|
metadata
Hi,
Since 5 weeks I've been trying to make an estimate about the number of active unique players each week. I'll be adding the totals for a month each 4-5 weeks (as the calendar dictates).
**Methodology**:
1. The ***total number*** of people who have gotten ***votes*** on a Sunday, less than 24 hours before voting closes.
2. ***Number*** of ***inactive players*** in my friend list to be substracted from the total.
3. ***Number*** of ***known alts*** in my friend list to be substracted from the total.
**Clarification:**
* Every week on Tuesday I start a list listing inactive players. This list is checked and rectified on Friday and Sunday. Inactive players who I strongly suspect to come back, are not removed. Inactive players below lvl 50 inactive for more than 5 days are removed, for inactive players above lvl 50 that period is extended to 2-3 weeks before removal. The remaining total number of inactive players left in my friend list (after removal) is substracted from the total.
* New players (or returning players that had been removed from my friend list) are added either when (re)appearing in the chatroom or in the high scores list (if they made one of the top 100 lists).
* I know this is not the most exact statistical method, which is why I've added a best case (3%) and worst case (10%) error margin, resulting in a range of active players. Without actual access to the players database this is the best I can do.
* I'm quite grateful to the people who questioned or even challenged my methodology -- those questions made me think and rethink the validity of those numbers, as well as my calculations. After having reread parts of statistics 101, I added two error margins (3% and 10%) in an effort to make the numbers more realistic, resulting in a +/- range. The 10% error margin may seem gross, but I consider it an adequate way to compensate for not knowing the exact number of alts as well as not all (new) players appearing in the high scores or the chatroom.
I'm not a statistics (or even a math) pro, so suggestions & improvements are most welcome!
MM
|
|
|
metadata
**November 2018 stats**
**4/11/18**:
inact:23
alts: 24
400 (votes) - 23 (inactives above lvl 50 or peeps known to come back) - 24 (known alts) = **353** active unique players
Best case error margin: 3% => range of 342-364, Worst case error margin: 10% => range of 318-388
**11/11/18: **
inact:25
alts: 24
393 (votes) - 25 (inactives above lvl 50 or peeps known to come back) - 24 (known alts) = **344** active unique players
Best case error margin: 3% => range of 334-354, Worst case error margin: 10% => range of 310-378
**18/11/18:**
inact:25
alts: 16
404 (votes) - 25 (inactives above lvl 50 or peeps known to come back) - 16 (known alts) = **363** active unique players
Best case error margin: 3% => range of 353-373, Worst case error margin: 10% => range of 327-399
**25/11/18:**
inact:32
alts: 13
400 (votes) - 32 (inactives above lvl 50 or peeps known to come back) - 13 (known alts) = **352** active unique players
Best case error margin: 3% => range of 341-362, Worst case error margin: 10% => range of 317-387
**02/12/18:**
inact:40
alts: 14
409 (votes) - 40 (inactives above lvl 50 or peeps known to come back) - 14 (known alts) = **355** active unique players
Best case error margin: 3% => range of 345-365, Worst case error margin: 10% => range of 320-390
**Weekly average for November: **
***353 active unique*** players.
Worst case: *310*, best case: *399*
|
|
|
metadata
idk if I'd show on it, I'm an ... hmm ... 'elder statesman' let's say, I won a long time ago, and now I hang around and when there's new content I nom it up. I don't solicit votes at all (and don't want any, if anyone's reading this post, give the shinies to people who will benefit). That might be nicely in your definition of inactive, not sure. Also, dk that there are many like me, so the methodology seems semi-decent to me.
|
|
|
metadata
> *Originally posted by **[mattender](/forums/10718/topics/1778998?page=1#12970836)**:*
> idk if I'd show on it, I'm an ... hmm ... 'elder statesman' let's say, I won a long time ago, and now I hang around and when there's new content I nom it up. I don't solicit votes at all (and don't want any, if anyone's reading this post, give the shinies to people who will benefit). That might be nicely in your definition of inactive, not sure.
Simple: when I see your icon not being greyed out and with a "5" on it, that means you're active :) And when you're not, you are counted as a "recurring inactive" because I've seen you come & go a few times, which is why I keep you in my friends list...
|
|
|
metadata
> *Originally posted by **[mattender](/forums/10718/topics/1778998?page=1#12970836)**:*
> idk if I'd show on it, I'm an ... hmm ... 'elder statesman' let's say, I won a long time ago, and now I hang around and when there's new content I nom it up. I don't solicit votes at all (and don't want any, if anyone's reading this post, give the shinies to people who will benefit). That might be nicely in your definition of inactive, not sure. Also, dk that there are many like me, so the methodology seems semi-decent to me.
matt... good to see you're still around... i do miss you and the other "elder statesmen that i no longer see here...
|
|
|
metadata
**December 2018 stats**
**2/12/18:**
inact:23
alts: 24
400 (votes) - 23 (inactives above lvl 50 or peeps known to come back) - 24 (known alts) = 353 active unique players
Best case error margin: 3% => range of 342-364, Worst case error margin: 10% => range of 318-388
**9/12/18**:
inact:45
alts: 24
425 (votes) - 45 (inactives above lvl 50 or peeps known to come back) - 24 (known alts) = 356 active unique players
Best case error margin: 3% => range of 345-367, Worst case error margin: 10% => range of 320-392
**16/12/18**:
inact:53
alts: 48
380 (votes) - 53 (inactives above lvl 50 or peeps known to come back) - 48 (known alts) = 279 active unique players
Best case error margin: 3% => range of 270-288, Worst case error margin: 10% => range of 251-306
**23/12/18:**
inact:70
alts: 48
463 (votes) - 70 (inactives above lvl 50 or peeps known to come back) - 48 (known alts) = 345 active unique players
Best case error margin: 3% => range of 335-355, Worst case error margin: 10% => range of 310-380
**30/12/18:**
inact:72
alts: 48
443 (votes) - 72 (inactives above lvl 50 or peeps known to come back) - 48 (known alts) = 323 active unique players
Best case error margin: 3% => range of 323-333, Worst case error margin: 10% => range of 290-355
**Weekly average for December:**
331 active unique players.
Worst case: 251, best case: 392
|