How many of these comments are made? How many are upvoted and down-voted? And who are the popular authors and voters of such 'nice comments'?
This is an analysis with an aim to find out if there is any benefit or not to making one of the short, spam-like comments mentioned in the title of this post, namely 'nice post', 'good post', and 'follow me'.
Contents
General / Assumptions
Six month counts
Success/Fail (upvote/down-vote) rate of the comments anaylsed
Popular authors / voters
Summary Analysis
Tools used to gather data and compile report
General
In section 1, the previous 6 months data is gathered.
For the remainder of the report, the previous 30 days only have been assessed. This is due to the load generated (and time taken) while executing such queries against SQL Server.
The comment length queried has been limited to 25 characters (chars) to also reduce load, and catch comments such as 'nice post, upvoted', 'follow me i follow you', and exclude longer comments which may include the words 'nice post'.
The vote counts do not include the self-voting of comments to aid in the assessment of if 'another user' is likely to upvote such a comment.
Not included in the analysis is the payouts/sizes of any votes issued.
1. Six month counts
In this section, a presentation of the growth of the analysed comments over the past six months.
Nice Post
As the chart indicates, we have had a substantial rise in comments of less than 25 characters containing 'nice post' over the past 6 months, which for January 2018 are double what they were for August 2016. September and December buck the rising trend which raises an eyebrow, but generally this comment is on the rise.
Good Post
Excluding November 2016, we have seen a rise in the comment 'good post' month on month and for January 2018, the number is almost 4 times the total of August 2016.
Follow Me
Although there are approximately 50% more 'follow me' comments when comparing January 2018 to August 2017, the chart/numbers are perhaps the most varied of the 3 comments analysed across the 6 month dataset.
2 Success/Fail (upvote/down-vote) rate of the comments anaylsed
In the section, a look at the likelihood of receiving a vote or not for the comments analysed.
Nice Post
Of the 11711 comments made totalling 25 characters or less which included the text 'nice post', 24% of these received an upvote - about 1 in 4. This 'success rate' is reduced to just over 1 in 5 when you include the number of downvotes such comments received.
Good Post
Of the 8517 comments made totalling 25 characters or less which included the text 'good post', 52% of these received an upvote - just over half. Only 90 of these comments received a down-vote, just over 1%.
Follow Me
Of the 1976 comments made containing the words 'follow me', over three quarters (81%) of these comments did not receive a vote. This percentage is increased to 83% when you include the 2% likelihood of a downvote when making this comment.
Summary
A comment made of 25 characters or less including the text 'follow me' is seemingly far less likely to receive a an upvote than a comment of equal length containing the words 'good post'.
A particular large percentage of comments containing 'good post' are being upvoted. This seems to be the best of the 3 comments to use when seeking a vote, but as section 3 highlights, their could be other factors not shown here.
3. Popular authors / voters
The final section takes a brief look at the accounts doing the voting on the comments analysed, and the accounts receiving these votes.
Are the votes widely spread, or is there an obvious group of accounts issuing the votes?
Nice post
The table above seems to show that 3 accounts are receiving the majority of the votes for the comment 'nice post'. It's interesting to find that the top 5 entries have been voted exactly 38 times by 5 different accounts.
Follow me
A much wider spread of votes on comments 'follow me', however, one account appears 9 times in the first 20 records.
Good post
Perhaps the most interesting set of data in this section goes to the comment 'good post'. The analyst shall remain impartial to the data displayed and let the reader decide what is happening here!
4. Summary Analysis
Each comment analysed has seen growth over the past 6 months, from 50% ('follow me') to almost 400% ('good post').
The percentage of upvotes for each of the 3 comments is higher than I expected, particularly for the comment 'good post'.
However, the data presented in the 3rd section of the report may well indicate the presence of 'sock puppet' accounts or voting circles, which is likely to be the reason for these higher than expected figures.
Also, the number of down-votes issued for these comments is not as high as expected, and this seems to show a high level of tolerance towards such 'low quality' comments. Perhaps the general community is opting to mute or ignore such comments instead of acting upon them.
To conclude, it is of the analysts opinion that too many votes are being issued to these 'spam-like' comments. However, a large proportion of these votes seem to be coming from and going to a concentrated set of accounts.
5. 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.
Example code
The charts used to present the data were produced using MS Excel.
This data was compiled on the 9th February 2018 at 11 am (UCT)
I am part of a Steemit Business Intelligence community. We all post under the tag #blockchainbi. If you have analysis you would like to be carried out on utopian-io/Steem data, please do contact me or any of the #blockchainbi team and we will do our best to help you.
Thanks
Asher @abh12345
Posted on Utopian.io - Rewarding Open Source Contributors