Steem: Claimed Account Activity Analysis

in #utopian-io5 years ago

Repository

https://github.com/steemit/steem


Contents

  • Introduction
  • Scope
  • An overview and comparison
  • Activity by Content and Up-votes
  • Activity by Posting Key
  • Conclusion / Summary
  • Tools and Scripts
  • Relevant Links and Resources



source


Introduction

This analysis is a follow up to Steem: Account Claims Analysis. Background information with regards to Claimed Accounts and their purpose can be found on this post.

At the time of writing, the Resource Credit cost of claiming an account is 9,289,019,776,360 RCs. This number was obtained by the output a failed account claim (due to lack of RC 'mana') using Steem Connect's 'claim account' web-page.

Using https://steemd.com/@abh12345 as a guide with a 'max_mana' (RCs) of 34,615,566,006,823 and the current RC cost of claiming an account, it is estimated that that a Steem account will require to hold approximately 4,640 Active Steem Power (And have 100% 'mana'/RCs available) to create a pending claimed account.

34,615,566,006,823 / 9,289,019,776,360 = 3.726503639804659

17,292.94 Steem Power / 3.726503639804659 = 4,640 SP

The aim of this analysis is to:

  • Review current claim and creation numbers and compare with the previous work two months ago
  • Take a look at some of the activities the claimed accounts are undertaking

Scope

The data for this analysis was gathered on the 15th March 2019 and unless otherwise stated, the data processed covers the period from the 25th September 2018 to the 14th March 2019.

An account created via a claim 'token' is considered to be one which has a 0.0000 fee and 0.0000 delegation attached.


An overview and comparison

The very first account claim was made on the 25th September 2018 by 'blocktrades'.

From this date up until 14th March 2019, 332,751 accounts have been claimed and this number is shared between 761 claimants.

The previous analysis covered 113 days from the 25th September 2018 to the 16th January 2019 and in that time there were 221,935 claims shared between 566 claimants.

The table below shows the overall, previous, and most recent period - which is the 56 days between the first and 2nd analysis.



We can see that the daily average number of claims is rising slowly as time progresses, but rising much faster is the average number of claim 'tokens' that are being used to create an account. This could be due to the increasing number of applications and games available, and there is further analysis here later in the report.

The next chart shows the number of accounts created from a claim 'token', grouped into weeks.



The average number of accounts created prior to 2019 on any given week is lower than 2019, but there are three busy weeks in 2018, and one in 2019. Looking closer at the busy period in early January 2019, we can see that a total of 8628 accounts were created via a claim token in the 4 days covering the 8th - 11th January, and 4087 of those were on the 9th January.



A quick check on this date alone shows that, as probably expected by some, 'steem' is the creator of the vast majority of these accounts via a claim token - 3992 of the 3087. Since the 9th of January, no more than 2000 accounts have been created using this method in a single day, and perhaps any backlogs of accounts requiring creation have been cleared.

From the 25th September 2018 to the 14th March 2019, 89,602 accounts have been created via a claim token. 'steem' have created the vast majority of these accounts, 76,537 in total.


Activity by Content and Up-votes

In this section, a look at the activity of accounts created via a claim token, grouped by the account that created them, as far as Posts, Comments, Up-votes, and Down-votes.

The table is made up of the top 20 claimed account creators and a selection of additional creators which are related to applications. .e.g. 'appics'.



As far as overall activity based on the criteria above, the accounts created by 'blocktrades' via claims show up well in all areas. Only 'actifit' average more posts per account created, but produce less than half the comments than the 'blocktrades' created accounts, on average.

The accounts created by 'fundition' show easily the highest number of comments, with almost 70 on average per account created. 'partiko', surprisingly for me, rank 5th on the average number of comments produced by an account created via a claim.

Potentially, the accounts mentioned above are the creators of the accounts producing content, as apposed to some of the creators with low averages for this criteria, whose accounts could be more focused on games, or curation.

'blocktrades' and 'actifit' also average highly on the voting numbers, and 'actifit' likely score well in this regard due to a curation initiative they have in place for application users:

...visiting the Actifitter's activity page on actifit.io, upvoting a minimum of 3 actifitter's reports with 20%, will earn you 3 AFIT tokens, daily! source


Activity by Posting Key

Using the table above for popular creators tied to applications, and scan through the Posting 'Auths' of various accounts to build a list of commonly used applications, the following table looks at the number of claimed accounts that are using these applications, and what they make of the total percentage.

These numbers are based on the 'posting' field in the 'accounts' of the SteemSQL.



'dtube' have by far the largest number of entries in the posting field, with 'busy', and 'steemauto' - the automation tool in 3rd.

It is 'drugwars' that has the largest percentage of posting auths attached to claimed accounts, and perhaps surprisingly, 'fundition' is a clear second. 'partiko' and 'peakmonsters' also show reasonably high percentages - about a quarter of accounts with posting auths for those applications tied to accounts created via claims.

The data here is far from conclusive as 'drugwars', for example, is one of the newer applications and so perhaps more likely to have a higher percentage of claimed accounts linked via posting authorization.

And considering the number of Steem accounts in existence, the overall numbers attached to these 'popular' applications does not look that great in my opinion. I would have expected to see higher figures all round.


Conclusion / Summary

Account creations, excluding those created by 'steem', started out relatively steady following the hardfork at the end of September 2018, but since the turn of the year have picked up noticeably.

Accounts created through 'blocktrades' show up best as far as content activity and voting, and it is 'drugwars' leading the way as far as claimed accounts registered via posting authorization playing the game.


This analysis didn't quite go the way I had planned it, and spanned over 3 days. The amount of data that could be presented on the topic is vast, and this is only a relatively small sample. Earnings grouped by account creator, and Vests owned / delegated to claimed accounts summed/averaged by creator could also lead to some interesting findings - perhaps it's best not to know :)




Tools and Scripts

This analysis was produced by gathering data from SteemSQL, a copy of the Steem blockchain data held in a SQL Server database which is managed by @arcange. The scripts were written in Linqpad 5 and the charts produced using MS Excel.



-- Total claims in Scope of analysis
select count(*) from TxAccountClaims where timestamp between '09/25/2018 00:00:00' and '03/14/2019 00:00:00'

-- Recent period
select count(*) from TxAccountClaims where timestamp between '01/16/2019 00:00:00' and '03/14/2019 00:00:00'


-- Accounts (all in scope) Created from Pending Claimed Account
select fee, delegation, creator, new_account_name, timestamp from TxAccountCreates
where fee = 0.0000 and delegation = 0.0000
and timestamp < '03/14/2019 00:00:00'
and timestamp > '09/25/2018 00:00:00'
order by timestamp desc


-- Accounts (first period) Created from Pending Claimed Account
select fee, delegation, creator, new_account_name, timestamp from TxAccountCreates
where fee = 0.0000 and delegation = 0.0000
and timestamp < '01/16/2019 00:00:00'
and timestamp > '09/25/2018 00:00:00'
order by timestamp desc

-- Week Creates
select cast(datepart(ww,convert(date,[timestamp])) as varchar(20)), count(*) from TxAccountCreates
where timestamp > getdate()-200
and fee = 0.0000 and delegation = 0.0000
group by cast(datepart(ww,convert(date,[timestamp])) as varchar(20))

-- Day Creates
select cast(datepart(month,convert(date,[timestamp])) as varchar(20))+'/'+cast(datepart(day,convert(date,[timestamp])) as varchar(20))+'/'+cast(datepart(year,convert(date,[timestamp])) as varchar(20)), count(*) from TxAccountCreates
where timestamp > getdate()-190
and fee = 0.0000 and delegation = 0.0000
group by cast(datepart(month,convert(date,[timestamp])) as varchar(20))+'/'+cast(datepart(day,convert(date,[timestamp])) as varchar(20))+'/'+cast(datepart(year,convert(date,[timestamp])) as varchar(20))

-- Creator on most recent busy day
select creator, count(*) from TxAccountCreates
where timestamp between '01/09/2019 00:00:00' and '01/09/2019 23:59:59' 
group by creator
order by count(*) desc

-- Creator totals
select creator, count(*) from TxAccountCreates
where fee = 0.0000 and delegation = 0.0000
and timestamp < '03/14/2019 00:00:00'
and timestamp > '09/25/2018 00:00:00'
group by creator
order by count(*) desc

-------------------------------------------------


-- Posts
select creator, count(*)
from TxAccountCreates
inner join Accounts on new_account_name = name
inner join Comments on name = author
where fee = 0.0000 and delegation = 0.0000
and timestamp < '03/14/2019 00:00:00'
and timestamp > '09/25/2018 00:00:00'
and depth = 0
group by creator
order by  count(*) desc

-- Comments
select creator, count(*)
from TxAccountCreates
inner join Accounts on new_account_name = name
inner join Comments on name = author
where fee = 0.0000 and delegation = 0.0000
and timestamp < '03/14/2019 00:00:00'
and timestamp > '09/25/2018 00:00:00'
and depth > 0
group by creator
order by  count(*) desc

-- up-votes
select creator, count(*)
from TxAccountCreates ac
inner join Accounts on new_account_name = name
inner join txvotes on name = voter
where fee = 0.0000 and delegation = 0.0000
and ac.timestamp < '03/14/2019 00:00:00'
and ac.timestamp > '09/25/2018 00:00:00'
and weight > 0
group by creator
order by  count(*) desc

------------------------------

select count(*) from accounts where posting like '%drugwars%'

-- Hardfork
select count(*)
from TxAccountCreates
inner join Accounts on new_account_name = name
where fee = 0.0000 and delegation = 0.0000
and posting  like '%drugwars%'
and timestamp < '03/14/2019 00:00:00'
and timestamp > '09/25/2018 00:00:00'

    -- 1st Jan
    select count(*)
    from TxAccountCreates
    inner join Accounts on new_account_name = name
    where fee = 0.0000 and delegation = 0.0000
    and posting  like '%drugwars%'
    and timestamp < '03/14/2019 00:00:00'
    and timestamp > '01/01/2019 00:00:00'


Cheers

Asher

Sort:  

Hi Asher, great follow-up to your previous analysis! Account creation is still strongly dominated from what 'steem' is doing and the graph on the accounts created per day again shows their change in strategy from creating new accounts once every 1-2 weeks to a more continuous creation process. I like the activity analysis of the created accounts! I think we've seen the blocktrades accounts in average more active than others in other occasions before? You've likely found the reason for the relatively high number of upvotes from actifit-created accounts, but I'm surprised by the comment rate of fundition-created accounts. Do you have any idea on the reasons there? The distribution of posting authorities is a cool addition to the analysis! This aspect can easily make a full analysis on its own, but it nicely shows which apps attract users. Drugwars is certainly a strong driver for new (alt-)accounts in the last couple of weeks. Going through the list of account creator names, there are some who even used enumerated account names for the game. I guess we won't be able to answer the question of how many individuals are actually behind :)

Your contribution has been evaluated according to Utopian policies and guidelines, as well as a predefined set of questions pertaining to the category.

To view those questions and the relevant answers related to your post, click here.


Need help? Chat with us on Discord.

[utopian-moderator]

Hi @crokkon

Thanks for the review!

Thanks for correcting me on 'steem's change to the way they create accounts, this looks like a much better approach by them.

'blocktrades' scores really well in all areas, and I think that second only to 'steem', these claims are going to more individual people, and less towards alts.

I agree that a more focused analysis on posting auths and activity would be a worthwhile venture, perhaps with some activity criteria such as last post/comment via that application, and not just related to claimed accounts.

Accounts created via fundition claims is certainly eye-catching, especially considering they don't actually have their own account creation process and redirect to Steemit sign-up. I'm not sure what we would find if that rabbit-hole was explored ;)

Cheers!

Asher

Thank you for your review, @crokkon! Keep up the good work!

Well apart from it being an interesting analysis, I would also like to thank you for making me realise I wasn't actually claiming accounts automatically as I thought I was (I swear I had it scheduled at some point) lol. Just set it back up again, although I don't really have any use for the accounts... well maybe (and hopefully) they will come in handy in the future!

This one was a struggle, and i'm not entirely happy with it, I think the idea was better than the execution.

Good to know it at least got you going on the claims again though - I've used a few of mine, no guesses what for ;)

Cheers!

@abh12345, After Hardfork and after Dapps empowered with Account Creation the pace is increased and in my opinion it proved very beneficial aspect when it comes to niche based users which inturn can prove very positive point because it will boost the Active users. When Dapps will bring users then most of the times users will visit specifically for that use case. Stay blessed.

Posted using Partiko Android

It would interesting to see how this look when looking at community based projects that do not have Dapps to see how they are leveraging the process. I believe the theory was that they can onboard users and also engage with them actively to improve engagement. A limitation of this whole process may also be the lack of RC delegation pools which could help new users with bandwidth.

Posted using Partiko iOS

interesting stats. would be interesting to know how much of those are real unique people

Posted using Partiko Android

Wouldn't we all :) Well, except steemit inc who are happy to flaunt the 1.2 million account total, and talk less about there only being 20k 'active' a week - alts, community accs, bots - 5/10k people imo.

there are probably more than 10k but definitely not 1 mill :D

Something needs to be done about these bot accounts. They will ruin Steem. I get the allure of using them to mine, refer, spam tiny curation upvotes, spam comments, etc. However, it will clog the network and make it even less appealing for newcomers. I hope one day something is done, such as unverified accounts, those not attached to unique phone numbers or which haven't done KYC are discluded from certain activities or promotions in the future.

Perhaps sometime in the future, for some of the apps their will be Oracles and some form of identification process in place to be able to contribute and earn from these places.

The RC cost to make one account is way to high, I would love to create an alt account that way but can't because of lack of RC... Guess I need to get some STEEM and create it the other way.

BTW I think drugwars also helped with the account creation spike, people were creating accounts left and right to farm drugwars... I've seen people with 7 accounts just to play drugwars!

I'll create you one, contact me on discord.

As for the price, I'm OK with it, or at least unsure on what 'price' I would set the cost at. Initially the cost in RCs was much cheaper and the demand was strong, the RC cost shot up (300000000000000000000 RCs more than at present) and is now fairly stable.

7? I would guess some are playing with 50+

🤔 ....that's a lot of data! 😯 Go Asher!! Work your magic. I love the quote you shared in the end. ✌❤🐔

Quite a lot yes, and tricky to present - There is so much on this topic. Drilling down to the activities of accounts created by 1 creator would just take too much time, but perhaps in some cases, yield the most interesting finds.

I did find the quote suited this data delve quite well, thanks!

The discovery is the feature is working as planned. :)

Posted using Partiko Android

This post has been included in the latest edition of SoS Daily News - a digest of all the latest news on the Steem blockchain.

Coin Marketplace

STEEM 0.31
TRX 0.12
JST 0.034
BTC 64742.01
ETH 3172.49
USDT 1.00
SBD 4.10