Topic: 2 MILLION users and their STATS (Read 846 times)

1) Right now unfortunately this is not possible since the scraper is scrapeing data which only is shown in user profiles. Information about user posts would be very useful, it will eneable  us to see many interesting things.
2) I am in process of scraping 2mil+ users . This thread shows that it would be useful to show activty of lower ranks also to see the difference between lower rank activty and higher ranks and intend to do so.

On lower ranks there would obviously be a lot of inactive members, aside from lower ranks who gave up on receiving merits and got banned for plagiarism some members actually here just created an account to ask one question and it is most likely their only reason why they created (you need to have an account in order to create a thread) their account after it is solve or the question has been properly entertained he won't be using the acvount anymore. Inactive account are really not a problem if that is what you are concerned as they do no harm in the forum.

I would be interested in the following breakdown:
1.) Active users who posted only in Altcoin discussion
2.) rest of the active users

And sorted by rank, from newbie to legendary...
Do you think I would be surprised by the result?

Yeah, I agree but this is a good thing in the end We're getting rid of the people who aren't interested in crypto.

I was trying to say that I was expecting more users to be lurking here (even if they don't post) but it seems the community isn't as big as I originally thought.

 Moving forward I think that I am going to show " deleted/skipped" accounts a bit differently but they are part of statistics and should be noted.



As I already mentioned before , i dont think that defining  active users as someone who has last logged in some time period is a correct estiamte. There are a huge amount of " Brand new" users relative to Jr. and above activty, who have logged in recently and possibly this is bot activty.

The ideal definiton would be user who has posted at least once in some time period. But current scraper version which is available does not have this kind of option,since it can scrape only information which is seen on profile page of any user.

Can we consider active user as someone who has just logged on within the last 90 days but not posted anything? Can we get a clear interpretation for the active user from "theymos" or why can't we all agree on one definition? because it will surely help for the members who actually scraped user data for their analysis.

btw this is what I got from DdmrDdmr for an "active user"

I'd suggest just removing the Deleted category. They are a combination of skipped user IDs when something went wrong in the signup process and accounts that were banned and deleted in the very early days of the forum before nuking became a thing.



The number seems to be highly inflated by brand new accounts that never post. Is it possible not to count an account as active until it has posted at least once?

Thank you for the breakdown. These are insightful metrics. The elephant in the room to this data are the alt accounts. Do you think there is a way to determine success earning cryptocurrency utilizing data from this set? Knowing these metrics, is there a dollar for dollar difference that can be quantified in the overall success of these rank statistics over time. Bc we know this many projects exist, and this many possible users can achieve this pay, we know there is this much stable work in the near future with this incoming growth. I want to know how safe my job is from this data.

I dont think that this data is proof of multiple account creation.
Also as it has been pointed out on several ocasions, it is reasonable to believe that good part of new users come here to discuss some details or ask some questions and dont have nor the ambition nor need to rank up. For example if they ask some questions regarding minig.


I used Pythons script , it is available on GitHub.
Even tough there is some collected data available online, there is no service and no data source publicly available to this extent.

The stats to a large degree reflects the true position of the activity on bitcointalk forum. Yes, there could be some variance but the reason for such variance are known and can be adjusted to suit the needed clarification.

This is a great effort. Keep up.

It clearly shows and can be also proof the massive creation of multiple accounts and making it until it reaches Jr. Member.

Some of it looses interest in the crypto and find an alternative way to earn income like faucets, captchas and trading. I can surely say that 75% of the Jr. Members are just informed by someone because they can earn income. The members, I think 40-50% of them manage to gain merits and continue their interest of crypto because they want to learn more about it or some of them are investors.

All of the percentage rate are just estimated

I count the most of users loose the interest in crypto currencies after they registered on the forum. I think the point here is that due to merit system they can't rank up so easily as before, so people just do not wish to bother with low ranked accounts heee, cause of it doesn't bring any profit.

What service did you use to collect data? How to see the percentage of ranks of the forum participants from the total number and their activity?

Its exactly why statistics are misleading. I've had more than 90 days from the forum in the past if I recall correctly. Actually, a lot longer than that. So it will naturally fluctuate, and will likely yield in different results throughout the year.

Like the other users have mentioned too...the damn bots.

I guess its because of the merit system, after this system introduce many cheaters and spammer's lose their hope to rank up and there are lots of users get negative trust because of merit abusing most of the users with negative trust quit the forum and I think some of them now are focusing on 1 account.

Im not shocked that the active users count now is lower than before because before the merit system introduce to the community there are users with 20 account's and higher that only owned by 1 you can see it on the topic created here in meta https://bitcointalksearch.org/topic/list-suspected-users-that-are-abusing-merit-30-2896910. Im sure after getting negative trust on their accounts they are not interested anymore to that account and abandoned it.


This is the benefit of the merit system in the forum we can now see the real count of legit users and the quality of post of other members have improved.
______________________________________________________________________________

Another view in this that affects the decreasing count of real active  forum member that I think is because of the bad market situation, many investors joined here last year with high expectations on bitcoin and some of them are now quitting the crypto world because they are seeing the market falling down especially those investors who just invest without knowledge about cryptocurrency.

Same way, last logged in ( Last activity) within some time period. Technically there are ~450k users who have logged in last 90 or so days.
But I think that there is reason to believe that good amount of these logins could be work of bots.

For example, if we use these parameters:
1) Rank: Brand New
2) Last activity within 30 days
3) Registered before year 2017
 
There are ~60k users which meet this criteria. Moreover when I scrolled fast forward, there was significant amount who were registered in 2014.
Do we believe that users who have registered more than a year or several years ago, have not made any posts and just log in every year or so ?

How do you define an active user?  I define them as someone who has logged on within the last 90 days.

There can be a some reasons of their inactivity:
1. Some users register here to create thread in scam accusation, once their problem gets solved, they forget this forum.
2. It's almost same here too, some users register to ask things about mining, hardwares or some technical things, then they get lost too.
3. Sometimes people forget password of their account when they visit forum once in a month or when they need.
4. Spam and spam...

Personally I was around this forum for year or two, was reading other threads/post, decided to ragister later. I want to say that number of members doesn't matters, guests are people who have interest in crypto usually.

You are right. Thanks for pointing that out.


Yes, i am aware of Vods great work.
But there is some important differences between this information and Vod's website. Vod's webiste shows that there are ~450k active users. If you compare it to number of last active users this year starting from Jr.Member, it is a difference of ~360k users. These are Newbies and Brand News and I dont think that most of these users are contrubuting to forum and I have good reason to doubt nature of their activty, thus i did not include them.

I chose to interpret missing number of accounts as  "deleted" and  the number of  " deleted" accounts are similar to Vods .  I agree that these accounts could include skipped UIDs.

A few things:

There are 3 Global Mods, not 4.

Most of the brand new accounts are the nuked accounts.

The 16,000 deleted accounts might also include the UIDs that were skipped. Ref: https://bitcointalksearch.org/topic/m.33862543

If you didn't know already, you could use Vod's website, it has a lot of data that may or may not help you: https://bpip.org/

I want to clarify some things. I analysed users from ID=1 to ID=2'000'000 the 2 million + users are not included . I have to note that there is a significant " Newbie" and " brand new" activity ( judging from "Last activty" ) .Since most of " Brand New" users have no posts, I doubt that the activty from these accounts have human nature but more likely they are bots and I decided to not include them this time.


First things, first. Huge portion of 2mil users are " Brand New" , ~1,4 million and they are not contrubuting to this forum, when this is taken in account, it is more easy to understand why the activty is so low in respect to user id count.  I would recommend to look at this thread: https://bitcointalksearch.org/topic/m.42983972


I recently restarted the scraper. I am planning to scrape all users this time. i will defintely show the  2 million + users atats also in near future.


And Thanks , I appreciate your feedback.  

Nice stats and great work


Guess 300 K of the Users are Bots or something like that way of the count!  Would be nice to see or know this stats too!

These stats look nice Thanks for taking the time to summarize them.

It's interesting to see that out of 2.3 MILLION users, the active users in the last month is roughly 65k. I was expecting more users to be active in the last month.

Not to mention that the users active in the last YEAR is under 100k... Not sure about you, but I was expecting at least 700k users to be active in the last year.

I wonder what is happening to them... Are they losing interest in crypto in general due to the volatility or are they moving on to other forums/social media groups?

What do you think?

Hello! At the moment of writing this post, there technically are 2333176 users registered from which I have scraped 2 million users. The 2 millionth user was registered on 2018/04/03. As most of us suspect or know the biggest part of these users are with lower ranks, I have looked at lower ranks more extensively in my last post  thus in this post I wanted to focus more on Jr. Members and above. My main goal with this post is to show distribution of first 2 million users , give an insight to user activity level while keeping it simple as possible.

Since PIE charts are the most common and one of the most intuitive charts to read, here are the distribution of bitcointalk ID's from 1 to 2'000'000 .



In addition I looked at the activity of rank Jr. Member and above. This time I characterised the activity by "Last activity" in certain time period . One plot for users who's last activity was last month and one plot for those who has been last active this year including last month.


Between these ranks last activity within last month was seen for 61836 users.

At this plot we can see that most ranks last activity within this year was in range from 70 to 80% . But Legendary and Copper members are above 80%, this is not an surprise since
1) Users with these ranks are in relatively small numbers
2) Legendary users are the biggest crypto enthusiasts
3) Copper members are relatively new rank type

If we sum up all the user accounts with last activity within this year from Jr.Member to Legendary + Copper members , we get 90993 users. In my mind this is a interesting result because most of us could easily think that there are much more active users in this forum since there are 2million + registered users.

This analysis was made from data which was scraped in start of July this year. I believe that the " last activity:last month" plot wont change much if we had data which would have been scraped more recently. Possibly we would see a bit more users in the " last activity: this year" plot but overall I think that this is a good representation of the situation.

In case someone wants to play with this data, here it is in more accessible manner:

Rank	Full N    Act.last m. N       %    Act.this year N     %
Jr.Members	60389        32035           53.05     47025         77.87
Members	27621        11772           42.62     18424         66.70
Full M.	18170        8959             49.31     12704         69.92
Sr.M.	8999          4441             49.35     6384          70.94
Hero	4140          2252             54.40     3139          75.82
Legendary	2144          1519             70.85     1892          88.25
Copper M.	1441          858               59.54     1425          98.89
Deleted	16000
Donator	109
Admin	2
Staff	26
VIP	37
Global moderator	3

P.s. If you have any ideas or suggestions what you would like to see, i will be happy to hear your ideas.