• 2763 Players on Java
  • us.mineplex.com
  • 12401 Players Online
  • 9638 Players on Bedrock
  • pe.mineplex.com
!
Attention Internet Explorer Users
To have the best user experience on our site please consider upgrading to Google Chrome or Mozilla Firefox

[Repost] Geography of the Mineplex Players

Discussion in 'Forum Discussion' started by FSCheetah, Mar 28, 2020.

  1. Hey guys, I decided to repost something I did 2 years ago because the old forums got removed. I'm thinking about making this for 2020 staff and possibly remaking the geography of mineplex players in 2020 (although I didn't find a database of players, if someone has a link please reply with it). I doubt the results will vary much, but let me know what you think.


    1. Determining the sample size

    Now, of course, getting all the users on mineplex would be nuts, so instead, we can get a proportion of Mineplex players and get estimated results.
    First of all, I calculated how many samples I would need for accurate results.
    Using -unapproved link-, I calculated which sample size I would need to figure where users of Mineplex come from. It suggested me to take a confidence interval of 3.1 (this means the margin error is ±3.1%). Out of 1139050 users on Mineplex, I would need a sample size of 999 users to accurately calculate where Mineplex users come from with an accuracy of 95%. To make it easier to read, I took 1000 users.
    [​IMG]
    Now, I needed to get 1000 users, luckily Mineplex has a database of players.
    I could of have written a program to get 1000 random Enjin IDs, but I figured it would be easier to do it myself. I sorted the database by Last Seen. I have a handy add-on for chrome which copies a whole column. So, after 20 pages of copying columns, I got a spreadsheet full of user IDs.

    [​IMG] [​IMG]

    Because some users were duplicated, I removed them and replaced them with random players on the 21st page.
    The way Enjin lists it’s users by Last Seen is… interesting, to say the least. I have no idea how it works because it doesn’t register by Last Seen. On one page it lists someone who was last seen in March and on the next one 3 minutes ago.
    Aforementioned doesn’t really affect the results because we need 1000 *random* players.
    3. Working with the 1000 IDs

    So, for those of you who don’t know, every user has a location set on their Enjin profile…

    After getting 1000 Enjin IDs, I needed to get all of their locations. Now, going into each page and copying the location would be *insane*, so instead, I wrote a small Python program which loads every page using selenium, copies the location of each user and extracts it into google sheets.

    If you didn’t get it, just imagine it this way: you load every page, then copy and paste the location of each user into google sheets. But instead of me doing it, the computer did it for me.
    [​IMG]
    This process took a while; my machine completed 1000 users in about 2.5 hrs.
    Now, you might be asking - surely there were better ways to approach this problem?
    Well… Yea, I could of have used the Enjin API, but I would need an API key, which I do not have, getting one from the Mineplex Staff? Meh… Buying an enjin domain? Meh… Writing in Python was easier for me because I worked with what I already knew.

    4. Working with more IDs

    After the machine processed 1000 users, I realized that a lot of them didn’t list anything in their location, in fact, about 27% didn’t. 27% without a location would really hurt the results, so I decided to get more users (starting from page 40) and run the program *until 1000 users have a location*, not random 1000 users.


    The machine finished after 45 minutes of work, and 373 users (270 of which had a location) were put into Google Sheets.
    [​IMG] [​IMG]


    5. Working with Google Sheets to create data
    Now that we have this massive google sheet with more than a 1000 users, it is time to sort and visualize them.

    For some reason, instead of 1000 users, I got only 951, but I decided whatever and called it a day there. The part where I work with sheets isn't really that interesting, so I do not include it.



    [​IMG]
    [​IMG]
    * The darker the blue, the more people come from there.
    ** If a country is gray it does not mean that nobody plays Mineplex from there. It means that less than 0.11% of players live there.

    Full table available here, sheet "By Country & Region"



    [​IMG]



    In the original thread, I also had proportion representation of staff and players, but I feel like that information is not relevant anymore. Either way, if you want to view it, you can go to the spreadsheet under the "Proportions" tab.
     
    Posted Mar 28, 2020
    Rerak likes this.
  2. i like this idea, but if you sort from last seen your results wouldnt be accurate since for example 8pm est would be 1am in europe, meaning that a lot of those players would be sleeping.
     
    Posted Mar 28, 2020
  3. Damn this is actually a pretty cool concept!

    Interesting for those who are into trivia.
     
    Posted Mar 28, 2020
  4. Indeed, good point. Well I guess it’s nice it sorted randomly.
     
    OP OP
    OP OP Posted Mar 28, 2020

Share This Page