Sampling the User-base

12/21/2021

Proof Of Concept: Sampling a very tiny portion of the user-base on a very specific time frame.

This is a POC (proof of concept) for analysis of the Gab user-base as a whole for 'inactive/removed' users among the user-base and to determine time frames for number of registered users.. This is merely a POC, so it will include a nano/micro (very small) sampling of the user-base at very specific numbers in order to allow for determination of time frames for analysis.

First, we need to consider and define what will be the criteria used for deeming a user 'inactive'. For starters, if a user has never made a post I will deem that an inactive acct., if they've not made 5 posts in over a year, I'll consider that inactive as well. If they've been removed (deleted accts. for whatever reason) from the publicly accessible database records (throws an error response), and I'll consider that inactive as well. *Sounds reasonable, though could include some lurkers. Again, this is a POC, so keep in mind the criteria for inactive could be expanded upon other factors, as could the sample group size and time frames.
( IE; replace 'is it inactive' w/ 'is it a bot' and change the criteria appropriately for determination.)

I'll start off with the very first 5 users registered to the platform. (UID#'s 1-5)
Then we'll move to the 1 million registered users mark, and take the first 5 from that mark forward (1mil - 1mil+4) and do the same from the 2mil, 3mil, 4mil and 5mil marks for a total nano/micro (tiny) sampling of 25 users.

Lets get started with the very first 5 users registered to the platform. (UID#'s 1-5) #1 = (The original lead programmer/designer and Co-Founder of Gab.ai

UID# 1 = acct. username e | Display name Ekrem Büyükkaya | Status:inactive |
Creation Date:August 10, 2016
UID# 2 = {"error":"Record not found"} | Status:inactive
UID# 3 = {"error":"Record not found"} | Status:inactive
UID# 4 = {"error":"Record not found"} | Status:inactive
UID# 5 = {"error":"Record not found"} | Status:inactive
* I'm not in the least surprised by the above results, as this was after all over 5 years ago as of the date of this writing. Also, a lot of the first accts. created were 'test' accts., in fact the acct. Andrew Torba uses now was actually the 31st acct. to be created.

SUMMARY: 5/5 inactive *One may be active behind the scenes.

Moving on to the 1 millionth acct. registerd/created on the platform.
UID# 1000000 = "username":"GoubrinSraya-san" | No posts in over a year = Status:inactive |
Creation Date:March 1, 2019
UID# 1000001 = "username":"glenbot" | "statuses_count":0 = Status:inactive
UID# 1000002 = "username":"tkeme2urleeder" | No posts in over a year = Status:inactive
UID# 1000003 = "username":"clg_lim" | "statuses_count":0 = Status:inactive
UID# 1000004 = "username":"WEFA_9" | "statuses_count":0 = Status:inactive

SUMMARY: 5/5 inactive
Time frame to 1 million users = 2 years 6 months 22 days

Moving on to the 2 millionth acct. registerd/created on the platform.
UID# 2000000 = "username":"SteverinoDC" | "statuses_count":0 = Status:inactive |
Creation Date:December 15, 2020
UID# 2000001 = "username":"19bldgnut58" | "statuses_count":1 = Status:inactive
UID# 2000002 = "username":"DortGoodman" | "statuses_count":0 = Status:inactive
UID# 2000003 = "username":"Shelzy69" | "statuses_count":2 = Status:inactive
UID# 2000004 = "username":"Zoran7577" | "statuses_count":0 = Status:inactive

SUMMARY: 5/5 inactive,
Time frame to 2 millions users = 4 years 4 months 5 days

Moving on to the 3 millionth acct. registerd/created on the platform.
UID# 3000000 = "username":"osacody" | "statuses_count":0 = Status:inactive |
Creation Date:January 11, 2021
UID# 3000001 = "username":"Toader71" | "statuses_count":0 = Status:inactive
UID# 3000002 = "username":"protonsin" | "statuses_count":5 = Status:active?
UID# 3000003 = "username":"Mlagl" | "statuses_count":16 = Status:active
UID# 3000004 = "username":"Oldlifter59" | "statuses_count":0 = Status:inactive

SUMMARY: 3/5 inactive
Time frame to 3 millions users = 4 years 5 months 1 days

Moving on to the 4 millionth acct. registerd/created on the platform.
UID# 4000000 = "username":"BryonLynn" | "statuses_count":1 = Status:inactive |
Creation Date:January 21, 2021
UID# 4000001 = "username":"Imakittyo" | "statuses_count":5 = Status:active?
UID# 4000002 = "username":"Mexicanos_Patriotas" | "statuses_count":28 = Status:active
UID# 4000003 = "username":"JNadal556" | "statuses_count":0 = Status:Inactive
UID# 4000004 = "username":"Helioril" | "statuses_count":0 = Status:inactive

SUMMARY: 3/5 inactive
Time frame to 4 millions users = 4 years 5 months 11 days
(+10 days from 3 mil mark ( 100k/day for 10days))

Moving on to the 5 millionth acct. registered/created on the platform.
UID# 5000000 = "username":"rondahubschman6625" | "statuses_count":0 = Status:Inactive |
Creation Date:April 8, 2021
UID# 5000001 = error | "statuses_count":0 = Status:Inactive
UID# 5000002 = error | "statuses_count":0 = Status:Inactive
UID# 5000003 = "username":"richardthomas351" | "statuses_count":0 = Status:Inactive
UID# 5000004 = error | "statuses_count":0 = Status:inactive
SUMMARY: 5/5 inactive
Time frame to 5 million users = 4 years 7 months 29 days
(+3 mos from 4mil mark)

OVERALL SUMMARY: From the info. presented above, we'll say 5 of 25 registered users are actually 'active'. (see above for criteria to determine inactive)
5 of 25 = 20%
20% of 5000000 (5mil) = 1000000 (1mil) 'active users'?

CONCLUSION: The only thing conclusive at this point is the number of registered accts. (according to public api DB records) and the time frames within which those numbers were reached. I don't consider the number of inactive users estimate accurate by any means, as it relies too heavily on a simplified criteria used for an 'inactive' condition determination and much too small of a portion of the user-base as a whole. As mentioned previously, there is room for expansion and this is just a simplified proof of concept.
*I.E.: Pro statuses, donor statuses, investor statuses, verified statuses are among the other criteria that could be considered as well as, a much larger portion of the user-base being pulled for consideration (by automated means).

Sources: https://www.easysurf.cc/ndate3.htm For calendar calculations
https://gab.com/api/v1/accounts/100000 For grabbing dates of creation