Curation Report (End-June) [with observations on HF19]

in #curation7 years ago (edited)

Two reports in ten days? It simply has been that turbulent in the world of Steem curation.

Be warned, as always, this will be an in-depth rant into the curation world, so it's only for curation geeks like myself.

In my last report, right after Hardfork 19, I noted that the payouts were way higher than expected. Things seemed too good to be true - and they were. The reward_balance had ballooned up during the HF18 era for some reason, and was being expended. This results in exaggerated payouts for everyone. Things are balancing out, though I know not when and how an equilibrium will be reached. The current reward_balance is 1.087 million Steem; the 30 day reward pool is 1.44 million.

During the last 10 days and the 7 days prior to Hardfork 19, I went on a curation rampage. I made over 1,500 votes, reaching out to over 800 new authors. I didn't have time or voting power to curate comments though, so this could have been much higher. Every night, my Voting Power would be in the 3%-5% range.

I also rented SP one week prior to HF19, and that turned out to be extremely profitable. It seems like everyone thought the rates were ridiculous, so I got lucky to be one of the few to see the value in it. I didn't think the reward pool would be so exaggerated, though!

Now, things are starting to balance out, and as such, so are my strategies. I can't afford to pay much more for rentals anymore, but I'm still in the market. In my previous report, I mentioned how using high voting strengths while taking your voting power as low as possible would be beneficial to begin with. However, in my simulations, from the third week onwards both approaches pretty much even out as the regeneration achieves an equilibrium for both. So, I'll start mixing and matching, giving out 100% votes to the most engaging posts, while going with lower strengths for lesser but still worthy content. Combined with downscale of curation, my VP will be much higher than the 5% it has been averaging. Today is the first time I've seen it above 10% since a few hours after HF19!

My general curation strategy remains unchanged, looking for undiscovered posts. This is mostly new authors, but sometimes I'd also run into a gem of a post by an older author that has gone undiscovered. I have finally integrated Esteem.life in my workflow. However, it still has some bugs which I'd like to see fixed. Still, with 20,000 posts being made daily nowadays, I can only consider reading a couple of hundred posts a day.

Digging through the depths of Steemit also gives you a great perspective to how Hardfork 19 has really impacted Steem. There are no real surprises here - it's pretty much all as expected, the good and the bad. Do note that this is all anecdotal observations.

  • Earlier, I often felt like I was the only one digging through the new posts in my timezone, as most posts were stuck at $0.00. But now, I can see minnow votes on many posts. $0 posts are far less common nowadays, they always seem to have a few cents on them.

  • Minnows are curating and resteeming much more. Linear rewards have definitely revitalized the community. The raw number of voters is approaching 20,000 - nearly 2x from a couple of weeks ago.

  • There's a VERY long tail of author rewards. The top authors are making much less than before - the days of one post taking up 10% (i.e. multiple thousands of dollars) of the reward pool are far behind us. A top post is currently taking less ~1% of the reward pool now, which seems like a healthy amount of inequality to me.

  • Similarly, there's a long tail of curators. Earlier, the top 100 curators on Steemdb.com accounted for 90+% of the R-shares. But really 98+% when accounting for V-shares - effective influence. Even during the experiment, this was well above 90%. Today, the top 100 curators only account for 70% of overall influence. That's an improvement of 1400% in distribution!

  • The square_root curation curve emphasizes early voting slightly less than the older system. It still ensures that voting on a trending post is a complete waste of voting power, but at the same time, you can vote on something which has $10-$50 and still make a decent curation reward. Especially if it's picked up by other voters. I feel this is a net positive, and the current algorithm seems pretty well balanced overall.

  • There's a scourge of self-voting and collusive voting. This is a serious problem that many of us saw coming, but the witnesses and developers turned a blind eye to. I'm seeing rampant self-voting and collusive voting, thanks to the 4x increase in the 100% vote. But wait, it's even worse than that. People are voting others with 10%-25% strengths, saving the 100% for themselves and their colluders. So, the self-vote is probably 10x as powerful as it was before. The end result is pretty grotesque. The abuse fighters simply can't keep up, and this is going unchecked. I have requested statistics earlier this week for the rise in self-upvotes. The voting power rate was just fine as it was. I was willing to be open minded about it, but it's now clear it's been an unmitigated disaster.

I've written a lot recently about curating responsibly, and I'm afraid that concept has been thrown out the window. To their credit, most curators are doing very well, but some are flirting with corrupt curation; some with large stakes or delegated stakes even. I'm also disturbed by the sheer amount of people getting in touch, offering me something in return for votes. All this is expected of a corrupt free market, of course, but is a dangerous slippery slope which might reduce Steem to a scammer's utopia. I use hyperbole because action is necessary.

Overall, HF19 has been a very positive story, except for one major blemish. There's no talk of fixing this either - I wonder if the witnesses and developers are even aware of this problem? I hope they are, and I hope solutions are in the works.

So, that's it. This will be my last curation report for the next few months, if all goes well. Feel free to get in touch with me on Steemit.chat if you have any further questions.

As always, I'd leave the shameless plug and request you follow my curation trail at Streemian.

Sort:  

There's a scourge of self-voting and collusive voting.

Yep. This was easily predicted but largely ignored. And the actual voting is now being ignored as well. This one particular user is doing this exact thing that you described - upvoting himself at 100% and occasionally upvoting some other users...at 1%. But worst of all is that his comments that he self-votes are mostly spam comments. I have tried to do what I can downvoting them at 1-2%, but I'm getting practically zero help from other users.

https://steemit.com/@crypto-p/comments

And I'm sure this isn't the only person doing this, but the rate at which it's being done and the amount of rewards he's pulling in by doing it is a little ridiculous. And when others have pointed out that this isn't appreciated, he flags them.

So, if anyone with some extra voting power would like to help kill some of his comment rewards, it would be appreciated.

As far as the actual hard fork changes...I think it will not play out well in the long-run.

Full linear rewards + 4x voting power = Perfect conditions for abuse/exploitation

How this was dismissed/defended by STINC is beyond me. They are turning this place into a mockery of social media and the rewards incentive, in my opinion. And there seems to be very few people who are willing to speak out against or combat the abuses/exploits. As @schattenjaeger asked in his post on the topic:

Is this what we want Steemit to be?

They are not the only one, I'm sure. This is an extreme case, but there are many, many that are less obvious. Earlier, I have thrown 100% flags the way of abusers like noganoo and matrixdweller, and would have done so here. Sadly, my voting power is so low and limited, I'll save much of it for curation. I've flagged some of their recent comments at 1%, but it barely makes a dent. So, yeah, whichever way you look, the lower vote target is a disaster.

The thing is, I can't even blame them. They are just upvoting themselves, which seems like a perfectly normal thing to do. It's just that the system is broken right now.

Ultimately, it's up to the witnesses to realize this is a problem and push the developers to rush out a solution asap.

I don't think the system is broken, I do blame the people self voting! I can't imagine any developer intervention to solve this. Any restrictions on self voting or repeated voting of same users will all be overcome by a wider circle of puppet accounts.
I've loved the 4x increase as my voting is now all manual and I'm able to meaningfully tailor vote strength. Sure it's made it easier to fully self vote, but if someone is posting 20 spam comments a day as voting targets, they'll still post 80 if it puts a couple thousand dollars in their pocket.
So even in the self voting regard the HF may have been beneficial by cutting spam posting and puppet account registration by 75%!

The solution is flags, and has to be community oriented.
I'd like to see a cheetah type bot that uses analytics like this to follow up posts and comments of those above a certain threshold, say 70% self voting, with a friendly message. "Hello! We've noticed you use 87% of your voting power on your own posts and comments! While we all need to pat ourselves on the back once in a while, it's great to get out in the community and mingle! You can find a regular rundown of engaging new content here.(insert link to good curation project.)"
The community needs to be made more aware. I know I'm not cross checking everyone's voting history, and the real kicker is that some of these 100% self voters are still getting sincere votes from other members. I even saw a post by one person who self upvotes 100% proudly proclaiming he was stopping his power down, and asking everyone to help build his Steem Power up since he's so committed to the platform!
A bot warning could give people the same pause that the duplicate content detection does for plagiarism.

The stigma around flags needs altered too. I could envision a "flag a day" educational campaign. I think everyone can and should throw a flag each day, as an important part of shaping the community.
Alternatively if an account is dedicated to combatting abuse, users should be asked to delegate SP to it and educated on how to do so. I would gladly contribute, and it puts a bit of seperation between a user and direct flagging if they are squeamish about it.

Collusive rings are harder to deal with and will probably always take a manual hand behind the controls, spotting the patterns and taking action.

And for those who equate egregious self voting with dividends and consider it a fair use of stake, you are in a sense right. But please think long term. If you can show me a company where the shareholders voted all profits be paid to them in dividends, putting a freeze on new hiring, employee raises, and all capital expenditures to support and improve operations... and the company flourished... I'll change my opinion!

Nice one,you have itemized and discusssed some of the ways to prevent these fraudulent ways of gathering votes,I hope the authorities will do something drastic.its pathetic the extent some authors will stoop so low

I fully agree. I think the linear reward curve was very helpful to improve the system, whereas the higher impact of 100 % votes was a wrong decision. I wrote a rather long article about the consequences of HF 19 myself and really had wished to get more feedback from some members who are involved in the decisions about future platform changes (I don't link my article here to get some upvotes - it is too old for upvoting anyway, but I consider the problem as massive).
Here is another example of blatant self-voting (amongst many others).

Well, you promised something for curation geeks only, but as a still kinda newbie I got a lot out of it. I stopped self-voting after HF19 hit because I realised I had to save my little upvote for the people I had encouraged to join Steemit and for those whose posts I was already enjoying. I really haven't been able to fully come to terms with the partial vote concept and your post has given me much to ponder. Thanks!

I remember you mentioned the same before. That's not just an altruistic gesture, but helps you build a reputation and follower base in the long term.

I'm afraid there are many myopic people on Steem, and sadly some with large stakes. They'll see quick buck and grab it without ever considering the consequences.

How do you invite someone over and not have enough wine for them to drink? Just saying 😜

Some would rather gulp down the wine themselves before the guests arrive...

Then they don't get invited to anyone else's parties. Short term gain, long term pain....

OK, I know in the real world often the nasty guys get ahead. Hey, I live in South Africa - we're a living laboratory for just that (if you've had the chance to read any news outside of curating posts, you'll know much of our country has been stolen by a family named Gupta from India - I kid you not). But it tends to bring out the "goodest" in the good guys.

I don't think I will ever get into the mindset of a curator.

The editor has a checkbox for self upvoting posts. So should I stop upvoting myself?

If you like your post, then it's ok to upvote it. However upvoting your own comments looks worse to me.

Yeah people upvote their own comments to get to the top. That makes sense.

HF19 is a huge success i think, but self upvoting for benefits must be limited i think. I am a minnow and thats why i feels so. All minnows feels same i think. And surely other whales wont agree i think, its just relative, so far so good.
Thanks for sharing such a detailed info.

Interesting read, it's good to have a more informed perspective. I still don't really understand curation or how it might have been affected by the Hardfork. I've seen a fair bit of criticism and praise for the Hardfork and both are valid. Most of the people I've seen posting about it seem to be pretty much on the fence, I haven't seen anyone say it's the best thing since sliced bread nor have I seen anyone claim that it's worse than Hitler, so it seems to have had a fairly lukewarm reception.

Personally, as a new user, the Hardfork has definitely had a positive effect on my posts, it's also meant that my vote is actually worth something now, on a purely psychological basis this is good for giving new users the perception of being able to affect things on the site which should seriously help with retaining more new users which is definitely a positive for the site. But that's only my opinion on the matter, and I don't have enough information to be able to weigh in on the much wider economic debate regarding self up votes and such, I see that it's not a good situation but I can't say anymore than that.

I guess, as with all societies, new laws/rules come into play when the need arises, and there will always be those who take advantage of loopholes in these new laws/rules to the detriment of the wider community. But in the end laws/rules can be changed, and it's up to the members of the community to help initiate that change.

Anyway, I hope that made sense and didn't bore you too much, I just wanted to give my opinion as a new user.

The general reaction is extremely positive, especially from the minnows and newbies. I guess they aren't really aware of the problems and abuse happening elsewhere. Yes, a hardfork is exactly that, new laws, or an amendment to the constitution. There was one bad change here, and I hope they remedy that soon enough.

You are doing a great job.

I think the biggest problem here was mixing the 4x vote increase and linear rewards into one bag, and confounding the output result. Without a proper control, we can't really identify what is causing what effect. For example, minnows earning more: is this due to linear rewards giving minnows more power? Or is it due to minnows, whom often voted very seldom, getting a 4x increase in influence compared to the bots who were maximizing their influence?
Hard to tell -- though I imagine the linear reward is the bigger influencer for that example, you get my point that the outcome is mixed.

I was hopeful for a voting power increase to be a positive effect. I'm starting to watch how often self voting or collusive voting has an impact. It's possible that people are used to the 40 votes per day and are still in the mindset (hence seeing how many 25% votes are going out nowadays), and perhaps this will change over time as people get used to it.

One advantage I found is that I am able to give heavy influence to fund projects I created (steemcleaners/cheetah), that I otherwise wasn't able to. This is great, because I don't have to worry for example about paying to run cheetah (she's getting seriously expensive).
But I realize this opens another can of worms: if I'm the only one funding my project, is it still truly a community supported and driven project?

You could take the data and simulate it for different algorithms, and I hope the Steemit Inc team is doing just that.

With a quadratic curve and 4x increase, it could be hellishly abusive, with a whales having the potential to drain out obscene portions of the reward pool. That said, a whale's abuse is easier to detect and police than minnows self voting. On the other hand, the minnow self upvote doesn't cost the reward pool as much.

I have seen a lot of minnows curate actively, so I'm inclined to believe the linear rewards is a net positive.

Come to think of it, the linear rewards would probably be better with a very gentle vote power cost. That'll encourage newbies and engaged curators alike to curate more, and drown out the self-upvote cost. The whole thing about the "so less people have 100% VP" makes no sense to me. I was open minded about it, but I don't see any benefit whatsoever. Engaged curators are basically discouraged to keep curating, while casual self-upvoters who couldn't be bothered to curate others' post gain massively. I know the bots are an issue, but there can be a better solution for that. (PS: For example, voting power tanks after a certain threshold which could only be bots or crazed humans. Currently, the voting power cost actually slows the lower your VP gets! So, let's say, I make 100 votes at the old 40 per day limit, I'd end up with 60% VP or something. The next 100% vote currently would cost 0.3%, which is less than the 0.5% it would cost at 100% VP. I suggest after this threshold, the votes start costing more, and the VP tanks quickly. Of course, the Rshares will need to be decoupled from the VP at this point, so even if a vote consumes 2% VP, the influence given will be the usual 0.5%*VP. This threshold can easily be observed by looking at the data, and should be a weighted percentile of overall activity rather than a fixed number. Just some thoughts, I'm hardly a developer or economist.)

Either way, as a top witness I hope you take this matter very seriously and discuss it with fellow top witnesses and Steemit Inc developers. I hope to see a solution proposed in time for HF20.

One advantage I found is that I am able to give heavy influence to fund projects I created (steemcleaners/cheetah), that I otherwise wasn't able to. This is great, because I don't have to worry for example about paying to run cheetah (she's getting seriously expensive).

You have a very useful service, but other self-upvotes are not so valuable - and most are in fact detrimental - to the community. So, you should count this as a big problem even if you find it personally advantageous.

So there is corruption, even in utopia. Sad, but not insurmountable, I guess. It's a shame that corruption infiltrates nearly all aspects of our lives and has become so commonplace that we barely notice it's existence.