Ranking the Top Curators on Hive | Number of votes, authors, vote diversification | The reward goes to ….

in #curators4 months ago (edited)

Curators function is essential for the blockchain since it is the method used to reward authors and distribute the inflation. The bigger the curator stake the larger the votes/rewards.

Add a heading.png

Here we will be looking into the top 20 curators. The ones with the highest voting power and the highest earnings. How are they doing in distributing their voting power. How many votes they made? How many authors they reward? Do they distribute the rewards equally to those authors?

The data presented here is for the period April 1 till April 27, 2020.

Top 20 Curators by earnings

Here is the table for the top 20 curators by earnings.

RankCuratorHP EarnedCurator HP [Apr 27, 2020]
1appreciator74,9177,460,000
2rocky129,4902,650,000
3upmewhale29,2112,900,000
4theycallmedan23,8602,690,000
5blocktrades20,2003,340,000
6bdvoter18,3431,720,000
7tipu16,9821,410,000
8buildawhale14,6031,380,000
9darthknight12,4852,580,000
10ocdb11,9551,400,000
11curangel11,4891,330,000
12trafalgar10,598979,517
13sweetsssj8,2701,040,000
14threespeak8,1271,080,000
15likwid8,021869,173
16themarkymark7,970546,029
17therealwolf7,205778,433
18tribesteemup7,136846,000
19steemmonsters5,849722,000
20xeldal5,0171,220,000


For reference I have added the curators HP that they had on April 27, 2020.

Disclaimer: The earnings are approximate and not totally accurate mostly because of the VESTS to HIVE ratio that is changed daily and I have a fix ratio for the period.

The @appreciator account is standing out with almost 75k HP earned in the period and almost 7.5M HP. Next are @rocky1, @upmewhale and @theycallmedan.
We can notice that the rank is not as the same as the HP that the accounts have.


Now that we have the top curators by earning let’s see how are they voting.

Number of votes from the top 20 curators

Here is the chart.

02.png

By this metric @tipu comes first with more than 4k votes distributed in the period. Next are @therealwolf and @curangel with around 2.5k votes.

Number of unique accounts voted from the top curators

The number of votes is not as important if those are given out to the same account.
Here is the chart for the number of unique accounts voted from the top curators.

03.png

@tipu comes first by this metric as well with 1249 unique accounts voted in the period. Next are @therealwolf and @curangel with just above 1000 accounts voted. The difference here is not as big as for the number of votes.

Votes Diversification

We have seen the number of accounts voted from the top curators. But is this metric enough? Some can vote a few accounts with large votes, and hundreds of others with tiny votes and the number of accounts voted will be large. But this is not telling the whole story.

To check how are the top curators casting their votes between the accounts voted I have calculated the standard deviation in the rewards given out from each of the top curation.

Wikipedia definition for standard deviation:
In statistics, the standard deviation is a measure of the amount of variation or dispersion of a set of values. A low standard deviation indicates that the values tend to be close to the mean (also called the expected value) of the set, while a high standard deviation indicates that the values are spread out over a wider range.

What this basically says is if the STDEV is low, the votes are distributed more equally, and if the STDEV is high, the votes are distributed unequally between the accounts. Low STDEV = good, high STDEV = bad.

Here is the table for the standard deviation of each curator.

RankCuratorAuthors Rewarded STDEV
1curangel0.08%
2ocdb0.15%
3appreciator0.17%
4tipu0.19%
5blocktrades0.20%
6therealwolf0.22%
7theycallmedan0.23%
8themarkymark0.34%
9buildawhale0.34%
10upmewhale0.37%
11bdvoter0.41%
12rocky10.44%
13trafalgar0.50%
14steemmonsters0.58%
15darthknight0.61%
16threespeak0.73%
17tribesteemup1.10%
18likwid1.15%
19sweetsssj1.58%
20xeldal3.51%

By this metric @curangel comes on top with a very low standard deviation of 0.08%, next is @ocdb with almost double the value of 0.15%, and then @appreciator.

When we use the standard deviation and normalize the number of accounts voted to the top value the final chart looks like this.

04.png

So, there you have it. The best curator on the Hive Blockchain @curangel! He has voted the largest number of accounts with the most equally rewards distributed to each of them.

@curangel is a community project run by @pharesim and a team of curators. There is a daily report on each of the votes given from the different curators. A dedicated web page was announced recently https://curangel.com/ for everyone that wants to delegate to the project. They are paying curation rewards to the delegators.

@tipu and @ocdb, another two community driven curation projects are on the second and third place. Great job from them as well.

Those on the bottom of the chart have some catching up to do 😊.


I find this topic quite interesting and valuable for the community and may dive to the curators with lower HP and do some ranking in that class in the future.

Hope this makes thing a bit more clear and it may serve as a guidance if everyone want to delegate there HP, earn some curation rewards from it, but still want to be sure that the HP is put in good use.

All the best
@dalz





Sort:  

Wow. That's even better numbers than I would have expected. Nice work curators!

They are frontrunners for a trail and a shower of votes follows them wherever they go.
They are not the top curators, but the top milkers.

@curangel doesn't front-run anything. It's actually the opposite of front-running, since new vote suggestions by curators are appended to a queue, and a vote only happens when the voting power of the account reaches 100%. Have a look at when @curangel votes on posts, and notice when those votes are happening in relation to the creation of that post. It's several hours afterward in most cases. Does this seem like milking behaviour to you?

There is a trail, yes. Every curation project has a trail of some sort. It's a way to entrust a project with your stake when you can't (or don't want to) curate as effectively yourself and you don't really care about maximising curation rewards. You can't even stop people from voting on a trail. Anyone can write a bot that follows votes it sees on the chain, and therefore, every account by definition has its own trail with at least 0 members.

I'd love to further discuss how you came to this conclusion that @curangel of all projects is milking. Certainly there are better examples of that?

I am tired of this discussion.

Just imagine you actually curated and voted on this platform for years and then you stumble upon a post titled like this.
I am not even talking about myself - I gave up on 'curation' a while ago.

This is just a cheap excuse for real curation and organic voting where popular material would automatically bubble to the top.

I understand your frustration. I'm also frustrated that curation is less organic than it should be. But it has gotten better from the era of bid bots. We have to consider that humans will be humans and we cannot go around expecting them to behave altruistically when behaving selfishly is strongly incentivised. That is an impossible dream, and we can either embrace that and try to iterate on the current system and build something a little bit better each time, approaching the world we would like to see... or we can give up like you say you did.

Which one is more productive?

I don't even understand what you think is particulary inorganic about @curangel's curation. It literally makes no sense. It's just like a big account doing its own human curation, except that there are several curators and not just one. Have you actually looked at the compilations and what is voted, or are you just damning the whole project because the votes come from a large account?

q morpheus: What if I told you ...

... that you are the one who gave up by embracing curangel.

I'd tell you that memes don't make myopic bullshit any less myopic.

So, who do you think benefits the most from curangel or ocd ?
Do independent curators, who vote the stuff they enjoy reading, benefit from curangel ?
Do influencers, who have a large audience (and seem to avoid Hive), benefit from curangel ?

Yep ... you are on the top :)
Congrats!

Low deviation isn't entirely good, it is actually a can be a sign of bad voting as much as good.

Curation snipers typically drop low votes on many authors for the best chance of curation rewards. Many of the curation snipers drop $0.00 votes to get 8x the curation reward of the typical user. It is fairly common to use the same weight votes (appreciator, for example, votes 9-12% 99% of the time, and curation snipers voting 0-1% most of the time with no specific favoritism as it is just sniping as many posts at ideal times as possible).

In the case of some of the largest "curators", it is a different problem, finding posts with little rewards and voting them very late to get near 100% curation. But it uses the same technique, large amount of similarly sized votes, none of them based on quality.

Both of these types of "curators" will perform well on these charts (and do).

In the case of @curangel, it is nearly impossible to be a sniper as it has a very delayed voting process. I think it is still only voting twice a day - but it could have changed. The posts are submitted by the community of curators and I am pretty sure it won't accept suggestions over a certain payout amount.

I understand that about Curangel, my point is correlation is not causation in this case.

Yep, that is fair - just pointing it out for others :)

Yes I know what you mean and you are right about this.

But in general having more accounts voted equaly, from the largest stakeholders is good for the distribution.

What is good curation can come on the same level of debate, as what is good content.

In my opinion the large stakeholders should play a totaly diferent game than the curation game :). Onboarding and development. That can increase their stake value much more than a few percent from the curation game.

The top 2 on this chart are not snipers, that I can tell for sure :)

In this list of curators, your personal account and buildawhale should not be even considered because you are also one of those curation snippers who vote others at 3-4 minutes to supposedly gain more curation on your personal acocunt followed by trail of buildawhale.
So in other sense, you just explained how you place those autovotes because they are popular authors.

I don't vote popular authors, a lot of the people I vote don't get many votes. I don't try to hunt anything that has no votes regardless of quality, I am very particular about who I vote for. If someone abuses my votes, I stop voting them.

You don't do anything yourself , its your curation snipe bot doing most of your automated snipping job for you.
Most of the cases I have seen you have removed your auto-votes are those who get sliglthly less popular later on so there are no more high curation rewards due to minimal automated efforts.
You can check how many popular and un-popular authors you have given automated votes to in last 8-10 hours.

You sound defensive.

If you want to turn this into a strawman argument, let's look at my last 3 votes and your last 3.

My votes:

Yours:

Notice a pattern?

I don't vote popular authors, a lot of the people I vote don't get many votes.

So you just proved my point to be correct and simply lied about yours.

Do you want to go meet at the bike rack after school or something?

Lol, not really. I have better things to do

Not sure where you got that, but ok if you say so.

I don't vote popular authors

I wish I have your vote. 😅

Exactly this. Quality curation will always mean lower diversification as quality by definition is rare.

76-85 is good curation. 100%? Imposible😀

Nah, you can do 800%. I was experimenting with an AI bot that was getting 150-800% even when curation was only 25%.

It is definitely possible to go over 100%.

Yes possible but if you know 100% the post will get big upvote. (Manual curation). Big whale never get big curation if curation manual. Often people get vote after 2 hour. The best curation 5minutes. That is only my philosophy😉😀

Curangel curators manually find the posts that are fresh, original and have almost no votes. The posts then are added to the queue for the Curangel account to vote throughout the day, so the voting is delayed in accordance to how many posts are in the queue. There is no way for curangel to snipe a post. Often times a lot of other groups have managed to vote the posts that have been in curangel's queue and as a result it votes after them.

I never said Curangel was sniping. I was saying that sniping behavior shows up as a "good curator" according to the above metrics.

Can you do a top 100 or even top 1,000?

T1000 is a bit too much :)

Thank you! That’s all I can come up with. My money is in the right place.

You are welcome :)

Hell Yah! I love me some curation and love being a part of Curangel. So many other great supportive communities out there and I can't wait to see the old bidbots engage the good community members and spread their votes among more accounts creating the good content.

All Hail Curators!

Well congrats to you as a curangel member!
On the old bidbods spreading their votes, I see we have some improvement already, but it can always be better :)

Amazing! @curators on #1? Wow!

It's nice seeing curators doing really hard to value others content. I think because @appreciator always go out and searching content. Many curators doing the same but I think it's because of their current power as well.

Stepping up the curation :)

Good job with this comparison @dalz and congrats @pharesim!
I'm glad to be part of @curangel. The power of Hive chain is in its users. There is no problem upvotes posts with coming huge rewards in the right time. But it is not case of @curangel, where we are looking for interesting, quality posts of various users.
I was suprised how many interesting posts are lost in the chain before I joined curangel ;)

Occasionally I open those daily reports from curangel and I'm surprised to see all the nice posts there.
Otherwise when I do the searching, its seems I cant find those!

Great job and hope you guys thrive even more.

Great work, I think focusing on the T20 makes this very accessible,

It'd be interesting to see the top 100 too, maybe just a supplementary chart at the end next time, or the T50 at the end?

I was just thinking that could reveal some of the old classic circle jerks - I know sweets is one of those - and what they're up to know.

I can't believe she's still got all that delegated HP, unless that's actually hers of course, she did earn a fair bit back in the day!

Yep I will try to dig in the lower HP holders next time :)

Although making a chart for a t100 would not be very readable. Maybe a table only.

Hey I think what you've got is the most accessible you can make it, I was just thinking something extra for those us that want to dig down a bit: a table would work just fine!

Keep up the excellent work!

I am happy to see Splinterlands (steemmonsters) up there! I am here curating your awesome post too, LOL! I love this, thanks for such great work! ~@clove71

Woot ... thanks a lot @clove71!

this is great work! 😁

Excellent the information you provide us dear friend @dalz, to be honest it does not surprise me to see the twenty main curators, they are excellent teams and people. congratulations to all especially
@curangel @apreciador @trafalgar @sweetsssj
I wish you all a great day

Hey @dalz, could you eloborate on this step?

When we use the standard deviation and normalize the number of accounts voted to the top value

I don't get what you are saying here.

Thanks for the interesting data, seems not much has changed...

Basicly I have multiplied the numbers of authors voted with:

1/(curatorX STDEV/best STDEV)

Really interesting. I'd love to learn how to maximize curation in some way. I see an account on here with a lot of power and have never posted on hive lol. To be honest I have been trying to figure out how much hive I need powered up to enjoy myself. So far I cant see any reason to have more than 1000 hp, but this has me thinking on the subject.

You are known for great statistics. Well done @dalz

I love manually curating on HIVE for Splinterlands / Steem Monsters!

You are great!

@dalz:

What do you think about giving us a small Dapp that calculates and graphs this Top 20 at any time?

It would be interesting if users could evaluate the performance of Hive curators.

That would require some time and skills that I dont have atm :)
But may get there at some point in the future.

I'm sure you can get someone who could help you with this.

Great job everyone! Curation is a main factor for our blockchain and all of them provided a big appreciation to the creators!

Btw: @diyhub didn't made so much HP because we can not compete with the big stake holders but we hit the 3000 upvotes yesterday :P I'm rly proud of our curation initiative which is fully non-profit :) which curation initiative pays out all curation rewards to the delegators? I think no one else ;) it would be great if you could measure the amount of upvotes next time too :)

Peace!

Thanks!

I'm thinking in digging into the lower HP holders and check those out.

This was a check for the T20 HP holders as they impact the platform the most, and I certanly agree that low HP holders can curate much better than the top ones.

Congrats to @curangel ,@pharesim - We need your trend to be followed by most to take this to beat BTC 😁

How would we go about finding the real heroes? - Those accounts who upvote (with clout behind them) in a truly diverse, and "demonstrably altruistic" fashion?

I guess each of us can point out and be grateful for the support he is receiving from some of those.

I dont see enough posts from individuals who are saying thank you to some of their supporters.

A large pool of data gathering and processing can be done, but I guess non of those analysis will be 100% right and they will miss something always.

I do not think that voting random posts can be considered curation. I checked the appreciator account, and I saw that usually it votes with high power (about $10-12) to posts that have just a few votes and there is no special quality to them. This is a kind of abuse and I think it gives a high percent of the big vote back to the voter.

Those are curation trail frontrunners and not the most successful curators.

With automatic curation like Steemauto?

Congratulations @dalz! You have completed the following achievement on the Hive blockchain and have been rewarded with new badge(s) :

You made more than 3000 comments. Your next target is to reach 3500 comments.

You can view your badges on your board and compare to others on the Ranking
If you no longer want to receive notifications, reply to this comment with the word STOP

Do not miss the last post from @hivebuzz:

HiveBuzz - Hive Gamification Experience
Vote for us as a witness to get one more badge and upvotes from us with more power!

@arcange Good to see you here, as haven't seen you around yet!

My pleasure and I think you are still on my witness list.
Glad that things will continue as before.
Blessings to your work.

Yes you are! thanks 🙂

Thank you @papilloncharity. I have been there, working in the background 😅

One day I shall one of the top 20 curators. There is a lot to admire here. Well done!

It pays to curate and you may not have to lift a letter, just click.
Everyone needs to pick their community to encourage #hiveans.

Is @curie not doing curation any more ? Strange to not see them in the list ?

Curie is very active, we just have less HP compared to these accounts. Our curation is very efficient though (just outside top 20), with large trail of followers. In the number of upvotes and votes diversification/distribution we are certainly among top 5.

They are not a t20 earner from curation reqards. Not sure how they operate, havent checked them in a while.

Because top curators has nothing to do with who earns the most curation rewards.. which is sort of the issue here and why analysis like this are comical.

You know, I may have replied in some thoughtful way to this, and ephasis that the fact that these are T20 curation earners, and it is important to check them out for transparency and for all of the community and add something more.

But when you come up and the first thing you do is calling someone else hard work that he has been doing for 3 to 4 days, comical, downvote it, it really kills any possibillity for any further discussion, and maybe finding some sort of important conclusion.

The downvote was for the, in my opinion, over reward placed By those mentioned in the list. It’s curation, I redistributed. If a downvote makes you not able to have a conversation, ok then. And I’m very familiar with those on the list.. my point was that was we call curation here is a bit skewed.

No I'm totally fine with the downvote :) Dont mind at all.
Your approach of starting a conversation is what I find wrong.

I didn’t say your post was comical, I said the fact that this is the method used to rank “top curators” (aka “best) is comical. As earning a shit ton of CR can be done by abusers and doesn’t mean they are helping the community or curating content, as it appears when displayed this way. I am saying the way we categorize curation is flawed.

At least this is way better than the bid bot model? new post are getting curated more than before. It's a good shift from what we have before. Certainly there's much more improvement to be done. But change of bid bot model towards curation sniping will contribute to better distribution of hive stake. Once hive stake is in equilibrium in this blockchain, hopefully we will no longer see only certain criteria of post will be curated. It s a positive change i would say. People need to see they earn more which is the main attraction why they're here.

Congrats @pharesim and the others that support our worthy cause.
You are champions!

Good Job curators!!!

Thank you

Nice numbers and nice work! Well rewarded :)

This is good compilation

Hello! This will be copy/pasted... but please pay attention to it... It's very important, I just need to spread the message quickly, before too many people are affected.

BITSHARES PROPOSAL SCAMS, BE CAREFUL!

There are people/a person, going around on bitshares and scamming people through proposals. I want to bring light to this as quickly as possible, for anyone who uses bts. It needs to be spread so that people are aware. The scammers seem to have already taken millions of bts and who knows how much more. :/

Thank you for taking the time to read this. It is very important to those who use BTS!

TL;DR: Yes, this is copy/pasted. It is to warn you about a scam going around. Please do not hate me for spamming this information around. :/ I am trying to possibly save people's crypto before it's too late!

Interesting statistical summary of healing; of these 20 curators, the community of quality content creators is largely sustained. I would like to take this opportunity to recognize and thank the expansive work that @appreciator does, which supports many authors regardless of whether or not they belong to the GEMS Community. This baby fish so far only receives significant support from @appreciator.

Thank you @bluemist, I am aware of the great work you are doing...

“Curators”