Just look at this:
http s://www.rankedftw.com/stats/population/1v1/#v=2&r=-2&sy=c&sx=a
For many years EU was on par with NA in terms of population. Nearly identical amount of players. Now it has dropped to 30k. What is going on ? EU is dead ? Or is it just bots failing to return proper data ?
Rankedftw is broken, it is written right on top of the main page.
EU should be more alive since it is where I see even NA players practicing besides KR server.
NA is the trashiest server contrary to what TehBAtz is trying to imply . NA is alive only for ladderers that are no different from the other server. The top of the server is absolute trash .
NA is the most active server by far. That’s simply a fact. 42% of the games played in SC2 are played on the NA server. EU is second with 36%. That might not seem like a big deal, but it means that wait times for a match will be much longer during the less busy times of the day and/or extremely long for the highest echelons of the ladder. I usually have to wait at least 2 minutes for a match as a Grandmaster, and it can’t find a good match. It’s often someone 1000 mmr below me. Part of this issue is the game duration because those who do play end up locked into a 45 minute game during which time they are unavailable to the matchmaker.
Find games within several seconds, that pretty far from the dead if you ask me.
Also we are better players than Americans 
I always had the opinion that EU is superior to NA in terms of skill (at least for masters/gm) but i don’t understand that significant population drop 90k → 30k. You can find games in seconds on any server, even on China which is the least populated server. I was just curious if this is simply data related error or indeed EU was dead.
That makes sense, but where did you get those numbers 42 and 36 % ? This would imply that KR server is almost completely abandoned with less than 22 %, assuming China server has some minor population as well.
Activity doesnt bother me right now on the D4 HYPE 
I would guess its some error in reporting… Or don’t they flush the player count to zero every time new season starts?
E.g. I have not played this season thus I don’t contribute to player count number.
PS. Nobody has fallen on my trolling attempt, nice job community 
Actually for me rankedFtw is the most important of all online tools. More important than aligulac and other sites. I am monitoring global population regularly to figure out whether it’s pointless for me to continue playing this game.
Numbers don’t lie (and they spell disaster 
)
october 2020 - 504 k
november 2021 - 359 k
november 2022 - 305 k
now - 152 k active players
Sorry bud, if these stats do not convince, i don’t know what will.
I am the dev of the most popular sc2 ladder tracker - sc2pulse, so I can share some info.
The global population and activity is slowly decreasing, which is not surprising because it’s an old game, but the game is nowhere near dead. The population is healthy and games played per day is close to what it was before the covid spike.
Here are a few points to consider:
Thank you for information.
Indeed if players are on many teams then rankedFtw can give invalid data or skewed data that does not reflect player count properly. So according to SC2Pulse current population stands at:
But still i don’t understand why EU population has dropped so drastically from 100k - > 30k. Where can i check EU population on SC2pulse ?
You can’t view players by region because player is a cross-region entity, i.e. player can have multiple teams in different regions, and it would interfere with new/old player numbers. You can untick other regions(checkboxes) and press “Show” to see region-specific stats for other charts(games per day and teams).
I recommend you to use the “Activity” tab, it shows normalized stats(x per day) which somewhat mitigates the fact that seasons have different length. But even in this case there could be false bumps(both negative and positive) just because how seasons work, but this is best stats we can get if we are talking about raw population numbers. But even if we compare raw numbers, the EU region peaked in s35 with 188k teams, and the current season has 93k(30k on rankedftw) teams, so it’s a 2x reduction, not a 6x reduction like on rankedftw. There was a big API update some time ago, rankedftw uses the legacy API that often doesn’t work, its numbers are just invalid most of the time.
IMO the main activity stats are:
There were 2 significant influxes of new players: f2p and covid. They somewhat revitalized the population but it will return to normal numbers over time. The population is decreasing, but it’s not as significant as some make it out to be, it’s not dead for sure.
Those are statements of opinion. Let’s get some actual data here and let people come to their own conclusion. Games played per day has decreased 143k to 81k over approximately 3 years. If that trend continues, it would be a decline of 2% per month. That decline is happening despite F2P and despite covid, both of which arguably should’ve increased the population. At most, you can say covid/F2P have decreased the rate of decline.
A reduction in games played probably has a close to 1:1 correlation with number of players, so if the games played has decreased by 2% per month then that means the available player-pool has decreased by 2% per month. In Grandmaster, that equates to 4 fewer players per month. That means there is a very rapid decline in the talent representation inside Grandmaster, which equates to less competition / an easier time to get into GM / less talent in GM. The GM threshold has dropped from ~5500 to ~4900 over the past 3 years, which confirms that this happening.
We can assign a win-rate decline to this MMR decline. If the GM floor declined from 5500 to 4900, then a modern LotV GM would have a 26% win-rate against a HotS-era GM. You say that this isn’t “significant” but that’s a statement of opinion and a pretty biased one at that. I think the vast majority of people would agree that this is an absolutely monstrous decline in talent.
As a Grandmaster myself, the decline in talent has absolutely ruined the game for me. I can literally spam nonsense build orders and get into GM. In the past, you had to have highly optimized build orders that required precise scouting and decision making. Now I can be like “so I am going to make 5 hatcheries before spawning pool, then I will mass slow lings and kill the opponent” and it legitimately works at GM level. Doing a ling flood allin like that in the past would be so hard to pull off it would come down to individual lings. The idea that you could do a slow zergling flood and/or waste money on 3 extra hatcheries was just plain insane. There is no way you would beat a HotS-era Grandmaster with this.
This decline in talent has also reduced the quality of matchmaking. The vast majority of people that I play are at least 500 mmr below my ranking. 1000 mmr below is quite common. The matchmaker just can’t find quality players because they don’t play anymore. 4 players per month over 3 years is equivalent to 75% of grandmaster leaving the game. Probably 90% of the games I play are a waste of time to play because the person I am playing against just doesn’t know anything about the game.
The biggest design mistake that Blizzard made and which precipitated this disaster was to artificially increase the game duration. It’s not uncommon for TvZ games to last an hour, for example. I spent a decade as a programmer in the game development industry, and everyone knows that the industry was moving towards shorter games. That’s why phone games are so popular. You can pop your phone out and play a quick game at any time and anywhere. Game companies realized the market was heading towards quicker games, and Blizzard was ramping up the duration of SC2 games. The way Blizzard’s designers handled SC2 was, from my perspective & experience, an absolute catastrophe.
The region lock was another absolute disaster because it alienated one of the most active player bases. Focusing the design of SC2 on mechanics instead of strategy/micro was an absolute disaster because it alienating people who weren’t strong mechanical players (which is the vast majority of the player base). The reason why League and Dota are so popular is because they’ve almost entirely eliminated the mechanical aspect of the game. Blizzard increased the mechanical requirements of SC2 repeatedly. The entire thing was one catastrophe after another.
In the far-off future, if I am a professor at a college and I am teaching prospective computer science graduates the concepts of game design, I will be pointing to SC2 as the perfect example of exactly what not to do with a game franchise. It went from being the game that created esports to having like 2,000 concurrent twitch viewers. Individual streamers in other games can pull in 10x that much. The way they handled this thing was an absolute catastrophe.
You are quoting my statements out of context of this thread. This thread is about EU being dead(30k players), which is not true at all. When I said it’s not that significant, I was referring to OP’s statement(30k players) and their source of information(rankedftw). I don’t deny that population is shrinking, but it’s not 30k players at all, it’s 90k(3x more). Rankedftw is broken, it’s a fact, not an opinion, but it’s at the top of google search results so ppl use it and draw wrong conclusions.
You are right that f2p and covid are just decreased the rate of decline, that’s what I meant that population will return to its normal numbers over time. I disagree with your reasoning. You took the covid influx as the starting point and compared it to the current season. The covid spike is not a normal situation/stats, you can’t use it because it returned to normal withing a year with 106k games/day at the beginning of 2021, so it just reverted to pre-covid numbers within a year. Covid players didn’t stay for long, unlike f2p influx that increased games/day from 70k/day to 105k/day and it was stable across several years. The population is declining but it’s not falling into oblivion like 143k->81k games/day(world) or 188k->30k teams(EU), it’s 105k(post f2p, pre-covid)->81k games, and 188k->90k teams.
I can’t make any comments about design decisions or masters/gm ladder because I’m just an ordinary pleb. I’ll take your words as they are because you are a GM and I’m not. In any case, this thread is not about high level players, it’s about population and activity. Masters+GM is only 4% of the population, 10% of games played, the rest are ordinary players that find a game within 10 seconds in prime time.
I get your sentiment, we see all those old yet successful games like cs or dota and we want sc2 to be among them. But it is what it is. It will be up the community to keep the game afloat. I’m doing my part by creating a replacement of old projects like rankedftw, sc2unmasked and matchhistory. IMO situations like this are really bad for the game when a player googles sc2 population and sees completely invalid data.
Nonapa . com
^— For accurate statistics, rankedftw is antiquated.
Sorry friend, but somehow i don’t believe that SC2 population decline is not significant. I think it’s huge! Even though rankedFtw may use some outdated and obsolite API and the data is not accurate i don’t think error margin is big. RankedFtw says global population is 152k, Sc2 pulse - 170 k. That’s just 18 k difference, not much. And given the fact that 2 years ago there were 500 k active players/teams current state of the game is in shambles - anyone who denies that, is delusional. But there is no stopping. Nobody and nothing will resurrect this dying game. Numbers will keep dropping until Microsoft shuts down servers for good. SC2 is not profitable any more. I think noone even buys skins/announcers etc.
season | duration(days)
--------+------
54 | 86
53 | 56
52 | 114
51 | 109
50 | 86
49 | 56
48 | 114
47 | 104
46 | 70
45 | 118
44 | 113
43 | 91
42 | 58
41 | 98
40 | 91
39 | 119
38 | 63
37 | 98
36 | 91
35 | 112
34 | 70
33 | 119
32 | 77
31 | 98
30 | 63
29 | 35
28 | 98
A really interesting statistic that I’d bet you could calculate is the probability distribution for the difference of MMR between opponents in 1v1. It’s the missing link to calculating how balance impacts the ladder. When a race is underpowered, they lose more, but the matchmaker puts them against easier opponents. This in theory creates interesting effects because terran being underpowered in TvP will also cause terrans to face against easier zergs even though it’s TvP driving the issue. That means terran being underpowered in TvP transfers a portion of the win-rate reduction to zerg. So, there is a dampening effect due to the matchmaking system. Knowing the distribution of MMR between opponents, especially if correlated with league, would allow for a much more accurate assessment of ladder balance.
I ran some monte carlo simulations of the ladder awhile back. They altered parameters of the ladder such as balance and race popularity to try to match the ladder statistics as closely as possible, and the most difficult aspect to get right was how the matchmaker matches players. This allows you to estimate things like PvT win-rates on the ladder.
I like your website by the way, it’s very useful for keeping track of smurfs. I check it after most matches. It allows me to pinpoint when I beat a pro player or streamer, which makes it fun to hop onto their stream and see their reaction.
Another missing parameter to my ladder simulations would be the distribution of games played as a function of MMR. E.g., being able to calculate the mean/variance of the number of games played for each MMR value on the ladder. The mean is the more important parameter by far.
I plan to eventually calculate these parameters myself, either through a random sampling of matches on nephest or sc2replaystats, but don’t have the time at the moment. I am absolutely swamped with work. I have 3 big projects delayed and one of my long time clients is wanting to resume a big project he put on pause about a year back.
You probably aren’t interested, but it’s worth asking. Would you be willing to sample 1,000 matches randomly per matchup, and dumping them to a CSV file for me? Having the MMR, games played for the season, and total games played for all seasons, for each player in each match, would allow me to calculate the missing parameters. It would have to be these parameters at the time of the match, obviously.
Sounds interesting. I’m busy myself, but I’d be happy to help you when I can. Add me on Discord nephest#5286.