Topic: sMerit Senders & Receivers – Weight of top 100 and 200 weekly contributors - page 2. (Read 641 times)

I don`t know what`s before in this case: the chicken or the egg. I`ve seen that the Spanish board is slowly being full of nonsense posters polluting some good discussions without even giving a thought to the opening thread; or by threads that are just a link to another source (these ones I rarely read, I find kind of annoying picking some mysterious link in order to understand an OP`s contribution).
In general, the Spanish mean has, as much, 10 good posters, and some of them are moving to the English one, due to the lack of quality in the local, I included.
But this is happening because it is easier to get merits on the English one? Or because of the lack of interesting posts? Are we losing our very best posters on the Spanish just because there is nothing for them in there anymore? And, do you think merits are related? Meaning: on the Spanish local the merited ones are always the same (because there are few good ones) but it can be uncomfortable for them, given all the Inquisitorial processes we have had. I don`t know, but maybe the locals are just disappearing into a centralization force because, given the lack of merits awarders, they prefer to receive merits from others, or because they prefer to contribute in some substantial discussions instead of the ones having a place now in the Spanish.

From a forum’s point of view, it is kind of bad really. If you turn the numbers around, it reads that 95,24% are spamming or not posting anything too remarkable….

It would be nice to know the definition of “active member”. When dealing with a large quantity of users, I often create a “state machine” where one user evolves/regresses between states over time and events.
For example (simple cases):
-   Active Heavy poster: At least 5 posts per day on average over last month.
-   Active Mid poster: At least 1 post per day on average over last month.
-   Active Low poster: Less than 1 post per day on average over last month.
-   Dorman Heavy poster: Was a heavy poster a month ago, but has stopped posting since.
-   .
The states help to get a general picture of the user base and their profiles. State evolution can be monitored and events triggered (in general) depending on the user state.

Quote from: joniboini on August 07, 2018, 07:37:02 AM

<...>
I think there is another problem. Some users who have smerit doesn't care about others. This is clearly evident in my local section. Most of the time the mods are the one who gives merits to members, and after that, it stopped. They save their smerit, probably to start their own merits giveaway thread. Or, there is no more good post to be awarded from their POV.

I’ve just looked over my local board after being away for a few days, and there’s nothing new recently worth meriting. Good posters there are rather scarce, and concentration of merit is really natural due to this factor (although it is also a concern, and often people pull the handbrake there for the same reason).

Also, awarding merit takes time, so playing the merit game for long may not be in line for everyone, and could decline overtime for lack of incentive from the awarder’s point of view.

I still think that there are quite a few that don’t know the difference between Merit and sMerit, believing that awarding substracts from your merit counter. I can’t put a figure to that, but I’ve encountered it often enough (although mainly with lower ranks).

I haven't seen recently a merit source application for the Indonesian board, hope it will be recognized as necessary.
Yep, also agree with you. Even tough mi local (Spanish) is not too bad about merit, I still think that most all the merited people have 1 or 2 merits as much per post, even if it is amazing. That`s a huge problem related to centralization, for this is going to be difficult for them to rank-up beyond member just because they can`t speak English.

On the other hand, I see every single day people like Jet Cash trying to work harder in their merit source task, but it doesn´t seem to go anywhere: most of the users are not deserves, for they are spamming the forum. There are also initiatives, like the "DMerit", proposed by JC, in which the creation of a "merit team" is proposed, meaning that we can implement a new system in where the merit sources can merit to some other legendaries or people they trust with a different kind of merit, so a team can be created and the merits can go to the locals.

Whatever, this is a problem in crescendo. From my perspective, maybe this is time for the people to change a little their mindset. I mean: those having smerits shouldn`t keep them as a "precious" treasure, and, in general, perhaps this is time to low down a little our standards: if a user has a pretty decent post history and if it is obvious that the user is trying hard, well, why don`t give this guy/girl more than one merit.

Anyway, it is happening. Suchmoon, for instance, has increased the number of merits he`s giving per post, as well as other people, taking a look at the post history and giving merits by the "general behaviour" of a user, instead of giving a single merit by post. I think this is a great mindset we all can have in mind when awarding someone: to check the history and to give merits based on that.

Now, the situation is pretty dramatic also. If we take a look at the beginner's board, for instance, it has been polluted with a lot of bounty threads, meantime the users are complaining about merits. Bounties or infinite guidelines are not very merited by now, for I think we all are quite tired of that. That`s why I come to conclude that the JC idea, the Merit Team is one of the best in the forum. I think there are enough merit sources by now, but badly distributed, maybe by the creation of a team, it could be fixed.

I think there is another problem. Some users who have smerit doesn't care about others. This is clearly evident in my local section. Most of the time the mods are the one who gives merits to members, and after that, it stopped. They save their smerit, probably to start their own merits giveaway thread. Or, there is no more good post to be awarded from their POV.

Let's assume we add more 20% with this number 18,292 to have a fair number.
18,292 + 20% of 18,292 = 18,292 + 3,658 = 21,950 the quality numbers

Once I mentioned on a post that, it's better to have 1 quality member than 100 spammers.
According to Vod's bpip the active member are 460,948 in total which means 4.76% users are actually making good posts. Not bad IMO.

I guess a lot of us also think the same. We need more merit sources to make the whole system decentralized. Somewhere I have seen someone proposed an initial target is to have 200 sources and then gradually the number can be increased.

The main challenge for theymos would be the selection process. It's very hard to pick a good merit source I guess although we already have few applications are pending. For some reason we are seeing theymos is a bit quite here, possibly his hours are already booked with some other tasks. It's really not easy to manage a community of 460,948 active members. I wonder how he keeps himself calm. 

1. Introduction:
The initial Merit spirit I believe is that of a decentralized system, where many members are responsible for meriting other user’s posts. I started to get the feeling that Merit was getting more concentrated over time, in terms of who is awarding it, so I went on to see just how concentrated it was, and if the ratios steepened as time goes by (as it turns out,  they do).


2. General summary:
Total sMerit:   201.035 sMerits have been awarded in total.
Total Txs:         90.737 distinct transactions.
From:              15.980 members have awarded at least 1 sMerit.
To:                  18.292 members have received at least 1 sMerit.

That gives us the broad picture of how many forum members are in the Merit game of awarding or receiving, not too large a scope in relation to all the Bitcointalk number of members, but considering the amount of spam that the forum has, it even seems surprising that 18.292 forum members have been merited.
This is nothing really new so far though, since we’re a few that report this data on a weekly basis. It does serve us though as a context to the following sections.


3. Top 200 Senders
This is where it starts to get interesting. What I’ve done here is, for each natural week (Monday to Sunday), aggregate the amount of sMerit sent by the top 200 sMerit Senders, and calculate the percentage it represents out of the total sMerit awarded for the given week.
Weeks are numbered 4 onwards, relating to the natural calendar week (4 is therefore 22/01/2018 .. 28/01/2018, and so on). I’ve use the week number and not the full week date interval due to a better representation on the charts.

Bear in mind that the top 200 Senders are different each week, and are not a fixed set. For example, I’m computed in the top 200 Senders on 17 of the 28 weeks, but I did not qualify for all the weeks. In the top 200 we’ll likely include all Merit Sources (84 now) as well as other members.

The overall average sMerit that the top 200 Senders have awarded is 58,99% of the total sMerit, but this varies from week to week as can be seen on the following graph:


The vertical bars represent the total sMerit awarded by the weekly top 200 Senders. The line chart represents the % of the total sMerit for the given week. Both of these data numbers figure in the table at the bottom of the graph. The yellow bar is last week, which is incomplete due to data (merit.txt file) being published on Fridays.
What we can see is the evolution: the top 200 Senders started off being behind the mid-forty percent range of total sMerit during the first few weeks. The tendency has slowly evolved and is now days in the 75-80% range. That’s a 30 point increase since the beginning!


4. Top 100 Senders
Likewise, the top 100 Senders (a subset of the above) has the following trend line:

The evolution for the top 100 Senders started off being around 30% of total sMerit during the first few weeks and is now in the 60-65% range. Here again we have the 30 point increase comparing to the beginning.
The overall average sMerit for the top 100 Senders is 45,21% of the total awarded sMerit.


5. Top 200 Receivers
Similarly, the top 200 Receivers has the following trend line:

The top 200 Receivers started off being around 50% of total sMerit during the first few weeks, but has slowly moved into the 60% range and lately boosting a bit into the 70% area.
The overall average sMerit for the top 200 Receivers is 53,67% of the total awarded sMerit.


6. Top 100 Receivers
Likewise, the top 100 Receivers (a subset of the above) has the following trend line:

The top 100 Receivers started off being around 35% of total sMerit during the first few weeks, but has slowly moved into the 40% range and lately boosting a bit into the 50% area or above (partial week though still in yellow).
The overall average sMerit for the top 100 Receivers is 39,35% of the total awarded sMerit.


7. Conclusion
Concluding numerically is simple: As time goes by, there is more concentration of sMerit sent by the top group of sMerit Senders and Receivers, although more pronounced on the Senders side of the scale.
This is likely due to an increasing importance of Merit Sources as times go by. Now I would venture to say that this was not the initial intention of the Merit System, but rather an organic sway from a more distributed system to an increasingly more concentrated system. This I figure puts extra pressure on current Merit Sources, as a large portion of the Merit System now depends heavily on them (bare in mind though that amongst the top 200/100 percent there are also non merit sources too as I said in the previous sections).
Maybe adding more merit sources would be a means to at least having more pairs of eye in the meriting awarding process, even if the total overall amount of sMerit assigned to them doesn’t budge at all.

On the receiver’s side, I find it difficult to conclude objectively. Concentration is rather heavy too, that is obvious. Subjectively nevertheless, I find my Merit Network (those I award sMerit to) growing at a very slow pace, since the members I merit tend to be the best and, once added to the network, new member addition seem difficult due to post quality of others by comparison, or by the spam forest blockage effect.