw one data point added so somebody has to have played a lot of SC2 this season.
That vertical line is super cool. Think those cutoffs are for d2-d1, d1-m3, and m3-m2 respectively, that's a cool feature to see.
I guess it's too dense to see the differences between the races, it would be super cool to see something like "terrans gain MMR fastest, relative to # games played, since their macro is most intuitive/similar to other games" but that's probably too hard to quantify.
Thanks again for keeping us posted dude ur awesome <3
|
On September 13 2020 10:24 Ben... wrote:I did the scatterplots for fun also (y being games played, x is MMR). They're kinda neat. For this first one, I did log scale for y axis since otherwise it's incredibly cramped. I set it ridiculously wide for this because I noticed something (spoilered because this one's kinda biggish download-wise): + Show Spoiler +Those vertical lines are interesting. A smaller scatterplot reveals the three most prominent lines on the right are around 4000, 4400, and approximately 4700-4800 MMR, which I'm guessing are cutoffs for various tiers of leagues, probably Master or Diamond 1. A bunch of accounts seem to cluster around the MMR cutoff for league tiers: + Show Spoiler +Cool stuff. edit: Yes, that giant space on the left is populated. There's someone with ~200 MMR. double edit: Cleaned up the plots a smidge and made the y axis more reasonable. I had to do it manually since 10e2 was too small and 10e3 left a bunch of empty space so I went with a range of 10 to 3500 games. When I went from 3000 to 3500 I saw one data point added so somebody has to have played a lot of SC2 this season.
This is far less meaningfull than I expected
Is it possible to plötzlich lifetime games against mmr as well?
Also thx for putting together the actually meaningfull Graph with the Log scale. That one actually Shows, how good the Balance is in Leagues with the most Balance whiners.
I m allways pleased to see them proved wrong, as for 99.99% of your average SC2 game your own mistakes are the only thing to blaue for your loss.
|
On September 13 2020 14:09 dbRic1203 wrote:
Also thx for putting together the actually meaningfull Graph with the Log scale. That one actually Shows, how good the Balance is in Leagues with the most Balance whiners.
I m allways pleased to see them proved wrong, as for 99.99% of your average SC2 game your own mistakes are the only thing to blaue for your loss.
Well, one thing you can be sure with is that the whiners will never accept actual facts like the races being very well balanced and that there's no big swings at all. They also won't ever accept the logically simplest explanation that Zerg just plays differently than other races, so they are less represented in ranks below 3k MM, which was also supported by the first graph, showing the curve of Zergs is rising much slower from 2k -> 3k than that of Protoss/Terran.
Also it kinda baffles me to see people really trying to deduct anything from those charts related to balance, in a range of MMR, where people have problems with Scouting, constantly building workers+army, Supply blocks, unit compositions, unit control, macro in general and most importantly: dealing with cheese/all-ins (scouting + defending), because in those ranges, cheese simply reigns supreme and literally everyone can cheese every game and rank up very fast to like 3500 MMR within days.
Another thing would be insta-leaves: some people lose a few games to cheese from one race and then decide to just leave every game they get matched against that race again. Many people have problems holding 12pool/ling-bane allins, so they get traumatized and just leave every game they get matched vs Zerg (instead of learning how to deal with it). This also gets shown clearly that the MMRs converge very quickly around ~4k MMR, where people mostly have learned dealing with cheese/all-ins.
In the end, the whiners will also never accept that every game they lose, is because of their own errors, that's why we will still hear "ZERG OP!" in every thread, in evrey game.
|
[QUOTE]On September 13 2020 18:56 suriel- wrote:
[QUOTE]On September 13 2020 14:09 dbRic1203 wrote:
because in those ranges, cheese simply reigns supreme and literally everyone can cheese every game and rank up very fast to like 3500 MMR within days.[/QUOTE]
What??? Dude dunno where you get your information from, but you have no idea what you talk about. This is only your stupid assumption and you construct the whole argument on this, which fails because the assumption fails. You have no data to back this up and your limited narrow view is bias af on this regard.
|
On September 13 2020 22:14 xsnac wrote:
What??? Dude dunno where you get your information from, but you have no idea what you talk about. This is only your stupid assumption and you construct the whole argument on this, which fails because the assumption fails. You have no data to back this up and your limited narrow view is bias af on this regard.
I mean, i've played in those ranges of MMR and like every second game was cheese. You can also create a new account and go Proxy/12pool every game and you will have a huge success, ranking up to 3500 MMR fast, that's what i did when i was fed up with people cheesing me, so i cheesed back and literally had like ~70% winrate in every matchup at one point.
Didn't even think one can be "biased" in that regard? Isn't it generally known that the 3k MMR ranks are just filled with cheese and general random plays of people not understanding the game?
|
only 8000 players who played at least 10 games ?
|
On September 13 2020 23:09 suriel- wrote:Show nested quote +On September 13 2020 22:14 xsnac wrote:
What??? Dude dunno where you get your information from, but you have no idea what you talk about. This is only your stupid assumption and you construct the whole argument on this, which fails because the assumption fails. You have no data to back this up and your limited narrow view is bias af on this regard. I mean, i've played in those ranges of MMR and like every second game was cheese. You can also create a new account and go Proxy/12pool every game and you will have a huge success, ranking up to 3500 MMR fast, that's what i did when i was fed up with people cheesing me, so i cheesed back and literally had like ~70% winrate in every matchup at one point. Didn't even think one can be "biased" in that regard? Isn't it generally known that the 3k MMR ranks are just filled with cheese and general random plays of people not understanding the game?
If you are a 4,5k player, you can get to 3,5k easily with absolutely anything, just like pros like BeastyQT troll low master players with thor drops etc.
Remember that cheese execution and follow-up will also be at plat/lowdia level, so it is not for granted that cheesing is that much easier than playing standard at different levels.
RL example, I just offraced toss and went proxy 4gate in 5 pvzs, got 2-3. It was nothing spectacular.
|
The mental gynastics being done here by Zerg players is just too funny. Wasn't it ''just Serral''? Why is Zerg dominating from bronze to pro play then?
This reminds me of 2011 when there were 20 Terrans in Code S and people said ''Its because Flash and Boxer play Terran, its only natural that BW bonjwas inspired the best players to pick Terran.
This graph might not reflect balance, it probably reflects what is common knowledge: Zerg is the easiest race to learn and be good with.
|
On September 13 2020 23:11 GSTL wrote:
only 8000 players who played at least 10 games ?
No. ~8000 was the number for accounts ranked when using random with 10+ games.
I did a quick check, and there are 68055 unique entries in the dataset I have for NA for names with more than 10 games (this would only count accounts with ranked using multiple races once). There's probably some duplicates that just happened to have the same name but are unrelated that got removed so the actual number is likely higher. This data isn't from Blizzard so I don't have unique IDs, which would make this more precise. Also keep in mind this ranked only so there are likely thousands more accounts that only play unranked, archon mode, co-op, etc.. The takeaway from this can be that a good chunk of people still play SC2. Across all regions, we're probably looking at into six figures of active accounts, which is pretty good for a 10 year old game.
On September 13 2020 14:09 dbRic1203 wrote:Show nested quote +On September 13 2020 10:24 Ben... wrote:I did the scatterplots for fun also (y being games played, x is MMR). They're kinda neat. For this first one, I did log scale for y axis since otherwise it's incredibly cramped. I set it ridiculously wide for this because I noticed something (spoilered because this one's kinda biggish download-wise): + Show Spoiler +Those vertical lines are interesting. A smaller scatterplot reveals the three most prominent lines on the right are around 4000, 4400, and approximately 4700-4800 MMR, which I'm guessing are cutoffs for various tiers of leagues, probably Master or Diamond 1. A bunch of accounts seem to cluster around the MMR cutoff for league tiers: + Show Spoiler +Cool stuff. edit: Yes, that giant space on the left is populated. There's someone with ~200 MMR. double edit: Cleaned up the plots a smidge and made the y axis more reasonable. I had to do it manually since 10e2 was too small and 10e3 left a bunch of empty space so I went with a range of 10 to 3500 games. When I went from 3000 to 3500 I saw one data point added so somebody has to have played a lot of SC2 this season. This is far less meaningfull than I expected Is it possible to plötzlich lifetime games against mmr as well?
I can't do that with the dataset I have unfortunately. The site in which the data is from gets their data from Battlenet for current season only.
|
On September 14 2020 01:30 Morbidius wrote:
The mental gynastics being done here by Zerg players is just too funny. Wasn't it ''just Serral''? Why is Zerg dominating from bronze to pro play then?
This reminds me of 2011 when there were 20 Terrans in Code S and people said ''Its because Flash and Boxer play Terran, its only natural that BW bonjwas inspired the best players to pick Terran.
This graph might not reflect balance, it probably reflects what is common knowledge: Zerg is the easiest race to learn and be good with.
And they won't quote statements like these that pretty clearly show what's happening. I'm sure a lot of zerg players had quite a bit of pride tied up in making it to diamond league. It must be difficult to discover that diamond zergs are as commonplace as silver protoss and silver terran players (Literally. By the numbers. Globally.)
But there's really no sense in trying to force a horse to drink water.
Can't complain about the balance in master's league though, so that's good
On September 13 2020 10:22 Alejandrisha wrote:Show nested quote +On September 13 2020 09:42 Ben... wrote:On September 13 2020 09:31 yubo56 wrote:On September 13 2020 09:18 Ben... wrote:I saw OP filtered out people who played fewer than 10 games (I didn't in my previous post) so I did also (it's trivial since the website provides number of games played as a column) just to see the difference. It's not really all that different but for consistency's sake here's the data: The graph looks basically identical: data is quite close too: Number of accounts with 10+ games by race:
protoss:31247
terran: 33855
zerg: 27118
random:8203
mean:
protoss:2848.8236310685825
terran: 2778.8631516762666
zerg: 2985.8191975809427
random:2872.1861514080215
median:
protoss:2759.0
terran: 2674.0
zerg: 2920.0
random:2818.0
overall mean MMR: 2864.1406052398356
population median MMR: 2793.0
Mean number of games played so far this season:
factoring in accounts with fewer than 10 games played: 55.53022382348439
ignoring accounts with fewer than 10 games played: 90.96791571651913
Hi, I asked for a log y axis plot but OP doesn't seem to have responded; since you seem to be checking this topic, would you be able to post one with a log y axis as well? Would be able to show the long tail at higher/lower MMR much better I definitely agree that using mean/median/mode to draw balance conclusions is extremely dubious, but it should be fun to theorize nonetheless Does this work? It's for accounts with more than 10 games. All I did was add ".set(yscale='log')". Nothing fancy. Yeah I don't think anything meaningful can actually be drawn from this data. I just like tinkering. edit: On September 13 2020 09:31 yubo56 wrote:
How are y'all scraping the data, did you just download all of the pages and parse the HTML? Pretty much. I just used the requests library to download each page (you can use the offset GET parameter they use for pagination to hop between pages), then used BeautifulSoup to parse the HTML and then cleaned stuff up a bit before chucking it in CSVs. Nothing too advanced. thank you, this is a much better graph. still don't think we can draw real conclusion from it. but this is better in that we can actually see the differences
Also.. Log transformation on the y axis? That just collapses differences in the middle of the graph and exaggerates differences at the ends of the graphs. There's no basis for applying a log transformation to population data in that way. It's just dishonest.
|
On September 14 2020 01:30 Morbidius wrote:
The mental gynastics being done here by Zerg players is just too funny. Wasn't it ''just Serral''? Why is Zerg dominating from bronze to pro play then?
This reminds me of 2011 when there were 20 Terrans in Code S and people said ''Its because Flash and Boxer play Terran, its only natural that BW bonjwas inspired the best players to pick Terran.
This graph might not reflect balance, it probably reflects what is common knowledge: Zerg is the easiest race to learn and be good with.
At top pro level, TvZ actually seems to be in a very good spot right now. The Clem vs Reynor series was a nice and very recent example.
After the last round of nerfs and a reasonable map-pool, Zergs actually have to make plays to win now, and can't win of the back of only defending harassment, macro and creep anymore. Good strategic and technical zerg play is also still rewarded.
|
On September 14 2020 01:30 Morbidius wrote:
The mental gynastics being done here by Zerg players is just too funny. Wasn't it ''just Serral''? Why is Zerg dominating from bronze to pro play then?
This reminds me of 2011 when there were 20 Terrans in Code S and people said ''Its because Flash and Boxer play Terran, its only natural that BW bonjwas inspired the best players to pick Terran.
This graph might not reflect balance, it probably reflects what is common knowledge: Zerg is the easiest race to learn and be good with.
Yet 45% of GM is protoss, while only 25% of GM is zerg.
https://www.rankedftw.com/stats/races/1v1/#v=2&r=-2&l=-2
|
On September 14 2020 02:02 ThunderJunk wrote:
Also.. Log transformation on the y axis? That just collapses differences in the middle of the graph and exaggerates differences at the ends of the graphs. There's no basis for applying a log transformation to population data in that way. It's just dishonest.
Hey, I'm the one who requested it. The reason one would do that isn't to make assessments of means / modes, or areas under the curve any more, but just to understand the shape of the tails of the distribution across races. It's possibly dishonest put the plot out there and say "this is the MMR distribution," since a lot of people don't know how to read log scales, but it's pretty common practice to do this when you're not interested in the peak of the distribution but the tail and scaling of the tail.
For instance, any curve with a peak looks plausibly like a normal distribution, but if you log scale it you immediately can tell whether it's quadratic and over what range it's normal. The question I was trying to answer was whether the tails of the distributions look any different between the three races, since we can already see the peaks very well, and a log plot is a great way to see both tails on the same plot. I think it's a useful tool if you know what you're trying to read
|
On September 14 2020 06:55 yubo56 wrote:Show nested quote +On September 14 2020 02:02 ThunderJunk wrote:
Also.. Log transformation on the y axis? That just collapses differences in the middle of the graph and exaggerates differences at the ends of the graphs. There's no basis for applying a log transformation to population data in that way. It's just dishonest. Hey, I'm the one who requested it. The reason one would do that isn't to make assessments of means / modes, or areas under the curve any more, but just to understand the shape of the tails of the distribution across races. It's possibly dishonest put the plot out there and say "this is the MMR distribution," since a lot of people don't know how to read log scales, but it's pretty common practice to do this when you're not interested in the peak of the distribution but the tail and scaling of the tail. For instance, any curve with a peak looks plausibly like a normal distribution, but if you log scale it you immediately can tell whether it's quadratic and over what range it's normal. The question I was trying to answer was whether the tails of the distributions look any different between the three races, since we can already see the peaks very well, and a log plot is a great way to see both tails on the same plot. I think it's a useful tool if you know what you're trying to read
Fair enough
|
On September 13 2020 23:22 Slydie wrote:Show nested quote +On September 13 2020 23:09 suriel- wrote:On September 13 2020 22:14 xsnac wrote:
What??? Dude dunno where you get your information from, but you have no idea what you talk about. This is only your stupid assumption and you construct the whole argument on this, which fails because the assumption fails. You have no data to back this up and your limited narrow view is bias af on this regard. I mean, i've played in those ranges of MMR and like every second game was cheese. You can also create a new account and go Proxy/12pool every game and you will have a huge success, ranking up to 3500 MMR fast, that's what i did when i was fed up with people cheesing me, so i cheesed back and literally had like ~70% winrate in every matchup at one point. Didn't even think one can be "biased" in that regard? Isn't it generally known that the 3k MMR ranks are just filled with cheese and general random plays of people not understanding the game? If you are a 4,5k player, you can get to 3,5k easily with absolutely anything, just like pros like BeastyQT troll low master players with thor drops etc. Remember that cheese execution and follow-up will also be at plat/lowdia level, so it is not for granted that cheesing is that much easier than playing standard at different levels. RL example, I just offraced toss and went proxy 4gate in 5 pvzs, got 2-3. It was nothing spectacular.
That was also a RL example of mine: when i started playing the game in Fall of 2019, i got placed into Plat3. Playing games there, i got cheesed very much, so i just started to cheese back.
I'm currently a ~3900-4100k player, but it doesn't take great effort to cheese someone. You can name 20 other former GMs that smurf their way into masters, it doesn't change the fact that cheese is easier to execute than it is to defend:
- you need to scout cheese
- after having scouted, you need to know what kind of cheese it is
- after knowing what kind of cheese it is, you need to know how to react to it
- after knowing how to react to it, you need to know how to handle it (micro)
things needed to know for executing cheese:
- learn the build order, which only covers the first ~3 minutes most of the time
- learn to handle it through repetition (micro, multiple games)
Cheese is so strong, that even highest pros have success with it (see Scarlett throwing out Rogue of GSL, or the countless ProxyRax games, or Stats/Trap losing to a few lings in GSL, because not having a proper wall)
Your anecdotal sample size of 5 games is cute, but it doesn't change the fact that doing this a few times more, you'll have a high win rate. I mean, people literally get GM with just cannon rushing (printf, DemusZero for example).Likewise, it's very trivial to go for a 12pool or a banebust every game and collect free wins. That's a RL example, because i did just that: 13/12 every game against Protoss and Zerg and i had about a ~70% win rate and got to 3.5k really fast.
And please, let's not shit outselves here: 3k MMR is basically garbage since people have huge other problems other than balance, like proper macro and 0 understanding of the game.
If you want a Pro to confirm how ridiculous it is to bring the "imbalance" argument into 3000 MMR range, check out some of Harstem's "Is it IMBA or do I suck" - series, there are even some "low masters" players sending in replays, but during the analysis, you can clearly see they still make Gold league mistakes. For example the "Spines are IMBA" one
|
On September 14 2020 17:22 suriel- wrote:
Your anecdotal sample size of 5 games is cute, but it doesn't change the fact that doing this a few times more, you'll have a high win rate. I mean, people literally get GM with just cannon rushing (printf, DemusZero for example).Likewise, it's very trivial to go for a 12pool or a banebust every game and collect free wins. That's a RL example, because i did just that: 13/12 every game against Protoss and Zerg and i had about a ~70% win rate and got to 3.5k really fast.
This is true. I just got back in the game 2 months ago and I went up to 5k3 by just 1base allining all my games with WoL units.
Ladder shouldn’t be used to determine balance. Mid GM and under players are just not good enough, and high GMs don’t even ladder seriously. Objective Pro feedback and High level tournament results is the best data for balance imho.
|
On September 13 2020 12:33 yubo56 wrote:Show nested quote +On September 13 2020 10:24 Ben... wrote:I did the scatterplots for fun also (y being games played, x is MMR). They're kinda neat. For this first one, I did log scale for y axis since otherwise it's incredibly cramped. I set it ridiculously wide for this because I noticed something (spoilered because this one's kinda biggish download-wise): + Show Spoiler +Those vertical lines are interesting. A smaller scatterplot reveals the three most prominent lines on the right are around 4000, 4400, and approximately 4700-4800 MMR, which I'm guessing are cutoffs for various tiers of leagues, probably Master or Diamond 1. A bunch of accounts seem to cluster around the MMR cutoff for league tiers: + Show Spoiler +Cool stuff. edit: Yes, that giant space on the left is populated. There's someone with ~200 MMR. double edit: Cleaned up the plots a smidge and made the y axis more reasonable. I had to do it manually since 10e2 was too small and 10e3 left a bunch of empty space so I went with a range of 10 to 3500 games. When I went from 3000 to 3500 I saw one data point added so somebody has to have played a lot of SC2 this season. That vertical line is super cool. Think those cutoffs are for d2-d1, d1-m3, and m3-m2 respectively, that's a cool feature to see. I guess it's too dense to see the differences between the races, it would be super cool to see something like "terrans gain MMR fastest, relative to # games played, since their macro is most intuitive/similar to other games" but that's probably too hard to quantify. Thanks again for keeping us posted dude ur awesome <3
Is there any correlation between amount played and mmr? this graph seems to indicate that no.
|
On September 13 2020 00:22 yubo56 wrote:Show nested quote +On September 12 2020 22:47 Dangermousecatdog wrote:
By the way those numbers aren't the node either. That can only be shown from the actual data table. Like I said, the more I look at the graph, the more nonsensical it appears. I'll have to see the data myself, but if it's from that same link then what we have is just a bullshit graph. There is only 7 datapoints per race so how could he make such a graph to present the data? Wait, why does rankedftw only give 7 datapoints per race? https://www.rankedftw.com/ladder/lotv/1v1/mmr/?f=eu If you go here, there are 160239 datapoints for EU alone, it seems plausible that that OP generated the plot from rankedftw.com Edit: your argument seems to be that the graph doesn't represent the 7 leagues correctly. Yes, indeed, if you want the % of gold players, you should integrate the curve between min(gold MMR) to max(gold MMR), and you'd get the correct percentage. But that's just b/c the transformation between leagues and MMR isn't linear, which makes sense since they're based on percentages and you have long tails. In other words, the statement "most protoss players are gold" is wrong, but "most protoss players are in the range 2300-3100 MMR" is perfectly correct, one is based on a league PDF and the other is based on an MMR PDF.
Compare this graph and the ones posted by ben.... ...
...The difference is astounding. One looks like an artifact produced by low data points and smoothing with it's near poisson distribution which is not what whould be expected from what we know of MMR, which is why I say it looks like 7 data points or super smoothed out. The other look like a graph. i;ll be interested to see when EU looks like as produced by ben... .
As for " "2300-3100 MMR" is perfectly correct" you basically wrote that at least half the players are between silver and plat MMR.
No shit sherlock! Blizzard has consistently stated that Silver-Diamond are 20% each since WoL. You don't need to look at a graph to work out that any three leagues between Silver and Diamond will be at least 50% of a population. I could be blindfolded and shown any randon graph and I could say Silver- Plat MMR will be 60% of the population or Gold- Diamond MMR is 60% of the population. Because it's a fact irrelevant of whatever you think the graph may or may not be showing.
On September 14 2020 20:41 skdsk wrote:Show nested quote +On September 13 2020 12:33 yubo56 wrote:On September 13 2020 10:24 Ben... wrote:I did the scatterplots for fun also (y being games played, x is MMR). They're kinda neat. For this first one, I did log scale for y axis since otherwise it's incredibly cramped. I set it ridiculously wide for this because I noticed something (spoilered because this one's kinda biggish download-wise): + Show Spoiler +Those vertical lines are interesting. A smaller scatterplot reveals the three most prominent lines on the right are around 4000, 4400, and approximately 4700-4800 MMR, which I'm guessing are cutoffs for various tiers of leagues, probably Master or Diamond 1. A bunch of accounts seem to cluster around the MMR cutoff for league tiers: + Show Spoiler +Cool stuff. edit: Yes, that giant space on the left is populated. There's someone with ~200 MMR. double edit: Cleaned up the plots a smidge and made the y axis more reasonable. I had to do it manually since 10e2 was too small and 10e3 left a bunch of empty space so I went with a range of 10 to 3500 games. When I went from 3000 to 3500 I saw one data point added so somebody has to have played a lot of SC2 this season. That vertical line is super cool. Think those cutoffs are for d2-d1, d1-m3, and m3-m2 respectively, that's a cool feature to see. I guess it's too dense to see the differences between the races, it would be super cool to see something like "terrans gain MMR fastest, relative to # games played, since their macro is most intuitive/similar to other games" but that's probably too hard to quantify. Thanks again for keeping us posted dude ur awesome <3 Is there any correlation between amount played and mmr? this graph seems to indicate that no.
It only take into account of the one season. It doesn't have total games played. Looking at the scatter plot, playing 1 game a season to keep the rank is still a thing. Anecdotal experience has shown the fastest way to improve is to play lots of ladder games.
Logically, you can't physically improve MMR unless you play ladder. No matter what you think your real skill level is, your MMR is nonexistent till you ladder and MMR can only be increased by laddering. Has anyone ever seen a master player with less than a hundred games? A GM with less than a thousand games? The actual mean of games plays of those ranks would be vastly higher than that. I would expect there is a very strong correlation between laddering and MMR until the higher end.
|
Mexico2170 Posts
I just want to comment on two things that have been bothering me for a while.
People at 3k understand the game. I'd go as far as saying most people in gold and silver have a base good understanding of the game. People that play StarCraft nowadays don't play it super casually. Most of the people I know in those levels are aware of tournaments, Build Orders, pros etc. There are other things that can impact MMR, things like the time you have available to play. You can watch GSL and read TL every day but if you can only play 1 game a day your MMR will improve more slowly. Also mechanical ability. Some people just have a lower skill ceiling than others, and might not be able to move their hands as fast , that's a fact. So how about we stop the snobery and stop discriminating lower leagues and feeling superior?
Second point: balance problems are real and they affect every league.
Sure, if you played better you could win more games. But you don't play better. Not yet. You improve over time.
If a unit, Build order or strategy is overpowered and an equally skilled openent uses it, you will lose games that you wouldn't have lost otherwise.
Yeah you can always improve and if you suddenly have 400 apm you could beat your opponent. And improving is something you can control, whereas balance is something out of your hands, so you should focus on getting better. But that doesn't mean balance doesn't affect players from every league.
There's also the fact that balance is not a constant. Some things that aren't very strong at one skill level can be slightly stronger at other MMR.
Even strategies have an impact. In parts of WOL and Host for example protoss had a good or even great winrate in PvZ, mainly due to cheeses and all ins. If you were a macro protoss? Though luck. Likewise other races. Muta ling might have a different winrate than roach Hydra. Mech has a different winrate than bio. So if you don't like a particular style or can't play it for some reason your winrate will vary.
To give a pro example, Mvp couldn't play bio anymore due to shoulder problems. If mech wasn't strong at that point in time he would do badly even if Terran as a whole was doing good.
Depending on how many people play different styles it could mask balance problems. For example maybe Terran has 40% winrate in lategame TvP but since they have a 60% winrate with early all ins the overall winrate looks "balanced".
Balance is much more nuanced than it appears and even overall winrates may be missleading.
|
TL-DR: zerg is the easiest race
|
|
|
|
|
|
EPT![[image loading]](https://i.imgur.com/kHr47aH.jpg)
![[image loading]](https://i.imgur.com/P7k89Wc.png)
