Quote from: LoyceV on October 01, 2018, 07:04:37 PM

I would have expected a much higher number. Apparently there's less high ranking (inactive or spamming) members than I expected.
Can you also produce the numbers for Full Members?

Sure, I'll create it as soon as I have some time on my hands, this week.

Quote from: LoyceV on October 01, 2018, 07:04:37 PM

Considering that less than 1% of all users has received Merit, the percentage of high ranking accounts is much higher. Of course, many of them had sMerit to share with their alt accounts too, which means "merit received" doesn't necessarily mean the merit was "earned".

Yes, there may be incomplete data in the UserData table so, there could be inconsistencies.

Yeah... I agree with the "earned" part, for the sake of this experiment I considered all users to be fair and not cheaters. Obviously this can't be true Cheesy

Quote from: eddie13 on October 01, 2018, 07:18:56 PM

Something is wrong making your list incomplete.. Must not cover all inactive users..

I was just looking at kwukduck https://bitcointalk.org/index.php?action=profile;u=654

Legendary, no merits, not on your list..

Yes you are correct, that user is missing because it is not available in the DB. Unfortunately I do not control the DB. Most probably some users are missing.
If you do a simple search in the DB, you won't find any data for him in the UserData table:

Code:

SELECT * FROM UserData WHERE UserName = 'kwukduck'

Good catch though

If you search for mine, you'll get a result:

Code:

SELECT * FROM UserData WHERE UserName = 'vlad230'

Quote from: DdmrDdmr on October 01, 2018, 07:39:20 PM

I think that the data in the OP is completely wrong. The lists are not members that were airdropped merit and have not sent any, but rather members that have given merit and not received any themselves.

I guess you write something like this on Piggys SQL:

select rank, count(*) nUsers
from UserData u
where (select sum(merit)from MeritData m where m.ToID = u.UserId)=0
group by Rank

or
select rank, count(*) nUsers
from UserData u
where not exists (select 1 from MeritData m where m.ToID = u.UserId)
group by rank
[..]

I understand your concerns but I believe my queries are correct. I'm using a more complicated approach, research SQL JOINS.
It should be easy to check though, if you find any user listed in the OP that has received merit prior to Sept 28th (latest merit data date) then I'm wrong

Regarding the queries you listed there, I'm sorry to say they are wrong. The first issue I see, is that you're trying to sum() something which does not exist (merit received for these users).

Quote from: DdmrDdmr on October 01, 2018, 07:39:20 PM

You cannot know who got an initial amount of sMerit (less of all how much on an individual level), since the formula depends on their activity during the year previous to sMerit System kick-off, and you cannot know this since there is no available full DB snapshot (not just merited) dating one year before and on the date of the merit system kick-off.

Yes, that's true you don't have that info but you need to see it in a different way.
We don't have any records of the initial merit that was airdropped but we do have the ranks of the users already so, that means they received airdropped merit (since they already have a higher rank) and we don't care how much they received either. Also, we need to keep into account that they didn't earn any merit and still have a higher rank.

Let me know if you have any questions.

Just wanted to add that Piggy's tool is really nice but I could create far more complex data analysis queries if I had more data for the user table like activity, post count, latest date active, registered date etc. I hope he will add them soon but I know the limitation here is the size of the DB that is stored for free on github .