tl;dr
Active users is inaccurately calculated on country leaderboards probably by getting every user that has set at least one ranked score. This is plausible due to how the leaderboards is likely counting the ranked score of every user that has submitted scores, therefore, ACTIVE USERS VALUE IN https://osu.ppy.sh/rankings/(INSERTMODE)/country IS FALSE, which is really unfortunate. Furthermore, there is no way there are 10 million active CTB players. I didn't know osu! would count somebody that's only played once in a game mode as an active player. Many have believed in these values, and I am posting this issue here to say that this should all be fixed.
So, I wanted to see how a percentile system would work in osu! where you grab the ranking you are in a mode and divide by total amount of active players in that mode. As there is no direct number that osu! directly provides you with, I started looking and realized that https://osu.ppy.sh/rankings/osu/country, https://osu.ppy.sh/rankings/taiko/country, https://osu.ppy.sh/rankings/fruits/country, and https://osu.ppy.sh/rankings/mania/country shows the active users in every country.
What does that mean? It means I could add up the active users in each country to get the active users of the whole world, or so I thought...
You see, this is not the first time somebody has thought about what I did. It was done once already around March 10, 2021 using Excel (https://www.reddit.com/r/osugame/comments/m1s442/there_are_15_millions_active_standard_players/), however, I released a program that pulls the live active users from each country live at that moment my program is ran and adds all the active users up in a gamemode by using the active users of each country as a guideline. You can find my project here which also talks about the issue I am going to mention here in this issue: https://github.com/TomyDoesThings/osuUserPercentileFinder/releases/tag/Experimental
As of around 12 PM EST of Feburary 1, 2023, these are the current values of the active users of each mode:
osu! - 22,241,423
Taiko - 11,601,962
CTB - 10,289,321
Mania - 11,282,923
Notice something odd? I did not at first as I did not think this could be a possibility, but how does a mode like CTB have 10 million active players? They are not just players. They are active players.
To put it into perspective, a player at the time of writing this is at 0.88 PP in CTB. That player goes by the username of Chr0ll0Lucilfer (https://osu.ppy.sh/users/23857539/fruits). At 0.88 PP in CTB, a player is at global ranking #2,227,932 in that mode. If there really are 10 million active CTB players, there is approximately 8 million players just like that player who are considered active. That simply is NOT an active player, so something is very wrong here, which leads to me to my next point.
So where is this active users number coming from? Well, me and probably many other osu! players thought it was a live count of all the active players in osu! which by opinion can vary from having set a play within a month or three months, but I just disproved that. I assumed then that this value could perhaps be some false approximation being used, so I thought of a logical way somebody would implement this.
That is when I saw ranked score and average ranked score.
By getting ranked score of any country and dividing it by the average score of that same country, you get nearly the same exact active users value for that country. By getting the performance of any country and dividing it by the average performance of that same country, you get nearly the same exact active users value for that country.
As I just proved to you that the active users value must be false, that average score and average performance is certainly also false because they are based off the active users of each country which I have just proven to you are false. Due to how a country's PP is probably added up from all the players that have gotten PP, this makes me think active users are players that set at least one ranked score in that mode, which is awfully off from the truth of what an active user/player really is.
We don't have 10M active CTB players (that play monthly) only growing. Even 1.2M active United States CTB players is very suspicious. So many of them are not even playing this game mode, but I will say this, many players do check out what other modes look like and may submit one score.
With this issue I have shown to you, I recommend changing what Active Users really means in https://osu.ppy.sh/rankings/osu/country, https://osu.ppy.sh/rankings/taiko/country, https://osu.ppy.sh/rankings/fruits/country, and https://osu.ppy.sh/rankings/mania/country. They are far from the truth and many have fallen into the trap of actually believing the active users values are true.
I made my program initially to see how good a percentile system would for this game, but with values this false, I can see why the osu!team could maybe a very hesitant on an idea like this, for medals such as the ranking medals.
tl;dr
Active users is inaccurately calculated on country leaderboards probably by getting every user that has set at least one ranked score. This is plausible due to how the leaderboards is likely counting the ranked score of every user that has submitted scores, therefore, ACTIVE USERS VALUE IN https://osu.ppy.sh/rankings/(INSERTMODE)/country IS FALSE, which is really unfortunate. Furthermore, there is no way there are 10 million active CTB players. I didn't know osu! would count somebody that's only played once in a game mode as an active player. Many have believed in these values, and I am posting this issue here to say that this should all be fixed.
So, I wanted to see how a percentile system would work in osu! where you grab the ranking you are in a mode and divide by total amount of active players in that mode. As there is no direct number that osu! directly provides you with, I started looking and realized that https://osu.ppy.sh/rankings/osu/country, https://osu.ppy.sh/rankings/taiko/country, https://osu.ppy.sh/rankings/fruits/country, and https://osu.ppy.sh/rankings/mania/country shows the active users in every country.
What does that mean? It means I could add up the active users in each country to get the active users of the whole world, or so I thought...
You see, this is not the first time somebody has thought about what I did. It was done once already around March 10, 2021 using Excel (https://www.reddit.com/r/osugame/comments/m1s442/there_are_15_millions_active_standard_players/), however, I released a program that pulls the live active users from each country live at that moment my program is ran and adds all the active users up in a gamemode by using the active users of each country as a guideline. You can find my project here which also talks about the issue I am going to mention here in this issue: https://github.com/TomyDoesThings/osuUserPercentileFinder/releases/tag/Experimental
As of around 12 PM EST of Feburary 1, 2023, these are the current values of the active users of each mode:
osu! - 22,241,423
Taiko - 11,601,962
CTB - 10,289,321
Mania - 11,282,923
Notice something odd? I did not at first as I did not think this could be a possibility, but how does a mode like CTB have 10 million active players? They are not just players. They are active players.
To put it into perspective, a player at the time of writing this is at 0.88 PP in CTB. That player goes by the username of Chr0ll0Lucilfer (https://osu.ppy.sh/users/23857539/fruits). At 0.88 PP in CTB, a player is at global ranking #2,227,932 in that mode. If there really are 10 million active CTB players, there is approximately 8 million players just like that player who are considered active. That simply is NOT an active player, so something is very wrong here, which leads to me to my next point.
So where is this active users number coming from? Well, me and probably many other osu! players thought it was a live count of all the active players in osu! which by opinion can vary from having set a play within a month or three months, but I just disproved that. I assumed then that this value could perhaps be some false approximation being used, so I thought of a logical way somebody would implement this.
That is when I saw ranked score and average ranked score.
By getting ranked score of any country and dividing it by the average score of that same country, you get nearly the same exact active users value for that country. By getting the performance of any country and dividing it by the average performance of that same country, you get nearly the same exact active users value for that country.
As I just proved to you that the active users value must be false, that average score and average performance is certainly also false because they are based off the active users of each country which I have just proven to you are false. Due to how a country's PP is probably added up from all the players that have gotten PP, this makes me think active users are players that set at least one ranked score in that mode, which is awfully off from the truth of what an active user/player really is.
We don't have 10M active CTB players (that play monthly) only growing. Even 1.2M active United States CTB players is very suspicious. So many of them are not even playing this game mode, but I will say this, many players do check out what other modes look like and may submit one score.
With this issue I have shown to you, I recommend changing what Active Users really means in https://osu.ppy.sh/rankings/osu/country, https://osu.ppy.sh/rankings/taiko/country, https://osu.ppy.sh/rankings/fruits/country, and https://osu.ppy.sh/rankings/mania/country. They are far from the truth and many have fallen into the trap of actually believing the active users values are true.
I made my program initially to see how good a percentile system would for this game, but with values this false, I can see why the osu!team could maybe a very hesitant on an idea like this, for medals such as the ranking medals.