Quote from: LoyceV on August 09, 2018, 11:39:38 AM

Can you get me a list of all 1033 Member accounts that lost posts? I'd like to cross-reference them to the ones that earned exactly 10 Merit.

I've created two google sheet documents:

a) https://docs.google.com/spreadsheets/d/1fVkH12ZVS50LEJWG5WjuvLqdAB144zWxP0AIY3f5N2o/edit?usp=sharing
It contains all the users in segment '06. Negative' (not just Members, but Rank is included so it can be delimited by filtering).
There are very few cases there that have earned 9 or 10 merits in the past month and deleted a buch of posts, so I think it wont really be useful to you.

b) https://docs.google.com/spreadsheets/d/1u9bgcvKRDJwCi4NsuxFKjo57pPPWXCD-crsBwBJwJhA/edit?usp=sharing
This file contains all users with 9 or 10 merits on their profile (as of last scrape 06/08/2018) and 9 or 10 merits derived from the cumulative merit.txt files (therefore, earned merits with no airdrop). The total is less than the amount you pointed to (Ive got 1.672) , so there must be a criteria difference. The interesting column is the last one (MeritReceivedFromNUsers), since it tell us how many people have awarded each of these users with merit.

Quote from: vphasitha01 on August 09, 2018, 01:01:35 PM

<>
1) How you derived the posting segments (eg: Very Heavy >=5 posts per day). Because I believed that threshold number matters the most and it will clearly affect the data when we move to the comparison or talking about percentages.

The posts per day is calculated as an average per user per month. It is done by simply substracting the most current Posts value on User Profile from that of the DB snapshot a month ago, and dividing by the number of days between the two snapshots.

Quote

2) Can we cross-referenced the posts of the users(636+1033+973) actually having negative posting state with posts that got deleted on ANN/ Bounty threads?

Not with the data I have, since Im just using basically the User Profile information, without actually looking at posts themselves (Im not even sure if deleted posts can be read on a log on some sort, but it would be rather tedious to download, classify etc.). What I did is go through 20 or so profiles and by seeing what posts are still there you get a pretty good idea of what must have been deleted (i.e. some profiles have plenty of bounty reports still in their profiles, others one-liners trash, etc).