Bitcoin Forum
October 20, 2018, 10:45:23 PM *
News: Make sure you are not using versions of Bitcoin Core other than 0.17.0 [Torrent], 0.16.3, 0.15.2, or 0.14.3. More info.
 
   Home   Help Search Donate Login Register  
Pages: [1]
  Print  
Author Topic: Analysis - Ranked-up Users – Section/Subsection profile (& Local Boards)  (Read 215 times)
DdmrDdmr
Sr. Member
****
Offline Offline

Activity: 280
Merit: 1120

There are lies, damned lies and statistics. MTwain


View Profile
September 06, 2018, 05:46:12 PM
Merited by Welsh (5), dbshck (3), malevolent (1), Piggy (1), finaleshot2016 (1)
 #1

1.   Introduction.

On the Merit Dashboard there’s a section where we can see the current (as of last Friday)  2.561 total users that have ranked-up, needing merit as a requirement. I recently looked-up how many were from my local board (a measly 22), and thought it may be interesting to see a summary for the other local boards too. Since that would only be a subset, I went ahead and did the exercise for the whole 2.561 users to see if there was any interesting information that could be derived, and there is ….

Note: I am considering here all the current merit received for the above set of users up to date, not just the exact amount needed just to reach their current rank.


2.   Ranked-up users.

The ranked-up user summary is as follows as can be seen on the Merit Dashboard (see Tab labelled Ranked-up):



The detail of the usernames behind each of the above numbers can be seen on the Dashboard, so there is no need to replicate it here.

I will be treating the ranked-up users as a whole in the following sections, without separating by rank. This is due to the fact that 93% of ranked-up users are those that have reached the Member rank, and only 7% have reached higher ranks.  Replicating the below information tables by rank may make the post too long, although if required I can add it later on.


3.   Number of Merited Subsections

The following is a distribution of the number of Subsections a ranked-up forum member has been merited in:


It turns out that 58,65% of ranked-up users have been merited in a single subsection (be it a local board or otherwise as we’ll see further down), 28,50% have done so on 2 subsections, 12,57% on 3 subsections, and 10,27% on 4 subsections or above.


4.   Subsections in which ranked-up members have been merited in.

This is the most interesting table really, and we can derive some interesting insights based on its content.


Columns:
Section: Forum Section.
Subsection: Forum subsection.
nUsers: Number of ranked-up users that have been merited in the subsection. The total of this column is more than the number of ranked-up members, since some are merited on multiple subsections.
%RankedUp: % of ranked-up users that have been merited in the subsection.
SingleSS: Number of ranked-up users that have been merited only in the subsection (and nowhere else).
%SingleSS: % of ranked-up users that have been merited only in the subsection.
nReceivedMerit: Total Received sMerit by the ranked-up users in the subsection.
avgMerit: Average merit per ranked-up user in the subsection.
stdDevP: Standard deviation associated to the avgMerit.

Example of interpretation:
471 ranked-up users have received sMerit in the Altcoin Discussion subsection. That is 18,39% of all the ranked-up users. Singularly, 77 of these users have ranked-up solely based on the sMerit received in the Altcoin Discussion subsection (that is 16,35% of the 471 users ranked-up merited in this subsection).
On aggregate, the 471 ranked-up users have received 3.504 sMerits in the Altcoin Discussion subsection. That is an average of 7,44 sMerits per user (with a high standard deviation of 10,96 sMerits).

What stands out:
a)   30,89% of ranked-up users have received at least 1 sMerit on a post that has later been deleted. That is a rather high percentage, that makes it more difficult to track merit abuse unfortunately (even though it is not moderated). Nevertheless, only 9,73% of these ranked-up receive all their sMerit on deleted posts (77 people).

b)   The subsections most related to ranked-up members are Russian (28,66%), Altcoin Discussion (18,39%), Ann Altcoins (15,03%) and Bitcoin Discussion (12,65%). Meta is related to 8,98% of ranked-up users, so although meta is a rather well merited section, it is not to date a key section to ranking-up.

c)   Subsections such as Mining (Altcoins) with 2,11% influence, Bitcoin Technical Support (0,86%), Mining (0,74%) and Project Development (0,55%) are the least influential in the overall figure of ranked-up people.

That says something: the most Bitcointalk/technological sections are the sections are the least influential for ranking-up.
That does not mean you cannot do so, don’t read this wrong, but rather that the profile of people on Bitcointalk that have so far ranked-up with the Merit System is not too technological.

d)   Some subsections have a larger degree of people that have received sMerit solely on that subsection: Announcement Altcoins (33,77%) or all those that received sMerit in that subsection, Mining (29,63%) of those ranked-up members merited in the subsection, and Mining Bitcoin (26,32%).
Although the user base is not really too large to draw solid conclusions, it tells us that mining profiles are rather specific and have a rather large fidelity to their subsection, and similarly in Altcoins (although in Altcoins we’ve all seen shady stuff going on and that may well be the reason).
e)   I would not pay too much attention to the AvgMerit really, since the standard deviations are rather high in most cases. This means that ranked-up people awarded merit in a given subsection do so in a widespread fan of sMerit amounts,  differing quite a bit in many cases from the average.

f)   I’ve deliberately left the local boards to be treated last but not least. Roughly, a bit more than half of the ranked-up forum members are related (through merit) to a local board. That I figure is good, and it means that an international forum such as Bitcointalk has contributions from all around and, consequently, people from different countries are ranking-up (slowly).

Nevertheless, if we look at this in detail, there are some locals that numerically count with a large share of ranked-up local members (I’m talking about absolute ranked-up numbers here, not in comparison to the population on each local board). Russian local boards stands out with 734 ranked-up users, followed by the Turkish Local Board (215), the Indonesian (160), and the Chinese (122). They are surely the largest local boards, so the higher quantity of local ranked-up members are theirs (even if in reality the figures are probably rather small, compared to the number of the distinct local posters there).

There’s a whole array of mid/low-level local boards (Croatian, Spanish, Italian, French, German, etc.), and then there’s the nearly impossible boards (Arabic, Dutch, Greek, Indian, Korean, Polish, Romanian and Scandinavian). Again, we do not know to total distinct posters on these boards to compare and see the ratio, but the absolute number is measly and either indicates a grave merit problem, or a potential to boost Bitcointalk in those areas (as in the above mid/low ranked-up level countries really too).

Even between countries, there are large “fidelity” differences. For example, out of the 122 local Chinese that have ranked-up, 88,52% of them have done so being merited only on their local board. The Turkish board follows a similar patter, while the Russian  local board has more than half of its ranked-up members that have been awarded sMerit on other boards too.

The lower the percentage (column %SingleSS), the more that the local ranked-up members have emigrated to the English boards in order to be merited for their contribution there. A certain command of English is required there, so it is logically related to the educational system in each country in terms of learning the English language.


Notes:
- I’ve assumed that people on local boards are from the local board if they have been merited there at least once. That is true for the most, although there are a few exceptions (19 to be precise) that have been merited on more than one local board. The most extreme case is Leteravian, who has been merited on 7 local boards (often by translating himself his posts into multiple languages).

- Bear in mind that I’ve analysed all the merit for the ranked-up members, not just the merit that took them to their next rank. The latter is a much more tedious task, and the results would probably not differ too much from the above, since most users that have ranked-up are Members with not too much merit above the required baseline.

1540075523
Hero Member
*
Offline Offline

Posts: 1540075523

View Profile Personal Message (Offline)

Ignore
1540075523
Reply with quote  #2

1540075523
Report to moderator
1540075523
Hero Member
*
Offline Offline

Posts: 1540075523

View Profile Personal Message (Offline)

Ignore
1540075523
Reply with quote  #2

1540075523
Report to moderator
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction. Advertise here.
1540075523
Hero Member
*
Offline Offline

Posts: 1540075523

View Profile Personal Message (Offline)

Ignore
1540075523
Reply with quote  #2

1540075523
Report to moderator
1540075523
Hero Member
*
Offline Offline

Posts: 1540075523

View Profile Personal Message (Offline)

Ignore
1540075523
Reply with quote  #2

1540075523
Report to moderator
1540075523
Hero Member
*
Offline Offline

Posts: 1540075523

View Profile Personal Message (Offline)

Ignore
1540075523
Reply with quote  #2

1540075523
Report to moderator
RivAngE
Member
**
Offline Offline

Activity: 294
Merit: 81


View Profile
September 12, 2018, 07:26:35 AM
 #2

A great analysis as always!
There's one thing I noticed that was missing from your post though: You don't mention which period of time is taken into account in the above numbers, only that it's until last Friday.

I guess the calculations are for the period since the merit system was introduced, but (A) it's not clearly stated and (B) I don't remember which month was the merit system introduced!
I apologise if I missed something.

Welsh
Staff
Legendary
*
Offline Offline

Activity: 1386
Merit: 1374


Away on expedition until 5th of Nov


View Profile
September 12, 2018, 07:47:42 AM
 #3

31% (rounded up) of users have received 1 merit on a deleted post? I don't think that's too bad honestly. Do you have any idea how many of these deleted replies were in self moderated threads? I think that could be an interesting stat. Also, considering the amount of merit which gets rewarded in Altcoin announcement threads this could be exaggerated, and doesn't really mean there's much abuse at all. A few replies in altcoin announcement threads get merit. Sometimes when calling out the project, and later on the thread might be trashed because they were offering an incentive to post. It quite often happens with the scam projects.

DdmrDdmr
Sr. Member
****
Offline Offline

Activity: 280
Merit: 1120

There are lies, damned lies and statistics. MTwain


View Profile
September 12, 2018, 08:23:24 AM
Merited by Jet Cash (2)
 #4

<...>
There's one thing I noticed that was missing from your post though: You don't mention which period of time is taken into account in the above numbers, only that it's until last Friday.
<…>
 
Yes, you are right: the period of time involved is from the Merit System Kick-off, until last Friday prior to OP date (specifically: 2018-01-24 22:12:21 .. 2018-08-31 02:41:54 forum time). Since calculations are limited to those that needed merit to rank-up, the merit considered is all-time merit.

31% (rounded up) of users have received 1 merit on a deleted post? I don't think that's too bad honestly. Do you have any idea how many of these deleted replies were in self moderated threads? I think that could be an interesting stat. <…>
It’s really not that hard to get merit a post that is later deleted or to be merited in such (I got merited at some point for a post that was later deleted), although logically probability will be higher for the regular user when he get merited often. Being merited on later deleted posts, where the number of merits is low or the number of posts is low could sometimes be a cover-up.
Out of the 791 cases, 705 belong to a Member rank, and out of those 98 have received 9,10 or 11 merits on later deleted posts. Some are bound to be cover-ups, but picking cases apart when the posts are deleted and pursue capabilities are limited is perhaps not very fruitful.

I did take a deeper look at the deleted merited posts back in may (see Analysis of Deleted Posts with sMerit). The focus there is not on those that have ranked-up, but rather overall.

I don’t think there is a way to distinguish self-moderated threads when scraping the data. Normally I only scrape as far as the title, without entering the specific content (I did do it once for an analysis on post length, quotes and so on, but it was rather cumbersome), so I figure I cannot really easily see any sort of mark that would indicate the thread to be a self-moderating one with the data I retrieve.

Welsh
Staff
Legendary
*
Offline Offline

Activity: 1386
Merit: 1374


Away on expedition until 5th of Nov


View Profile
September 12, 2018, 12:27:10 PM
 #5

Sure, I've probably had posts that were merited and, are no longer available because the thread was trashed or something. I have noticed a few threads where someone has asked a fairly routine question, someone answered, and was merited for it, and then the post gets removed. I guess my original point was I don't think there's too much merit abuse going on with deleted replies. I was expecting a much higher percentage.

vlad230
Sr. Member
****
Offline Offline

Activity: 308
Merit: 274



View Profile
September 12, 2018, 02:08:55 PM
Merited by Zapo (1)
 #6

Yes, you are right: the period of time involved is from the Merit System Kick-off, until last Friday prior to OP date (specifically: 2018-01-24 22:12:21 .. 2018-08-31 02:41:54 forum time). Since calculations are limited to those that needed merit to rank-up, the merit considered is all-time merit.
[...]
I think I found a bug in your app.

Steps to reproduce:
1. Select the start date as the start of the merit system 1/24/2018 (or close to it)
2.  Check the people who ranked up to Full Member and look for my user

Expected:
I am listed as one of the users who ranked up to Full Member.

Actual results:
I am not listed there, although I was a Jr. Member when the system was implemented.
There may be other users that are missing from this list.

P.S. Any chance you could make the font a bit larger or thicker? It's a nice tool but a bit hard to read.

Check the image below:
DdmrDdmr
Sr. Member
****
Offline Offline

Activity: 280
Merit: 1120

There are lies, damned lies and statistics. MTwain


View Profile
September 12, 2018, 03:15:10 PM
 #7

<...> I think I found a bug in your app.<...>
It’s really down to a misinterpretation: on the Ranked-up tab, the date slider (filter) is relative to ‘Date Created’, meaning the date the account was created. It does not act as a filter on the actual Merit TXs as do other Tabs on the Dashboard. Since your account was created on the 31/10/2017, by placing the slider on the 24/01/2018 you are effectively leaving your account out … (I was debugging it now using my profile and thought there was something wrong, but then I just realized I seem to have ranked-up to Sr. Member yesterday..).
 
Quote
P.S. Any chance you could make the font a bit larger or thicker? It's a nice tool but a bit hard to read.
I have it ready to adapt to the user’s screen resolution (forget phones though), so theoretically it does it’s best to adapt to any screen resolution. I’ve tried it on a couple of devices and it adapts well, although they are 1920x1080 and up and I tend to user 150%++ scaling on high definition screens (otherwise everything is too small, not just the Dashboard). Try zooming the browser to see if that helps.

DdmrDdmr
Sr. Member
****
Offline Offline

Activity: 280
Merit: 1120

There are lies, damned lies and statistics. MTwain


View Profile
September 12, 2018, 05:17:40 PM
 #8

Sure, I've probably had posts that were merited and, are no longer available because the thread was trashed or something. I have noticed a few threads where someone has asked a fairly routine question, someone answered, and was merited for it, and then the post gets removed. I guess my original point was I don't think there's too much merit abuse going on with deleted replies. I was expecting a much higher percentage.
It’s not trivial to detect cases of merit abuse, but I would say that it occurs more frequently on relation to the 10 merits required to become a Member, since the number of merits implied is low and easier to “obtain” by non-standard procedures. I would therefore focus on the overall picture of those that have received 10 merits being a Jr. Member (in anticipation) or a Member.

Out of all the 19.268 people that have received at least 1 sMerit (non-airdropped), the following are where I’d say we have more probability of such cases:
Code:
               rank           merit          nMeritReceived nTX            nInitialMerit  nUsers
               Jr. Member     10             10             1              0              109
               Jr. Member     10             10             2              0              16
               Jr. Member     10             10             3              0              13
               Jr. Member     10             10             4              0              9
               Jr. Member     10             10             5              0              26
               Jr. Member     10             10             6              0              5
               Jr. Member     10             10             7              0              2
               Jr. Member     10             10             8              0              6
               Jr. Member     10             10             9              0              1
               Member         10             10             1              0              249
               Member         10             10             2              0              293
               Member         10             10             3              0              209
               Member         10             10             4              0              184
               Member         10             10             5              0              161
               Member         10             10             6              0              146
               Member         10             10             7              0              99
               Member         10             10             8              0              71
               Member         10             10             9              0              50
               Member         10             10             10             0              32

Obviously many of the above cases are legit, but this is where I’d place my bet to find the largest pool of merit abuse cases, especially in those cases where they are stuck on 10 gained Merits in one or two TXs.
That gives me 667 users with a conceptually higher probability of having abused their merit (3,46% of total merited users), out of a potential of 1.681 users that have received their 10 merits in 1 to 10 TXs (8,72% of total merited users). The question is how many of those really did abuse the Merit System and how many didn’t, but the above is the pool with higher conceptual chance I'd say.

vlad230
Sr. Member
****
Offline Offline

Activity: 308
Merit: 274



View Profile
September 13, 2018, 09:28:17 AM
 #9

It’s really down to a misinterpretation: on the Ranked-up tab, the date slider (filter) is relative to ‘Date Created’, meaning the date the account was created. It does not act as a filter on the actual Merit TXs as do other Tabs on the Dashboard. Since your account was created on the 31/10/2017, by placing the slider on the 24/01/2018 you are effectively leaving your account out … (I was debugging it now using my profile and thought there was something wrong, but then I just realized I seem to have ranked-up to Sr. Member yesterday..).
Ok, that makes more sense now.
I originally wanted to look up accounts that have ranked up under the current merit system, not since the beginning of the forum. I doesn't really make sense to have it like that if you ask me - no offense.

If you set the date from the beginning of the forum you'll have a lot of misleading information - making you think that a lot of people have ranked up in the current merit system when in fact that's not really true.

I have it ready to adapt to the user’s screen resolution (forget phones though), so theoretically it does it’s best to adapt to any screen resolution. I’ve tried it on a couple of devices and it adapts well, although they are 1920x1080 and up and I tend to user 150%++ scaling on high definition screens (otherwise everything is too small, not just the Dashboard). Try zooming the browser to see if that helps.
I'm using 1920x1080 on a 24" monitor and I think this could be improved and not requiring people to zoom in.

But maybe that's just a bad professional habit, working in the industry for too many years now. Smiley
DdmrDdmr
Sr. Member
****
Offline Offline

Activity: 280
Merit: 1120

There are lies, damned lies and statistics. MTwain


View Profile
September 13, 2018, 09:46:23 AM
 #10

<…>
I originally wanted to look up accounts that have ranked up under the current merit system, not since the beginning of the forum. I doesn't really make sense to have it like that if you ask me - no offense.

If you set the date from the beginning of the forum you'll have a lot of misleading information - making you think that a lot of people have ranked up in the current merit system when in fact that's not really true.
<...>
No, we are still not on the same page here. The forum members displayed as having ranked-up have done so requiring merit to do so. I do not show members that have not needed merit to rank-up. You can see this by looking at the detail and observing that merit is the required amount (or above) for each new gained rank, whilst initial merit is generally 0 (Initial rank of Member or above is not cero, since they had default initial merit).
Therefore, all now 2.608 forum members shown on the tab have ranked-up through merit, regardless of when the account was created.

The Date Created slider filter is useful to delimit precisely that. Say you wanted to see what new accounts created in 2018 have achieved, you would move the slider accordingly to focus on this subset.

vlad230
Sr. Member
****
Offline Offline

Activity: 308
Merit: 274



View Profile
September 13, 2018, 10:11:19 AM
 #11

No, we are still not on the same page here. The forum members displayed as having ranked-up have done so requiring merit to do so. I do not show members that have not needed merit to rank-up. You can see this by looking at the detail and observing that merit is the required amount (or above) for each new gained rank, whilst initial merit is generally 0 (Initial rank of Member or above is not cero, since they had default initial merit).
Therefore, all now 2.608 forum members shown on the tab have ranked-up through merit, regardless of when the account was created.

The Date Created slider filter is useful to delimit precisely that. Say you wanted to see what new accounts created in 2018 have achieved, you would move the slider accordingly to focus on this subset.

Ok, I get what you mean now. Then would you agree that the date picker would need to start when the merit system was implemented (Jan 24th, 2018) and not from the beginning of the forum?

Otherwise it's a bit misleading or something does not really add up.
DdmrDdmr
Sr. Member
****
Offline Offline

Activity: 280
Merit: 1120

There are lies, damned lies and statistics. MTwain


View Profile
September 13, 2018, 10:27:25 AM
 #12

Ok, I get what you mean now. Then would you agree that the date picker would need to start when the merit system was implemented (Jan 24th, 2018) and not from the beginning of the forum?

Otherwise it's a bit misleading or something does not really add up.
Date picker is fine being able to cover all time period. For example, @Jet Cash is a late 2015 account, and @hilariousetc a late 2014 account. Both are therefore prior to the Merit System kick-off, but both have needed merit in order to move from Hero to Legendary. From a global perspective, these two forum members have needed sMerits to rank-up (500 each), so it is correct to say that they have ranked-up under the merit system; specifically, from Hero (pre-merit rank) to Legendary (needing merit rank).

Now if you want to focus only on people that have ranked-up after the Merit System was initiated, and that created their account at the time, then you can move the Date Created slider to focus specifically on those created from the 24th of January onwards.

Both focused are valid, it depends on where you set the focus (thus the Date Created slider).

Pages: [1]
  Print  
 
Jump to:  

Sponsored by , a Bitcoin-accepting VPN.
Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!