Analysis of Steemit Bidbot Business - Steemit Business Intelligence

in utopian-io •  7 years ago  (edited)

The Bitbot business has been growing on steemit for some time now. Some people love them and some people hate them, but facts remain. For better or for worse they are here and they are providing a service that people are using.

For that reason I am developing a weekly report to analyse the amount of SBD and Steem Vests that are recycles through steemit via the use of Bidbots. The aim of this analysis is to track the growth or decline in the use of these services over time. By creating a weekly report, accurate information will be freely available on how much these services are actually used and we can also begin to spot trends or changes in trends between these bots.

As this is the first report, to improve it moving forward it might need some tweeks. So I will detail the actual findings and I will fully detail how I went about this analysis. I hope that you will read the steps taken to carry out this analysis and help me fine tune the results.

This analysis has been carrier out on 1 week of data from the 18th March to 24th March.

All of the Bidbots included in this analysis are listed on https://steembottracker.com

How much SBD has been sent to Bidbots for Votes?

image.png

Looks a bit messy right? But stick with me because there was a bit of work involved in calculating how much SBD was sent to bidbots for the week in question.
First I ran a query to gather all of the transfer's of SBD to these bots. The SQL query used here was

Select 
*
From TxTransfers (NOLOCK)
Where timestamp >=CONVERT(DATE,'2018-03-18') and timestamp <=CONVERT(DATE,'2018-03-25')
and [to] in ('kittybot', 'isotonic', 'getboost', 'booster', 'lightningbolt', 'nado.bot', 'shares', 'peace-bot', 'postdoctor', 'spydo', 'singing.beauty', 'minnowhelper', 'jerrybanfield', 'sunrawhale', 'sleeplesswhale', 'zapzap', 'cryptoempire', 'pwrup', 'redwhale', 'lovejuice', 'foxyd', 'noicebot', 'minnowfairy', 'honestbot', 'seakraken', 'upboater', 'whalebuilder', 'mrswhale', 'upgoater', 'smartsteem', 'dailyupvotes', 'pushup', 'allaz', 'sneaky-ninja', 'upmyvote', 'dolphinbot', 'minnowvotes', 'chronocrypto', 'thebot', 'inciter', 'oceanwhale', 'promobot', 'mitsuko', 'bearwards', 'voterunner', 'brupvoter', 'payforplay', 'edensgarden', 'redlambo', 'mercurybot', 'appreciator', 'slimwhale', 'moneymatchgaming', 'boomerang', 'childfund', 'buildawhale', 'youtake', 'megabot', 'authors.league', 'upme', 'alphaprime', 'steembloggers', 'msp-bidbot', 'lost-ninja', 'estream.studios', 'upyou', 'postpromoter', 'rocky1', 'bluebot', 'flymehigh', 'aksdwi', 'puppybot', 'onlyprofitbot', 'boostbot', 'discordia', 'canalcrypto', 'therising', 'ebargains', 'fishbaitbot', 'luckyvotes', 'brandonfrye', 'estabond', 'upmewhale', 'hotbot', 'adriatik', 'steembidbot')

However what I found with this query was that there were transfer's made between accounts that had nothing to do with bid, these looked like funding type payments, so I needed to exclude these. From looking at the memos on these transfers, most of these had no memos, and if they did, the was not url included in the memo.

So using Power BI, on the query above, I also added a filter to remove any transfers where the memo did not include some sort of url.

This has returned a value of 79,410 SBD sent to bidbot. In the above visualization the table on the left shows this 79K by bidbot.

I then looked at the memos for these transfers again. You can also see a section of this table in the above visualization. From here it was clear that there were some transfers still to be filtered out. What it appears from the data set is that all transferred for votes begin with a url (so the first few characters were https).

To account for this, I used power bi to tot the value of all transfers that where the memo does not begin with Https and I deducted this value from the total transfers. You can see in the above a value for total SBD transferred adj.

The next step in calculating the total SBD transferred to these bidbots was to look at refunds. The SQL query used here was

Select 
*
From TxTransfers (NOLOCK)
Where timestamp >=CONVERT(DATE,'2018-03-18') and timestamp <=CONVERT(DATE,'2018-03-25')
and [from] in ('kittybot', 'isotonic', 'getboost', 'booster', 'lightningbolt', 'nado.bot', 'shares', 'peace-bot', 'postdoctor', 'spydo', 'singing.beauty', 'minnowhelper', 'jerrybanfield', 'sunrawhale', 'sleeplesswhale', 'zapzap', 'cryptoempire', 'pwrup', 'redwhale', 'lovejuice', 'foxyd', 'noicebot', 'minnowfairy', 'honestbot', 'seakraken', 'upboater', 'whalebuilder', 'mrswhale', 'upgoater', 'smartsteem', 'dailyupvotes', 'pushup', 'allaz', 'sneaky-ninja', 'upmyvote', 'dolphinbot', 'minnowvotes', 'chronocrypto', 'thebot', 'inciter', 'oceanwhale', 'promobot', 'mitsuko', 'bearwards', 'voterunner', 'brupvoter', 'payforplay', 'edensgarden', 'redlambo', 'mercurybot', 'appreciator', 'slimwhale', 'moneymatchgaming', 'boomerang', 'childfund', 'buildawhale', 'youtake', 'megabot', 'authors.league', 'upme', 'alphaprime', 'steembloggers', 'msp-bidbot', 'lost-ninja', 'estream.studios', 'upyou', 'postpromoter', 'rocky1', 'bluebot', 'flymehigh', 'aksdwi', 'puppybot', 'onlyprofitbot', 'boostbot', 'discordia', 'canalcrypto', 'therising', 'ebargains', 'fishbaitbot', 'luckyvotes', 'brandonfrye', 'estabond', 'upmewhale', 'hotbot', 'adriatik', 'steembidbot')

What I found running this query was the amount of bitbots using other bidbots, but as that is not the aim of the analysis, I needed to filter these transactions out, as they have been included in the first query above. To do this, I placed a filter in Power BI to remove all transfers with memos that did not include the word ‘refund’

This refund value can also be seen in the visualization above
From here, I deducted the refund amount from the Total SBD sent adj. to come up with a fully adjusted SBD transferred value.

For the week 18th – 24th March, the amount of SBD sent was 75.70K.

What % of SBD Rewards Claimed on Steemit were sent back to Bidbots?

All curation rewards are paid in vest or SP. Only author rewards are paid in SBD. Now that I know how much was sent to the bidbot businesses I wanted to see what % of actual SBD rewards claimed were sent back to bidbots.

To gather this data the query used was

Select * From TxClaimRewardBalances (NOLOCK)
Where timestamp >=CONVERT(DATE,'2018-03-18') and timestamp <=CONVERT(DATE,'2018-03-25')

Using the data returned from that query and the total SBD sent, I was able to calculate the % of Rewards Claimed that was sent to bidbots

30% of all Author SBD Rewards were sent back to bidbots last week.

image.png

What % of Steemit Vests are paid to bidbots for Curation.

Vests are paid to both authors and curators. All curation payments are in vests. For authors if you power up your post 100% then all the authors rewards are in vests, but if you take payment 50/50 then 50% of your rewards are paid in vests. Therefore Vests make up a very large portion of the rewards.

To get the curation rewards for bidbots only I carried out the following SQL query

SELECT
reward,
timestamp,
curator
FROM VOCurationRewards (NOLOCK)
Where timestamp >= CONVERT(DATE,'2018-03-18')             and         timestamp<= CONVERT(DATE,'2018-03-25')
and [curator] in ('kittybot', 'isotonic', 'getboost', 'booster', 'lightningbolt', 'nado.bot', 'shares', 'peace-bot', 'postdoctor', 'spydo', 'singing.beauty', 'minnowhelper', 'jerrybanfield', 'sunrawhale', 'sleeplesswhale', 'zapzap', 'cryptoempire', 'pwrup', 'redwhale', 'lovejuice', 'foxyd', 'noicebot', 'minnowfairy', 'honestbot', 'seakraken', 'upboater', 'whalebuilder', 'mrswhale', 'upgoater', 'smartsteem', 'dailyupvotes', 'pushup', 'allaz', 'sneaky-ninja', 'upmyvote', 'dolphinbot', 'minnowvotes', 'chronocrypto', 'thebot', 'inciter', 'oceanwhale', 'promobot', 'mitsuko', 'bearwards', 'voterunner', 'brupvoter', 'payforplay', 'edensgarden', 'redlambo', 'mercurybot', 'appreciator', 'slimwhale', 'moneymatchgaming', 'boomerang', 'childfund', 'buildawhale', 'youtake', 'megabot', 'authors.league', 'upme', 'alphaprime', 'steembloggers', 'msp-bidbot', 'lost-ninja', 'estream.studios', 'upyou', 'postpromoter', 'rocky1', 'bluebot', 'flymehigh', 'aksdwi', 'puppybot', 'onlyprofitbot', 'boostbot', 'discordia', 'canalcrypto', 'therising', 'ebargains', 'fishbaitbot', 'luckyvotes', 'brandonfrye', 'estabond', 'upmewhale', 'hotbot', 'adriatik', 'steembidbot')

to get the Total Vests paid out I carried out the following query

Select 
Account,
Reward_vests,
timestamp 
From TxClaimRewardBalances (NOLOCK)
Where timestamp >=CONVERT(DATE,'2018-03-18') and timestamp <=CONVERT(DATE,'2018-03-25')

image.png

Almost 6.5% of all Vests earned on Steemit are paid to bidbots.

Which bidbots earn the most in Vests for curation rewards?

If we take a look at this by bidbot we can see that almost half of this is split between 4 bidbots.

image.png

Conclusion

With this analysis I have answered 4 important questions in relation to the bidbot business on Steemit.

How much SBD has been sent to Bidbots for Votes?
What % of SBD earn on steemit is sent to bidbots?
What % of Steemit Vests are paid to bidbots for Curation?
Which bidbots earn the most in Vests for curation rewards?

Moving forward I will be tracking this data an reporting on it here as it makes up a large part of the current steemit ecosystem.

What other information would improve this report? What do you think of the findings? Please do leave your comments and feedback below



Posted on Utopian.io - Rewarding Open Source Contributors

Authors get paid when people like you upvote their post.
If you enjoyed what you read here, create your account today and start earning FREE STEEM!
Sort Order:  

Thank you for the contribution. It has been approved.

You can contact us on Discord.
[utopian-moderator]

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!

  ·  7 years ago (edited)

Thanks for analysing these guys. They are extremely important in steemit economy.

My 2 satoshis:

  1. I believe if we take all transfers with steemit url in memo that will be it.
  2. Could we build an interactive report with for e.g. fields:
  • share of bidbots in total steem rewards
  • share of bidbots in total transfers
    ?

all transfers with url are not for paid votes, many are actually questions about a previous vote or a vote not given and would could not be classified as payments for votes.

amount of such transfer should be tiny

wow.. really did not see this coming i most say, bots on steemit is really complecated but make steemians lazzy

That is some solid procedural analysis right there.

We ought to be able to use the same code to analyze historical data, as well. All of the queries involve time bounds, so nothing really keeps them from having the balance shifted by a week or longer, or even longer periods for aggregation. Which might be useful to see if there are trends already visible in the blockchain.

Certainly it would be premature to make any sort of assessment on anything less than the last historical three-month period. Given that would span quite a lot of the rapid descent of the steem evaluation, that would be amazing and possibly revelatory.

It's also hard to say whether this is a reasonable amount of valuation to be stuck in bid bots. Given that the mechanics of curation are so heavily geared to rewarding behavior which can only be executed by an automated system, I think we need to know more before we can make a judgment there, too. And I really want to be able to make a judgment about that.

Otherwise, well done.

thank you @lextenebris - the bad news is that as I was testing the model I did look back at previous data and the use of these bots have been growing

That much we could assume, but by how much?

I may have to actually dig into the transfers between bots. That sounds – weird. And kind of interesting. I wonder if they have enough SP in motion between them to define their own ecology, separate from any manual activity occurring elsewhere on the blockchain.

Bidbots are actually a profit for investor. They want a passive income without touching anything while user want visibility and also the fastest way to build reputation.

User should not blindly use the service without even understand the purpose of it. I see a lot people just spend their SBD/Steem to bidbot thinking of getting profit of it without even bother to write a simple content.

I think this is one off the major problems, it doesnt matter what the content is like, the vote will come once paid for

Some do have white-list & blacklist after they check the content manually. Blacklist will be prevent user from abusing the bitbot. Whether is a quality content is another story.

Thank you for looking into these. The figures that stuck with me and i had to re-read it to understand it implications.

30% of the daily claimed SBDs (note to self: not the reward pool) were returned to the bidbots.
And 6.5% of the daily claimed rewards were also claimed by these bidbots.

I think it would also help the readers if the results (figures) were included in the conclusion (no longer need to scroll back to the findings).

@eastmael you are right, I should have included them also in the conclusion

What I'm interested in is how much SBD is going to the bid bots vs how much SBD in vote value is going out.

That stat would answer a lot of questions.

Did I miss it?

So, I am pretty much on the right track that this is going to eventuate in a very narrow circle owning vast amounts of the vests and the claim of distribution is a poor one considering they only make up ~1.5% of votes cast?

Any hope..??

Thank you so much for this.

I think there is hope. Steemit is self governed. did you see the vlog yesterday from @fullltimegeek? He shares his feelings on bitbots - worth a watch

will give it a look, thanks again.

Thank you again Paula for your efforts to provide us all with the quantitative insights on what is happening on this platform. This sort information allows for better rational discussion rather than purely perceived observation.

30% is hefty amount and now confirms the suspicion held by many that using these bots is indeed making a few people very wealthy for little or no effort.

I'm hoping some one such as @dan comes up with a better platform.

it is to my understanding that @dan is creating a new platform that will address issues like this, however this is the hear and the now. Steemit is the platform we are on and steemit is self governed. I believe we have the power to stop this if we really want to .

To stop it, the top witnesses need to be on board. But, it would appear that they don't care because some are in on it. I have my suspicions that the last hard fork was done with the intent that the mess Steemit finds itself in would happen, because, those in on the know would profit.

The other option is vote in other witnesses to steer the ship. But that is going to be a long road I think.

As for @dan, we can only hope that he's been taking careful note.

yes there are some top witnesses involved and making considerable money leasing SP to these bots. Eventually there will be a witness voting war over this.....I can feel it coming....

Do you think the witnesses not involved in bidbots here on steemit are not taking notice? forks are dependent on witnesses and as we self govern Im sure they have something to say. we should be shouting out to the witnesses here on steemit not @dan

The cynic would note that witnesses with access to broad swaths of bid bots have a ready-made army with lots of ammo ready to rollout in the case of any sort of voting war.

Let us not forget that the bots with huge amounts of leased SP carry just as much power when they put in a witness vote.

Ehat do you mean by a witness voting war @paulag. Do witnesses really have the power to stop it?

Well researched and analysed, u actually did a nice work here. Before now I ve known little or nothing about the bidbot business on steemit but now I think I ve grown bigger in its knowledge and hope to learn more in ur subsequent analysis.
Thanks

glad my post is of value to you

I may not be the type of audience that this post is aimed at @paulag but one thing that would be helpful to me is more explanation as to what these figures mean to those of us not using bots. As far as I can tell it's bad news.

I feel like I am getting further and further behind despite the generous support of yourself via @abh12345 by way of delegation.

If these bots are growing and they already take up such a large proportion of the reward pool it seems like there will come a day when there will be nothing left for humans. But maybe that's not the case. it looks to be moving in the wrong direction though.

It would be nice to understand the bidbots economics. How do they make money? What is their business model?

their money making model is rather simple. The vote costs them nothing, like it costs me or you nothing to vote. However many of these bots pay for leased delegation so they have a big vote value. So they get income from both the bid people place, and curation rewards. and their costs are the lease fees leaving them with a profit

Thanks, paulag!

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

Achievements

  • 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

Thanks for sharing! A link to your post was included in the Steem.center wiki page Upotes Bots. Thanks and good luck again!