Topic: Analysis - Merited on multiple Subsections and the Correlation Matrix - Updated - page 2. (Read 1084 times)

I totally agree with you this topic is very helpful in all the way. Merit is something that is earn with good work and comment not just with posting rubbish (seen  on local all over it).
My surprise was the same for this merit on local i expected somewhere else but one is good for start anyway.

I was going through the names, I saw names that are always on different lists about merit and reputation. Great names. So I was surprised and happy to see mine too. Thank you for creating this list. It is cool to know that I've been doing interesting topics and posts in different sections of the forum, instead of focusing on one place.


Have you considered the possibility of being 1000 times more comfortable writing deep texts in the native language? The time that is consumed, the difficulty of expressing everything you think, the difficulty of writing with the correct grammar, the difficulty even in understanding what others are writing. Even telling a joke or giving a funny answer is MUCH more difficult.

I do not believe that merits are only distributed among friends in those sections.

Cryptocurrency will only be adopted when greater incentives are given for the universalization and acceptance in every corner of the planet. Rather than being restricted to small bubbles where only those with excellent English, enough money for just Hodl without selling and deep knowledge of programming, encryption and technology related topics. I believe that there should even be a greater incentive for local sections to grow. With more sources of merit in them.

If we take global statistics, the global population is not fully represented here. What is a shame.


Indonesia+Brazil+Pakistan+Nigeria+Bangladesh=1Billion.

This is probably because in some countries the penetration of internet access is much higher. In the USA it is 85%, in Africa it is only 10%.

But the expectation is that the world will be completely connected in the next 10 years. That is, if the Bitcoin-related communities know how to open the doors to these people, instead of always trying to close them, we would have disruptive power over the banks. This has always been the concept of Bitcoin intended by Satoshi. And opening doors does not mean talking: learn English or disappear. It means allowing the growth of these communities.

It’s informative, and helps to monitor how things are going, but context is also required to interpret the information since not all is what it seems. For example, Ann Altcoin is the historically most merited subsection, although if you play around with the filters on the Merit Dashboard, you’ll see that it is losing its share of merit over time (in comparison to other subsections). There are probably no merit sources there, so I fell it will keel on losing its share gradually.
Nevertheless, merit distributed on the Ann section follows for the most it’s own patron on meriting, which may not concur with the post quality that often. A post above by @jackg breaks this down a bit for us.

By the way, I believe you have got a merit source on your local (Croatian) board, so if you write some good posts there you should get a decent chance. Another matter is if the merit ratio to post is adequate or not (0,113 on the Croatian Local over the last six months – see my last reply on this post to a link to the data source).
In my local board for example (Spanish 0,034 merits/post), the ratio of merit per post is rather low, although a bit better off than yours in terms of absolute merit. In my Local board, there are not very many decent posters, and if you put a bit of effort you will get merited. Nevertheless, the amount merited per post is low, so ranking-up on my local board by just posting local is a slow process. It also depends on what subsubsections you have within the Local board and how active they are posting (i.e. active local Ann threads for example will possibly bring the ratio down there and not boost it).

A bit of everything guess (somewhat like Anaxagoras' principle of "there’s a portion of everything in everything"), though opinion is based on observation and not empirical data. I guess a lot of ICO hype bares its toll too. Many of the Ann threads I’ve seen merit the OP, possibly in the hope to give more credibility to the ICO and thus improving the marketing of their investment to some extent. Also many Ops merit users within the thread, often for helping to keeping it bumped through their posts. Merit abuse/trading can easily be hidden there due to the amount of posts in the subsection, and some decent posts are to be found for sure.
In an ideal world, there should be plenty of decent post there since so many people are behind projects that relate to blockchain and its usage. In reality, the quick buck for some, and the quick million for others prostitute what could be a good section.

I appreciate your feedback to @Helana. A form veteran can always bring good insights on to motives behind the scenes.

Your questions are not bad for a newbie, and attitude does wonders in general. One tends to forget that the mapping of person to forum Rank is not linear to age, and seeing Newbie as the rank of a forum member is not the equivalent to a toddler in real life.

As to your question, I created a post on the matter just a month ago: Bitcointalk – Posts per Day - evolution during past six months. Specifically, there’s an entry called "4. Merit per Post Ratio" within the post that tells us the Merit per post on each subsection over the last 6 month (on aggregate) and the last month. If your interested, read the whole post as it gives you context to some strange negative numbers you’ll find on the embedded data table (quick mystery solver: it’s due to post deletion by forum members and mods in cleaning up processes).

The results in that post are very interesting, as we can see which subsections have the highest/lowest merit to post ratio. But again, context is important when interpreting the data. For example, Meta (0,3) , Serious Discussion (0,349)  and Bitcoin Development & Technical Discussion (0,286) have the largest merit/post ratio, but the degree of effort in the post there in general terms are noticeable. Market Place Altcoins (0,001) and Altcoin Discussion (0,009) has one of the lowest ratios, which is no surprise.

Maybe this happens due to the "bumping fever"?

Yep, I'm afraid that this is an issue, also there are external links offering bumping services like this one: https://www.seoclerk.com/Traffic/611685/Bump-Up-Your-Bitcointalk-Thread  (notice that they are bumping with different accounts!!! what a professional!!!)


Yep, you are probably right. One tends to think that the one's favourites boards should be the same for everybody, but of course, it doesn't. And yes, the Serious and the Ivory are quite silent, but, still, really merited. It will be interesting to see a correlation between the number of post on a board and the merits per board in the same table, just to be able to come to some conclusions.
If it has been previously done, please point me in the right direction, I'm just a newbie.

There are some gold mines in there sometimes but you have to be very lucky to find one near the top.
Some people with advanced issues with altcoins that get merit (often clients related to the forked coins that require advanced technical knowledge of bitcoin's design).
There are a few users who completely ruin everything for everyone and one user who was giving merit to supporters of the thread (who also conveniently turned out to be either his alts or part of the team - hence why they were in support of his ico).


Serious discussion and Ivory tower I guess are pretty small based on how much you need to rank up in order to be able to post there.

Well, in fact, is kind of surprising that the Altcoin and ANN sections are the most merited ones -if I did interpreted well the graphics. To me, those two are full of trash, in the best of cases.

To the OP, do you think that this situation is actually enlighting that there is a problem of merit trading -bigger than considered 'till now- in these boards? Or, on the contrary, that, hidden in between all the crap there are, in fact, some good posts in there?
For all that I can see so far -I've been a passive reader of the forum for about 5 months before even creating an account - I have been unable to find a single decent post in there, so when I first started to read your post, I thought that either Meta or Serious Discussion should be the most merited ones. I'm kind of surprised.

Also, the numbers obtained in the Deleted post are quite disturbing.

Anyway, thanks for this awesome work.

This is very good for me i probably for some other member with no merit or low merit. If we count out local board that are mainly for friendly merit giveaway i presume, ANN and altcoin section is mainly where merit is distributed.

Yes, the mining subsections are certainly for the very savvy on the matter, and require in-depth knowledge to understand the topics, let alone be able to add to the threads without feeling scenic fear. I bet you’ll be merited there though sooner or later…

Nicely spotted. As you say, Meta is a structural subsection for those on the list, with 49,64% of the received sMerits being for posts in that subsection. Of course theymos rules on Meta, and excluding his sMerits, the amount would be 35,84% for the forum members on the short list.
There’s an idea for yet another user profile badge! Something like the eye of an Oracle…

True, it can be rather off-putting in some of the sections. In the mega-thread areas, I normally tend to visually perform a filter on the number of posts in the thread. Any thread over 20-30 replies I kind of skip, unless they are reputably known threads.

It's an interesting analysis, and I didn't realise that I had been merited in so many sections. I try to visit a lot of them, but the problems in finding interesting thread titles puts me off opening a lot of threads. I don't open threads with titles like "Bitcoin", "problem", or even "help me". Maybe I should give some of those posters a chance to make their point. I did make a guide to starting threads, but nobody seems to read these topics, least of all newbies.

Extraordinary effort, glad to see people contributing valuable discussion to the forum. Usually among the people indicated as high merit achievers, you'll usually find one or two that are abusing merit. However, that list looks as legit as can be, no merit abusers there. I wonder who will be the first to achieve merit in every section (including archival!)

I don't need to repeat myself every time but man, this is great analysis, thanks. You deserve every single merit you got til now, don't listen to those who complain, they envy you.

Seems like the majority of those 27 have acquired the most merits in the meta section and this is not surprising as this was the "hottest" subject for the last couple of months.

I'm happy to see that just a bit more than the half of my merits (63%) are collected in meta, and big part of the rest are in the Beginners & Help (28%). So I'm on the right way in helping the community.I'm satisfied.

I can't wait to see your next idea. Really like surprises

Ah, It was strange to see myself in your list. I randomly open up all the boards every morning, and just read through them. Though I don't post often, I read a lot from various subsections(the latest one getting added is the Mining(altcoins)).

Bitcoin (Mining) section seems to be a dangerous board and I haven't made a single post out there. Unless we are really very familiar about mining and the mining equipment, we shouldn't post there. The mods are pretty strict out there and spam posting would lead us to the posts getting deleted more often or at times would get us banned too.

And I do hunt for some good tech based threads(which is literally very difficult) in the frog pond(Altcoin Discussion) which has helped a lot more people. Though I have requested a separate board for Altcoin Tech Discussion, I don't think it would happen as most of the users argue that it is a bitcoin forum.

You’re becoming a usual on my recent stats. That’s two out of two you’ve shown up on a highlighted list!
By the way, I’ve just added a third tab to the spreadsheet with the Rank and number of merited subsections for all merited users (thanks to @ theyoungmillionaire for the idea).

Yes, that is on my to do list too, since I’ve only been merited in 5 subsections myself. Way more space for knowledge here on the forum.

Yes, spam is kind of terrible on the Bitcoin subsections for example, but I persist on (only for now) reading there. The basically spam-free are the more technical subforums, better for reading than other subforums, but rather more challenging.

I did! See User's sMerit Network size – Data to select possible Merit Sources . I also included it as part of the Merit Dashboard to keep it up-to-date (Tab labelled Potential Merit Sources).

Finding a potential source (excluding current ones) is just one of the parts to the equation. It is just as hard to determine exactly where there is a lack of merit sources, and that is necessarily complemented with post content exploration, raising voices (backed by lots of real examples) where merit sources are required, and possibly a strategic vision.

For example, a given Subsection/Local Board may seem under-merited, and we may even find a few candidates numerically (pending manual whitelisting), but the number of unmerited posts in the subsection needs to be visually assessed by someone/a team in order to certify that a merit source is in fact required. That along with a cross reference to sMerit evolution in the subsection over time.

You make a lot of effort in these analyzes; I hope your analysis will contribute to make this forum a better place.
Maybe you can do the same thing to know who are giving merits to Subsections "help to choose merit sources."

Amazing analysis!
I personally always try to take part in discussions from almost every other section/subsection of the forum, but as you said, not all of us have enough knowledge to contribute useful opinions everywhere. Though some times I skip sections because of the excessive spam in there, as I don't want to get into the eyes of people as a spammer.

Your data analysis is amazing ddmr.

I am happy to see I participate in many boards. I didn't know that other users are used to visit only few boards.

There are many boards which I ignore completely, such as mining, reputation, scam accusations, gambling, etc...

Your post is a good incentive to try to visit other boards and change posting habits

Nice work mate. There are very few people on the forum who can play around with data like you. Fantastic effort.

PS: I am really happy to see me on the list which eventually proves that mdayonliner is not someone who talks about merit only LOL

Update 04/04/2021

Data is now accesible here (updated weekly):
https://public.tableau.com/views/BitcointalkMeritDashboard/CorrelationMatrix?:language=es&:display_count=y&publish=yes&:origin=viz_share_link

---

1.   Introduction.

I’ve updated this OP, which was elaborated in August 2018,.recalculating all the data, and additionally included the merit matrix by rank. All with data is now as of 20/12/2019.

In order to allow comparison between then and now, I’ve archieved the old version of the OP here: http://archive.ph/R2owd

---

A couple of days ago, there was a discussion going on related to whether people are being awarded sMerit solely on a specific section/topic, and how that seems unfair. While many of us would like to be Satoshi alike, the fact is that sMerit is meant to be awarded to any post we find contribute, be it on the topic it may be. It’s up to the awarder’s criteria.
Every individual has his own skillset (some strong, some weak at best), and we do not all converge on them, less of all on being highly knowledgeable on Bitcoin. In a company that produces shoes, only a subset of the employees will have any idea as to how to design them, but everyone knows how to wear them. The same applies here in terms of Bitcoin knowledge, and although it is improvable for all, only a specific set of people will really master it enough to be able to contribute with a depth charge of knowledge.

In fact, as can be seen on the Merit Dashboard (go to the Tab labelled "Section Subsection"), the Bitcoin aggregate section receives only around 10% of all the sMerit that has been awarded). There are multiple reasons behind this fact, but that is not the real objective of this post.

What the discussion got me to wonder is the following:

a)   How many people receive sMerit on how many forum subsection (i.e. how many are capable of being merited on more than one subsection, and who excels on multiple subsections).

b)   For a given forum subsection where people are merited, what other subsections correlate through those same people (a kind of correlation matrix).


2.   Forum members Merited on Multiple subsections.

The idea here is to see a summary of the number of subsections forum members (all ranks considered) have been merited on. By subsection I’m considering the main level of subsections I’ve used on previous occasions (Altcoin Discussion, Announcements (Altcoins), Marketplace (Altcoins), Mining (Altcoins), Speculation (Altcoins), Bitcoin Discussion, Bitcoin Technical Support, Development & Technical Discussion, Mining, Project Development, Deleted, Economics, Marketplace, Trading Discussion, Arabic, Chinese, Croatian, Dutch, French, German, Greek, Indian, Indonesian, Italian, Japanese, Korean, Other Languages, Philippines, Polish, Portuguese, Romanian, Russian, Skandinavisk, Spanish, Turkish, Archival, Beginners & Help, Meta, Off-topic, Politics & Society, Serious discussion). There are 41 subsections, 22 of which are local languages.

I’ve first created the distribution for all merited users, and then restricted it to those that have received >= 10 sMerits for comparison reasons. Posts that have been deleted cannot count in any section, but I’ve also included a count of those that have received sMerit exclusively on deleted posts, which is an area of interest on its own (excluded from the list below, but included in the merit matrix).



Things that standout:

a)   158 (27 last time) people have been merited on 10 or more subsections. That is a very difficult achievement, which implies a good knowledge in multiple areas as well as communication skills. This feat is in the hands of the following:

The list with the actual subsections where the above have been merited can be found here:
https://docs.google.com/spreadsheets/d/1CFMvMD4lDMnnkxWl9p02k0ZoJW044eRY4EGEqkYd-8I/edit?usp=sharing (second tab).
Each person has his own mix of subsections that contribute to his total received sMerit with different weight. I’ve included the column PMerit (percentage of Merit) to show this.I've now extended this information (on the Google Sheet) to all merited profiles, not just those listed here.


b)   The vast majority (73,53%) of people have only been awarded sMerit in one subsection, and 12,10% have in two subsections. The former amount is logical, due to the large amount of people that have been awarded a single merit.

c)   For those awarded >= 10 sMerits, 50,01% have been on a single subsection, and 20,74% on two.

d)   The total exclusively deleted is 1.801 people (5,62% of those merited), having 206 of those  received >= 10 sMerits. This set of users have gained every single sMerit on what are now deleted posts!

I guess a list in mandatory here: https://docs.google.com/spreadsheets/d/1CFMvMD4lDMnnkxWl9p02k0ZoJW044eRY4EGEqkYd-8I/edit?usp=sharing  (third tab)


3.   Subsection correlation matrix.

This has been a toughie. What I’ve done here is, for each subsection, determine the number of forum members that have been awarded sMerit for posts in the subsection. Then, I’ve analysed for each of them, in what other forum subsections they’ve also been merited in.
The idea behind is to see the correlation of sMerit between subsections, based on where the people are being merited. This should tell us the subsection affinity of the merited users. The less exclusive (see later on) a merited subsection is, the more we can see correlations to other subsections on the forum.

Since the number of subsections is large, I have grouped all the Local sections into one, thus simplifying by nearly half the matrix, and on the premise that correlation between subsections is very small. I’ve also brought “Deleted” to the table as a virtual subsection due to its weight.

The result is the following table (All ranks):



I’ve highlighted the top three percentages for each row (excluding the deleted column). I’ve also got the absolute value table, but the one displayed in percentages is better suited here.

Example of interpretation:

Altcoin Discussion has 3.870 merited people (12,08% of all merited profiles), 1.796 of which are exclusively merited just in this subsection and no other (46,41% of the 3.870 people).
Out of those merited here, 16,28% have also been merited on Bitcoin Discussion, 15,87% on a Local Board, and 14,86% in the Economics subsection. Only 0,90% have been merited in the Bitcoin Mining subsection. A fact aside is that 22,97% of these people have also been merited on a post that has now been deleted (or off-limits).

Table is best read by rows.

Things that are interesting:

a)   Column nMeritedUsers indicates how many distinct users have been merited in that subsection (not the total sMerit they received), and % Distinct Users the percentage of all merited users.  
It’s interested to see that 35,77% of merited users are merited on local boards, albeit with low sMerit values (seen on previous analysis). On the other end of the scale, Serious Discussion and Project Development have only 1,39% of all merited users merited there (archival is lower, but it’s not really a proper subsection).
 

b)   Column nExclusiveUsers (or deleted) indicates how many of those users have received their sMerit solely on a given Subsection. If the user has also got deleted posts, I do not count these here as a separate subsection, and thus consider that the sMerits are awarded exclusively to a single exclusive subsection.
 
Column %Subsection specifies the percentage of exclusive users in relation to all the merited users in the subsection.
77,90% of Local merited users are merited exclusively on local boards. Ann Altcoin is also a very closed circle (70,80%) and so are the mining subsections (64,53% mining altcoins and 61,13% mining in the Bitcoin section). On the other hand, subsections belonging to the Other section are less exclusive.


c)   16,37% of merited people have at least been merited on one post that has later been deleted. Out of those, nearly a third has been merited exclusively for what are now deleted posts, as seen before.

d)   Correlations that stand-out (summarized by Section):

-   Altcoin subsections have a strong correlation to local boards, as well as to the other Alternate Cryptocurrencies subsections.

-   Bitcoin subsections correlate more to Marketplace, Bitcoin Discussion and Meta.

-   Deleted virtual subsection correlates most to Local boards as well as Bitcoin and Altcoin discussions.

-   Economics section correlates best to Bitcoin Discussion, Marketplace, Trading and Meta.

-   Local Sections correlate more to Altcoin and Meta (although not much, since they are rather endogenic).

-   Other section components correlate best to Meta, Bitcoin Discussion, and Marketplace.



The merit matrix delimited to Legendary rank:




The merit matrix delimited to Hero rank:




The merit matrix delimited to Sr. Member rank:




The merit matrix delimited to Full Member rank:




The merit matrix delimited to Member rank:




The merit matrix delimited to Jr. Member rank:



Note: I've omitted the merit matrix for some standard ranks (Brand New, Newbies) and non-standard ranks (Administrators, etc.).