Merit earners & the Pareto principle

[tranthidung]

We have experienced more than 2 years and nearly a half with the merit system. Today, during lockdown and boring time, I would like to present another overview analysis on merit system. In particular, I present the distribution of earned merits.

Let's look at the Pareto principle: https://en.wikipedia.org/wiki/Pareto_principle

Data source:

• The last Friday merit data dump, at 29may2020 02:50:29 (GMT time)
• LoyceV's Merit data analysis (full data since Jan. 24, 2018; not just 120 days) (for reference)

Results:

• Total users who received at least 1 merits: 34282.
• Number of top 20% of merit earners: 6857
• Total earned merits by 34282 users: 678082
• Total earned merits by top 20 merit earners: 595288

Final result:
Once again, we see another good example and usecase of the Pareto principle.  

• The top 20% merit earners have earned 87.8% of total earned merits on the forum.

Code:

     +-----------------------------------+
     |      cat   earned       te   p_te |
     |-----------------------------------|
  1. |  top 20%   595288   678082   87.8 |
  2. | rest 80%    82794   678082   12.2 |
     +-----------------------------------+

[1714858928]

1714858928

[ChuckBuck]

Wow   Excited to waiting to see the sharing of your experiences during the last active period   I am not interested in dry, boring data, but I am excited about experience sharing to operate in various fields. It is good to learn  

[iHodler]

I think you wanted to say "top 20% merit earners" got 595288 (87.8%) merits.

Anyway, nice info-graph.

[Mbitr]

Wow, very interesting graphic ! Not sure if this is a good thing or a bad thing !
I don’t want to go off topic too much , but have these stats changed at all over 2 1/2 years , in the sense of there being an increased amount of members who are merit sources ? Has the increase in merit sources over time  changed this 80/20 ratio. Hopefully, this makes sense ?

[tranthidung]

Wow, very interesting graphic ! Not sure if this is a good thing or a bad thing !

Nothing bad at all!

Applied the Pareto principle on earned merits on the forum, we can easily see that about 88% of total earned merits are belong to only to the top 20% of merit earners. I don't naively say that those top 20% users are real assets of the forum while the others are not but it is something very closely to that judgement. Even passed-away users, I believe they have received many merits within the last 2 ^1/2 year.

I don’t want to go off topic too much , but have these stats changed at all over 2 1/2 years , in the sense of there being an increased amount of members who are merit sources ? Has the increase in merit sources over time  changed this 80/20 ratio. Hopefully, this makes sense ?

No. I extropolated the statistics incorrectly. They are for earned merits, not for initially airdropped merits or merit-sourced merits. Merit sources only received sMerits to distribute them to good posts on the forum. In reality, merit sources don't receive a single merits because they are merit sources.

[mu_enrico]

Total earned merits by top 20 merit earners: 595288

The top 20 merit earners have earned 87.8% of total earned merits on the forum.

Top-20 or top 20% mate?
Is it because the top 20% created 80% of posts? Since it's usually related to the work done and post count is one of the easiest metrics available.

[Raytheon]

Top-20 or top 20% mate?

The table at the end of OP clearly showed "top 20%": 1. |  top 20%   595288   678082   87.8

Is it because the top 20% created 80% of posts? Since it's usually related to the work done and post count is one of the easiest metrics available.

Not really. In term of posts, I think the 80% of posts (bounty reports) belong to top 20% of non-merit earners. Yes, it's true the more quality post you make, the higher possibility you gain merits.

[o_e_l_e_o]

Top-20 or top 20% mate?

Definitely top 20%. 595,288 split among only 20 users would be 29,764 merit each.

Is it because the top 20% created 80% of posts?

No. The majority of posts on the forum are from bounty hunters. The Bounties (Altcoins) board has 7.5 million posts in it. That is more posts than in Bitcoin Discussion, Development & Technical Discussion, Mining, Bitcoin Technical Support, Project Development, Meta, Politics and Society, Beginners and Help, Off-topic, Serious Discussion, and all their child boards, combined. Announcements (Altcoins) is even worse with 8.6 million posts.

I think the issue with these data is that over a third of the total merit earners, in the region of ~13,000 users, have only received a single merit. This is going to skew things largely in favor of the top 20%.

[DdmrDdmr]

We can throw in a couple more:

Top 1% merited -> 45,32% of total amount of earned merits.
Top 2% merited -> 57,04% of total amount of earned merits.
Top 3% merited -> 63,66% of total amount of earned merits.
Top 4% merited -> 68,13% of total amount of earned merits.
Top 5% merited -> 71,34% of total amount of earned merits.
Top 10% merited -> 80,33% of total amount of earned merits.
Top 20% merited -> 87,79% of total amount of earned merits.

So the classical 80%:20% is really 80,33%:10% here …

Delimiting the above to earned merits during 2020 (right up to last Friday) we get:
Top 1% merited -> 23,61% of total amount of earned merits.
Top 2% merited -> 36,82% of total amount of earned merits.
Top 3% merited -> 46,27% of total amount of earned merits.
Top 4% merited -> 53,89% of total amount of earned merits.
Top 5% merited -> 60,07% of total amount of earned merits.
Top 10% merited -> 77,36% of total amount of earned merits.
Top 20% merited -> 88,86% of total amount of earned merits.

So if we delimit the numbers to the top 20%, the overall amount of merit share is very similar if we compare the whole Merit System history with 2020 so far, but distribution within is different (i.e. the top overall 1% have received 45,32%, whilst delimiting to 2020 the figure is 23,61%).

Note that the top rated are so within each calculation set. That is to say, the top merited 1% in the overall picture will not be exactly the same accouts as the top 1% merited for year 2020 (up until last Friday).

Mind you, numbers do not display the reasons behind, and context is important.

[tranthidung]

Top-20 or top 20% mate?
Is it because the top 20% created 80% of posts? Since it's usually related to the work done and post count is one of the easiest metrics available.

To be exact, it is for the top 20% merit earners. I calculated total earned merits of all merit earners, then sorted them out descendingly. After that, categorising them into 2 groups, with the cut-off is at 34282*20% = 6857^th position

If you look at whatever posts below, you will know OP's stats are not for top 20 merit earners:

• Weekly earned merits (median) of top 100 merited users
• [CLUBS] Top Merited-Users Classified into 4 Clubs

DdmrDdmr already gave more details above

Mind you, numbers do not display the reasons behind, and context is important.

You are correct too. I have not play with data this way but I have a theory that if I exclude earned merits in the first one week after the merit system was kicked off, the results will become more closely to 80-20 Pareto principle. Why I think so? Because there are some potential biases:

• Trial of merit usages in early hours/ days
• Abuses in early days

people made some trials in early hours/ days before they truly understood what merit is, and how valuable merit is, then use their merits wisely and more correctly. The same thing for merit abuses.

After all of such activities are cool down, statistics highly follow the 80-20 principle. There are some other bias events: like the Art contest, Jr. member demotion, something like that. At least with an raw analysis, the principle looks fit and works well.

Is it because the top 20% created 80% of posts? Since it's usually related to the work done and post count is one of the easiest metrics available.

Not really. In term of posts, I think the 80% of posts (bounty reports) belong to top 20% of non-merit earners. Yes, it's true the more quality post you make, the higher possibility you gain merits.

There is no statistics for the ratio of merit per post and only earned merits are counted here.

[Mrengage]

We have experienced more than 2 years and nearly a half with the merit system. Today, during lockdown and boring time, I would like to present another overview analysis on merit system. In particular, I present the distribution of earned merits.

Let's look at the Pareto principle: https://en.wikipedia.org/wiki/Pareto_principle

Data source:

• The last Friday merit data dump, at 29may2020 02:50:29 (GMT time)
• LoyceV's Merit data analysis (full data since Jan. 24, 2018; not just 120 days) (for reference)

Results:

• Total users who received at least 1 merits: 34282.
• Number of top 20% of merit earners: 6857
• Total earned merits by 34282 users: 678082
• Total earned merits by top 20 merit earners: 595288

Final result:
Once again, we see another good example and usecase of the Pareto principle.  

• The top 20% merit earners have earned 87.8% of total earned merits on the forum.

Code:

     +-----------------------------------+
     |      cat   earned       te   p_te |
     |-----------------------------------|
  1. |  top 20%   595288   678082   87.8 |
  2. | rest 80%    82794   678082   12.2 |
     +-----------------------------------+

To be frank on this I definitely to really understand what all this is all about probably I will want some knowledge of what the chart graphs is talking about. Am just seeing figures all round. More light show be thrown for me to grab.

[snipie]

To be frank on this I definitely to really understand what all this is all about probably I will want some knowledge of what the chart graphs is talking about. Am just seeing figures all round. More light show be thrown for me to grab.

In da forum there is 34282 account (prefer to say account since there is plenty of alts) that earned/received at least 1 merit. From that BIG number, there is 6857 accounts that received 87.8% of the total merit sent.
So the other 27425 (34282-6857) accounts received only 12.2% of the total merit sent.
You know like a company, big bosses receive +$1,000,000 salary and the good gentlemen grab their $8 per hour and say thank you sir /jk
+ see DdmrDdmr post above (
https://bitcointalk.org/index.php?topic=5252747.msg54547341#msg54547341)

[The Sceptical Chymist]

Top 1% merited -> 45,32% of total amount of earned merits.

There are nearly 3 million registered users here, maybe more. So the top 1% spans 30,000 users.

Yeah, and that statistic kind of makes sense to me.  There are thousands of active/inactive shitposter accounts created when the ICO boom was happening, and I'd never expect any of those to earn any merits.  There are also a lot of accounts that went inactive before the merit system was introduced, and there are a crazy number of Newbie accounts that were created and never used for posting.

So it's not as though there are 3 million active users right now.  Not even close.  I can't even imagine that there are anywhere near 30,000 accounts consistently earning merits, but that's just based on my observation of post quality.

[KaneVWE]

The top 200 merit earners lose around 80% of their merits if you subtract that which they have sent each other.
Some perhaps  even more if you just take away their top 10 merit fans.

So basically merit sources and their pals send each other lots of merit.

It's far more akin to pal points or let's give each other all the merit so we can dominate DT and thus the best rev streams and enforce double standards on everyone else.

If you want to ask some of these " top" .merit earners to present their very best original thought provoking posts that made any real lasting difference here on this forum get ready for a deafening silence from 90% of them.

Here have a try with

Thepharmacist. - an moron defending his merit swag bag who has never made an notable post or achieved anything here.
Lfcbitcoin- his first 3000 posts are mostly 1 liner speculative crap or sports betting.  Youd get banned for that now.
Tman ? - clearly retarded.


Well most of them really.

Also your principle only considers those that have earned 1 merit.
Making the merit distribution look more reasonable than it really is if you consider the forums membership or even prior active membership.

Also why worry about the distribution of a meaningless mess like merit at all.

The real genius members posting here have earned in some cases perhaps 20 merits total and you have complete imbeciles with 3000 earned. Lol

So from millions of members those giving merits  out, give them out mostly to each other and their pals or alts.
I'm sure that principle is evident somewhere. ....like the central banking system, big corps etc

The selfish self serving colluding scum principle let's call it    ... pareto doesn't do it justice.

[DdmrDdmr]

<…>

I’ve created a Google Doc with all the accumulated data as of last Friday, in case anyone wants to take a look:
https://docs.google.com/spreadsheets/d/1_vYXRLlMQsdsw2teukQMZ-4vL3YE8J-tuR1zAP8sM9s/edit?usp=sharing

Columns:
meritReceived: Number of received earned Merits
nUsers:            Number of Users that have received meritReceived Merits
nUsersAcc:      Number of Accumulated nUsers (as we scroll down the list)
MeritsGroup:   meritReceived * nUsers
%AccMerits:    % of overall merits, Accumulated as we scroll down the list
%AccUsers:    % of overall merited Users, Accumulated as we scroll down the list

So in order to see how many users have earned at least 10 Merits, we’d go down to row 641 and we’ll see that 9.768 users have earned at least that amount. Likewise, row 643 shows that 13.857 users have earned at least 5 Merits.
The top 30K (earned) merited profiles are really read on row 647, rendering 1 Merit or above.

Additionally, we can see on row 316 that the top 1% merited users now accumulate 47,13% of the earned merits, and on row 637 we can read that the top 19,63% accumulate 88,70% of earned merits. The update for the overall data is now that of:

Top 1% merited -> 47,13% of total amount of earned merits.
Top 2% merited -> 59,08% of total amount of earned merits.
Top 3,02% merited -> 65,70% of total amount of earned merits.
Top 4,01% merited -> 70,07% of total amount of earned merits.
Top 5,02% merited -> 73,26% of total amount of earned merits.
Top 10,13% merited -> 82,07% of total amount of earned merits.
Top 19,63% merited -> 88,70% of total amount of earned merits.