Steemit 'Smack-down' analysis - Flag usage: September 2017 - February 2018

in #utopian-io6 years ago (edited)

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:

Steemit - Flag analysis 1


t2.png


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,

image.png

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:

image.png

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.

image.png

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:


steemprice.png
source

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

image.png

image.png

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

image.png

image.png

As with the average flag data, the total flags issued each month peaked in January after rising sharply in December.

New account registration Totals

image.png

image.png

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

image.png

@mack-bot, @cheetah, @blacklist-a, @spaminator and @steemcleaners are all (hopefully) known to many as the main accounts fighting abuse on the platform. It is a welcome sight to see these accounts present in the first 6.

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.

image.png

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

image.png

This data was compiled on the 2nd March 2018 at 7:00 pm (UTC)



I hope you found this analysis interesting.

Asher @abh12345



Posted on Utopian.io - Rewarding Open Source Contributors

Sort:  

.

Thank you!

Ahh really? In the first analysis I made a point of mentioning zappl as they seems to have a huge number of flags.

Sooo, who's is the owner key? Someone big don't like the zaps!?

.

Excellent work this, I understand it better than I did when you first published it.

Certainly some interesting accounts in there!

@crokkon I was going to ask the question about zappl because I was surprised to see that in the data above

Hey @crokkon, I just gave you a tip for your hard work on moderation. Upvote this comment to support the utopian moderators and increase your future rewards!

Thank you as always for compiling these things @abh12345; it's something I would never do, but I love to look at the analysis overall and try to put a 'human' spin on the numbers.

If it's safe to assume that the flags increase with new users, then the work that, for example, we are doing with @newbieresteemday, is very valuable. If we can continue catching new folks in our nets before they unknowingly do a faux pas, then that is time very well spent. The hope is that they will take their knowledge and pass it on to others. The bigger hope is that we are creating a continually changing, selfless foundation here.

I want to focus on unknowingly for a moment as well; not long ago, a fellow with a reputation of 25, commented "cooperate please" on my post. After teaching ESL for years, I assumed he was doing the "I upvote you, you upvote me business". He got a few jabs from others on my page, but I decided to go to his page to see his work, which by the way wasn't bad. So I left a comment, and his reply was the same kind of nonsense. I thought about writing him off then and there, but didn't. I instead wrote a lengthy comment explaining why his behaviour was in poor taste. He eventually thanked me profusely and came back to my post and apologised to the other guys there too.

I'm a firm believer in second chances; not third and fourth ones. I also think there are a lot of new people here who honestly just don't know any better, and if we are all in this together, I think it's all of our responsibility to teach others, or at the very least, give them an opportunity to improve.

I'm not naive enough to think that all the abuse here is unknowing; in fact, I know the exact opposite, But if we can focus on the newcomers, perhaps when they become "old-comers" (yes, I know that's not a word;), we won't have the selfish abuse.

Ideal and optimistic I know. But that's the way my head dreams!

One final thought; when people flag, I guess there's no where that a "reason for the flag" is also documented? It would be an interesting table/chart to analyse as well.

Geez, this is starting to look like a @davemccoy comment haha I'll leave it at that; thanks again for the post, and of course for reading all of this :)

Cheers!

When I saw this comment as I scrolled down I immediately thought Dave was here again! :D

Knowledge is key with regards to some of the flags, but I think the largest % of flags come to counter blatant abuse.

It's very mindful to give 2nd chances, and it worked really well in your example - well done you!

Sadly there is no where to give a reason, although the main counter-abuse accounts are probably very grateful for this - they are already reaching limits with regards to how many flags(votes) they can issue in a day, and so bandwidth worries would make their job even tougher!

Have a great weekend!

haha commenting like @davemccoy...I'll take that as a compliment ;)

Thank you for your kind response, and yes "sadly" is right. It would such useful information as well as a possible deterrent.

Have a great weekend as well!~

Nice collection of charts here. It gives a clear picture of who (whether a person or bot) is flagging. Obviously cheetah and steemcleaners would be mentioned.

I understand the basis and purpose for flagging or down voting; but have never done any myself. If someone writes a post that is derogatory or inflammatory, I generally skip over it and don't stop back by the person's page. I know you can mute people too, or choose to not follow them... and those are the routes I take.

When I am curating posts and see blatant copy/paste and plagiarism, I will report to steemcleaners but that's the extent of my negativity on the platform. I'm normally an upbeat person with a positive attitude. People do what they do for his or her own reasons but being a journalist by career, I can't go with the plagiarism aspects.

Hi Dawn, thanks for your comments :)

I've sat on the fence on the post, as this is how it should be for an unbiased view.

I do think that a lot of people feel the same way as you, and this is a go-lightly approach to take and one I adopt in the main.

The probably with this is that if too many of us turn a blind eye to obvious abuse, then it degrades the quality our platform as a whole.

Each person should care enough to police the place, with their own values / understanding of the reasons for a flag. But as you say, it is also possible to report the abuse and let someone else deal with it - this is a nice to have, and the people managing these services should be appreciated for this!

Thanks!

One thing I noticed, being a newb, is because this platform is worldwide, there are language, cultural and societal intricacies that can be misconstrued. Another reason I'm not a fan of "flagging".

Yes that does happen - a bit like with the 'sir' thing.

But I do think their needs to be a way to hide such things as porn/insults/worse.

Awesome @abh12345, thanks for doing some great analysis work and providing data for us all.

My pleasure Jeff :D

This is quite a funny piece...I just knew Bernie would make the first list and haijin the second. Those guys are always fighting.

But why would someone wana flag jerrybanfield and cheetha...I don't just get it.

Bernie made both lists :D

Haejin a new entry from the previous analysis.

Jerry - he does write a lot of stuff people don't agree with and is in the spotlight a fair deal.

Cheetahs flags are mainly retaliatory flags from abusers, I think.

Lolzz...so they all really do have reasons to be flagged .

Excellent analysis Asher, I totally agree on the correlation of spike in new accounts to the amount of flagging, because they are clearly creating a large number of spam accounts. I actually had a few posts on this and a few simple answers, take a look if you'd like: https://steemit.com/earthnation/@cryptkeeper17/my-thoughts-on-new-account-creation-on-steemit

Interesting idea for sure. However, I own at least 5 'dead' accounts for the future, and if you check crokkons link above, you'll see in the 2nd spreadsheet that other, certain high profile types do too!

I have no problem with people owning multiple accounts, what annoys me is that it seems that the word is out and people could be purchasing them, and either using them for what they want or handing them over to other people. But the platform 'giving away' a relatively high amount to people that are totally unknown to be effective or real additions to the platform, or a front for jerks to pull all kinds of shenanigans like it seems they are with your point that the new account numbers and the number of flagging is virtually identical, thanks to your great analysis that is now apparent to me btw.

Yes we need a new sign-up process, a lot of SP is sat in accounts that will likely never be used.

Cheers!

Great content. Thank you for taking time to produce this and sharing it with us.

I am a relative newbie and totally agree with your point regarding a steep learning curve. But on steemit, it's worth it.

My pleasure, thank you for your supportive comments, and welcome to Steemit!

Am really sure March is going to make that list looks like a child's play. I have been reading a lot these days from many author's and it looks like March is gonna be a month of bloodbath. From spamming comments to cave trolls, a lot of people are voicing out their engagement in the cleansing process.

The result of this: we still waiting to see.

The potential is there, but i think perhaps not due to a low STEEM price and fewer signups - lets see!

Thanks for some data on this, although it's all academic and philosophical for me at this time...

This has been a topic on several forums I'm involved in.

Data Mining Reveals How The “Down-Vote” Leads To A Vicious Circle Of Negative Feedback
A classic theory of behavioural psychology predicts that punishment should improve behaviour. But the first study of online voting behaviour in social networks shows exactly the opposite
...
"The evidence is that a contributor who is down-voted produces lower quality content in future that is valued even less by others on the network. What’s more, people are more likely to down-vote others after they have been down voted themselves. The result is a vicious spiral of increasingly negative behaviour that is exactly the opposite of the intended effect."

This fits within my understanding of behavior. What lay people "know" about behavior modification is a mix of pop culture and old wive's tales, and is often far removed from successful training.

I tend to agree with this hypothesis.

I find it funny that those flagged by cheetah and Steemcleaners would do retaliatory flags. Just sets them up for more flags.

I'm surprised with the flags for zappl and when I read @crokkon post it kinda explains things.

Things will be heating up in the coming weeks as war operations will be in full swing.

Yeah, more flags for them I think - might as well set a new account up and try to do things differently.

I think if I did the same analysis in 3 months time we'd see less flags, only because I think the price of STEEM will range for a while and new signups will be low in comparison to Dec and Jan. We'll see!

Hey @abh12345 I am @utopian-io. I have just upvoted you!

Achievements

  • WOW WOW WOW People loved what you did here. GREAT JOB!
  • Seems like you contribute quite often. AMAZING!

Community-Driven Witness!

I am the first and only Steem Community-Driven Witness. Participate on Discord. Lets GROW TOGETHER!

mooncryption-utopian-witness-gif

Up-vote this comment to grow my power and help Open Source contributions like this one. Want to chat? Join me on Discord https://discord.gg/Pc8HG9x

The master of stats does it again, this is actually quite fascinating - mainly because I have never even thought of using a flag! Maybe I'm just lucky and only ever find quality content, it's very interesting to see accounts in both the top flagging and flagged list and I think you might be right about the value of steem being linked to new account registrations.
c0ff33commentaimage.png
#thealliance

Flags aren't everyone's cup of tea for sure, and you can report abuse to steemcleaners to deal with if you want to remain a non-user of flags.

Yeah we'll see what happens with signups and flags over the coming months, I can see less of both in the near future...

No matter how you are treated to Berniesanders, but you will find his accounts in all charts: he is one of the main distributors of flags and one of the main post-promoters (I mean Randowhale). He is an active fighter against abusers, but at the same time a very active self-voter:-)
It's no surprise to see him here, as well as Steemcleaners, Haejin and many of spammers. This all is logical - the struggle of some for the purity of Steemit, and others - among themselves. But what makes a person to atack Zappl, I can not understand.

I agree.

And with regards to zappl, check Crokkons reply - seems it was one person with many accounts who isn't a fan of short content.

I assumed that the rise of STEEM correlates with the rise of registrations on here, but its nice to see some charts backing that up! I really like seeing all this data clearly presented. Thank you for taking the time to do this!

Well thank you very much @emmyem84!

I think its the case, and what I expect to see over the coming months is the general disengagement of 'abusers', and a reduction in flags, before we go round again (hopefully!) later this year :)

You know whats funny, I just joined Steemit (during the crash) so while other people may be hesitant, it didn't stop me from joining here. But I really do believe in the future of blockchain tech changing the world, so I had no hesitation. Now that I am here though I am wishing for a bull run and at the same time im hoping it doesnt because I feel like I need to buy more Steem. Crypto and Steemit are a rollercoaster lol.

Yes I think its a case of riding the downward swing, building up some followers and readying for the next move. It could come at the end of the year, or next month. I would consider buying a little STEEM and powering it up, and think that its best to do it in stages. e.g. The price looks good compared to January, but is still 3x more than it was last November.

Flagging is just like scolding, a means to rebuke or correct a mistake. But bullies misuse it. They use it to terrorize people

I would agree that flags can be misused, and would also suggest that up-votes are too.

I completely agree with this... I'd say that upvotes are way more abused than flags ever will be.

It's interesting to see the correlation between new accounts and flags. Just like you said, I think it's the new users who I think may not be aware of how things are done here who may be more on the receiving end of the flags. I have no idea that there number of flags are that many. Thanks for sharing.

A pleasure. The correlation is only a hunch but the numbers seem to support it. Cheers!

Interesting to take a look at this and it's a somewhat controversial topic.

I think there are a few big flag wars that flared up around the time of your data spikes so maybe the key here is to try and frame a data query around identifying and counting those flag wars. I'd expect these to be flags in both directions so there may be a clue in the data there.

I think the flagging with regards to 'wars' has only started in earnest the last week or two - flags have been falling for the past month or so, and so I don't think this has made much impact, so far!

Like you already wondered, I am also curious to know why so many people flag @steemcleaners?
The only reasons I could think of is that people try to take revenge? Or by flagging their posts they prohibit that the @steemcleaner account will become more powerful?

Only small accounts do the flagging of steemcleaners, mainly to retaliate. The account is very big so it will have no effect on the reputation.

I'm glad I follow your posts. You may think I have learned a lot (That is the case indeed), but there is so much more to learn, so thank you for your informative posts.

My pleasure, thank you!

This analysis is interesting, to say the least. Thanks for taking the time to gather and analyze the data. That takes time and effort that I appreciate.

Thank you :)

Yes so much time I miss comments and feel bad - catching up this morning!

Your statistics are skewed by the fact that amongst the ones doing the flagging, number 8 through number 27 are mostly the same entity, and should be added together to reveal the true top 3 flaggers. I know, you know why I know. Greetings!

Very true - but I shouldn't exclude those accounts should I ? :D

Wow !!
Excellent statistical information..
I think, Flagging is an in-built moderation measure in Discourse, intended to dissuade bad behavior and spam, as well as to promote civilised discourse
@abh12345 Nice explanation...
Thanks for giving us your essential post..

Sure hope they aren't just flagging for flagging sake. If so, what's the point of even being on the platform?

flagging for flagging sake

Can you explain this a bit more? Remember that issuing a flag drains your voting power in the same was as an upvote does, so flagging just for the sake of it is a strange way to operate imo.

People are petty and have vendettas against others and feel empowered to stifle expression. Seems like when a "whale" gets on a power trip and targets you, you might as well move on to a different medium. just my 2 cents which doesn't even mean that much here.

This is 'proof of stake' at work im afraid. Unfortunately what you state does happen at times, but is likely to be a really small % of flags given.

Thanks for the explanation. Still learning all of the ins and outs.

Yes indeed interesting mate, and thanks to the top 6 flaggers for fighting abuses in the platform.

i just hope they are guiding the spammers if they are real so that the activeness don't increase and they learn to use this platform

flagging is good they could learn to don't repeat the same mistakes again and they could learn more

the effort you have put into making this post is outstanding i can't even compare hat's off :)

Interesting data. Just taking a quick look at the accounts doing all the flagging... I saw @nextgencrypto show up at #10 and then related @nextgencrypto accounts showing up at 38,39,40... (and @nextgen accounts at #34-#37). I guess I'm still new here, but I just don't understand that type of Steemit behavior. It makes me question what people are really doing on Steemit, and how people are manipulating the platform in different ways. Am I missing something here?

Those are owned by the same user :)

I saw a new user who intentionally posted 14 times with the tag introduceyourself to get upvotes. She eventually got flagged.

14 times? thats too much,guess she was enjoying the upvotes..lol

I try to warn people first with misuse of tags, but I have flagged repeat offenders.

Hmmm... ok, here we go to research who's this quarry guy.

Did you uncover anything? :D

excellent your post Thanks for sharing i will done up vote

This post has received a 67.98% upvote from @lovejuice thanks to @abh12345. They love you, so does Aggroed. Please be sure to vote for Witnesses at https://steemit.com/~witnesses.

half of them are due to the flag wars

flags trend goes with prices huhh that's kind of interesting

@abhi12345 your analysis are so precised no one can do such a deep research on something but you surely do and that is why i love to read your posts

whales are fighting hard to combat again spam that is great and thank to you bro for such a detailed post

thanks a lot for the information sir... i was finding about this flagged thing.

I think this is too much information for my brain lol

Very interesting article. I was surprised at the proportional ratio of the number of flag and the price for STEEM. Thank you for doing such a great job.

Очень интересная статья. Я был удивлен пропорциональным соотношением количеством флагом и цене на STEEM. Спасибо, что проделали такой огромный труд.