A look at the data behind the seemingly increasing number of flags/down-votes that are been distributed on the Steem Blockchain over the past 6 months.
Are the number of flags rising?
Who are the top 'distributors'?
Who is on the receiving end?
Previous analysis of this type can be seen here:
A brief introduction to 'Flagging'
Flags, or down-votes are a method to reduce the rewards on a post and or make the post less visible on the Steemit user interface.
This can be for a number of reasons including a disagreement on the rewards payout, which may be due to plagiarism or spam, as well as just a general lack of approval on the pending rewards.
Additional flags may be given due to general disagreement on what is being said, or a general dislike to the user making the comments. Flags could also be given in retaliation to a flag, and although there is criteria behind their distribution, a reason is not compulsory when a flag is being issued.
With up-votes, we have down-votes. Voting power is consumed either way, and the user flagging is entitled to do as they wish.
Flags - A six month high level data review
Data in this section has been captured from the 1st September 2017 through the final day of February 2018,
When comparing the full date range, we can see that from September to early December, the day-to-day flag numbers were generally around 2500, with a few spikes to around 10000.
From the middle of December onwards though, the average rose considerably, touching 10000 on many occasions, and there are a number of days when the total count finished above 20000.
The above chart in numbers:
From this chart and data, it is easy to spot an overall rising trend with regards to flag issuance.
And, by looking then at the new account registrations, it is possible to see correlations between flags and new account sign-ups.
The charts do look remarkably alike, even with regards to the 'spikes' seen in, most noticeably, in January and February of this year.
If we now take a look at the price of STEEM over the past 6 months, we can see the same pattern present again:
What is interesting to note to the analyst at least, is that there is a lag on the Price of STEEM starting to rise, before the new accounts start to significantly increase.
It is difficult to suggest whether this is down to people not jumping on the Steem train quick enough, or that if there was a bottleneck with regards to new account creation, or both.
Averages and Totals
Flag averages
The average number of down-votes fell from September to November but rose sharply in December, peaking in January 2018 in the data range analysed.
Flag Totals
As with the average flag data, the total flags issued each month peaked in January after rising sharply in December.
New account registration Totals
Again, if we take a look at the new account registrations on the Steem Blockchain, we can see that the same pattern is displayed in regards to the total flags issued over the data range represented.
This seems to suggest beyond doubt that the influx in new users caused the increase in flags.
Who's flagging?
The analyst would again like to reiterate that flags do serve a purpose, and that it is the freewill of any accountto the use to both the 'up' and 'down' vote for content they agree/like and disagree/dislike.
Top 40 accounts issuing flags over the last 6 months
Together, they have issued over 225,000 of the total flags (851300) given in the past 6 months - over a quarter of all flags.
Is this list how you expected it to look?
Who's being flagged?
Here is a list of the top 20 most flagged accounts.
It is interesting to note a few names appear on both lists. Is it possible that an account could be producing so much 'flaggable' content, as well as finding so much 'abusive' content to flag?
Summary
With an influx of new users, which seem to follow a rise in the price of STEEM, we have a new set of 'uneducated' accounts joining the Steem Blockchain.
The learning curve in general has been noted by many people as being high, and with this there is also a requirement to adapt to what is acceptable, or not, with regards to content.
It is reasonable to suggest, that as the price of STEEM falls following a spike and levels off, the same pattern occurs in the the number of new accounts registering, and the number of flags being issued.
The accounts that stay through the price lulls, have mostly learned what is accepted by the general populous and so the average flag levels fall.
We can expect that when the price of STEEM rises considerably again in future, the number of new account registrations will rise, and following that, the number of flags issued.
Finally, the analyst would like to suggest that flags are an integral part of the steemit.com platform (and other varieties), and without the option to flag a post, the level of 'abuse' would be overwhelming and unlikely retain the accounts, both producing and reading quality content.
Tools used to gather this data and compile report
The data is sourced from SteemSQL - A publicly available SQL database with all the blockchain data held within.
The SQL queries to extra to the data have been produced in both SQL Server Personal Edition and LINQPAD 5.
The charts used to present the data were produced using MS Excel.
A selection of SQL Scripts used
This data was compiled on the 2nd March 2018 at 7:00 pm (UTC)
I hope you found this analysis interesting.
Posted on Utopian.io - Rewarding Open Source Contributors