Updates to Hive Curators Dashboard

Introduction
In this update I primarily wanted to answer a couple questions that my Top Six Hundred Curators Dashboard left unclear.
Number one question that I had was: How much of the total Hive Curation Rewards do the top six hundred curators represent? The answer turned out to be 95%
The second question I wanted to have answered is how do the curation rewards distributed among the top six hundred curators? As you can see from the two new charts pictured above the top 25 curators earned 61.48% of the curation rewards or 147,064 HP from the total of 239,199 HP earned by all 600 top curators. The chart below the first chart shows the distribution of rewards among various rank bands, top ten curators earned 105,673 HP, the next 15 earned 41,390 HP and so on.

Reward Distribution
Another interesting chart showing curation rewards distribution is the one above. As you can see the curve starts out really steep and then quickly flattens out. This means that curation power on Hive is highly concentrated among just a few top curators. Top ten curators earned 44.18% of the total curation rewards, top twenty five accounted for 61.48% of the rewards and top 52 accounted for 72.89% of curation rewards while top 100 accounted for 81.95% so the last 48 curators in the top 100 curators accounted for less than 10% of curation rewards.
Going up to top 253 curators the total share of the 253 curators is 93.22% of all curation rewards meaning that the other 347 top curators accounted for less than 7% of total curation rewards. And top 100 curators had about 82% while the additional 153 curators accounted for about 11%. It was interesting to see that while I am ranked at 272nd place and earned 87 HP in the last seven days, I am in the band beyond the first 253 curators in that really flat portion of the curve.

Top 600 Curator Rewards vs the Rest of the Hive Social Network Curation Rewards
I really wanted to see a visual representation of how much do the top 600 Hive curators represent in terms of the rest of Hive curators. The chart above clearly answers that question.

I also added the new seven tiles in the rows two and three pictured above. As we can see there were 8,944 total curators in the last seven days the total HP all curators earned was 251.87K HP from which 239.2K HP was earned by top 600 curators on this dashboard. So top six hundred curators on Hive accounted for 95% of all Hive curation rewards.
These are interesting numbers it shows how Hive inflation via curation is controlled mostly by top twenty five curators who control 61.48% of inflation and by the time we account for the top 600 curators there is only 5% of Hive curation rewards left to distribute among the remaining 8,344 curators!

I have also added a new chart that looks show the APR spread in each of the ranking segments. As you can see our top curators by the number of curation rewards have a very good curation efficiency as well as a pretty narrow APR spread. For example the top ten curators have a median efficiency of 8.51% and the spread is between 5.52% earned on curation and 10.11%
The next fifteen curators have a higher median efficiency of 8.58% and a spread that has a higher low of the band and lower high of the band at 6.68% and 9.77% curation APR earned. As we navigated down the ranking segments we see that Median APR tends to drop and the spread widen with the bottom 350 curators having the widest range of APRs between 0.64% and 15.07%

Another new chart lets you compare stake vs rewards, with bubble size showing payout count and color showing APR bucket. Interestingly higher number of payout events (or how many times the curator has voted) tends to correlate with higher efficiency at all staked Hive levels.

This other new chart is trying to answer the same question and provides a different view on the same subject matter.
There are a few more updates to the Top 600 Curator dashboard, so click here and check out the latest updates!
I am a Hive Witness and would really appreciate your vote for me as a witness: https://vote.hive.uno/@seattlea
If you prefer your Hive Witnesses to never power down I am one of those Hive Witnesses. I believe that unless you are a top 20 witness and running more than just a Hive Witness node then you don't need to power down.
"So top six hundred curators on Hive accounted for 95% of all Hive curation rewards."
Those are some eye opening stats!
Question, has it always been this way or has this stat crystalized as the overall user base has shrunk somewhat? O.o
That is an interesting question, I will try to look into the historical data to see if the stats really changed with the dropping userbase. Here is what happened to the said userbase in the last five years:

https://seattlea.z5.web.core.windows.net/weekly-authors-dashboard.html
I have looked into the data from 2018 and it appears that there is a slight trend towards a greater concentration of curation rewards over time, overall you can see that it "has always been this way"
The share of curation rewards for top 100 Hive curators never dipped below 70% since July 9, 2018... It did go up from low 70% to low 80% As you can see similar dynamics are at play when it comes to top 10 and top 20 curators. Top ten curators saw the greatest concentration of curation rewards from a low of around 24% in 2018 to the current 44%
Thanks for going the extra mile, appreciate the effort!
Interesting data, perhaps one way to look at it is that it points to a lack of balance or over concentration of influence at the top over time? Obviously only if said actors heart wasn't in the right place.. which we'd like to think is not the case.
Its a curious thing 2018 was a long time ago, would be a far harder thing to graph but based in your revelation, would be very interesting to learn how many user accounts have 'ascended' the ranks over that same time frame? I mean is there possibly some sort of imperceptable glass ceiling, or are the top curators simply too far ahead to ever catch, who knows..
I have also added a chart of top 600 vs the rest over time and it went up from 85% to 95%:

I have thought of a chart showing top curators over time... I might share that in the future if I am able to build it.
Fantastic job on today's Ecency leaderboard quest, @seattlea! This update provides excellent clarity regarding the distribution of curation rewards on Hive.
Tipped with Ecency POINTS.
I've been asking for this chart for 9+ years.
I've tried to tell people that things were bad, but they didn't want to hear it.
Well, now they can see it.
The top 20 have been taking 50% of the pool since at least the fork.
It may have been even worse before the fork.
It was definitely worse during the n2.
This is the reason that I have long called for a 1000mv soft cap on voting.
Not hard coded, but respected by those with more stake than that.
A 1000mv soft cap gives newbs that want to invest and 'try us out' a way to do so without getting reaped by those with 10x their stake.
As long as the top sucks out more than the crowd can survive on,we will always have a small crowd and a bad general reputation in the market.
Thanks for finally making these numbers accessible to the restivus.
Well, I am just starting on Hive analytics, more is coming... I think that we should have a view of what is happening on Hive right now and what it has been historically as well.
I'd like to see the curators rated on how much of their curation got dumped.
Seems like some of them want it to be dumped.
It would be easy to show how much they earned curating all time and what they have staked now, but if they transferred it to another account that would be harder to account for. I have a dashboard rating all witnesses based on KE that accounts for witness earnings and some look pretty bad...
You'd have to check it against known exchange accounts, but that would omit those that send it to another account before sending it to an exchange.
The k/e ratio of the accounts curated should do it well enough.
This isn't returning an expected result?
I think you are trying to search the weekly trends dashboard there are no curator or authors names there just the numbers of weekly active authors... Try this search on the Curator and Author dashboards...
The prompt says 'search by account', so I thought I would try it.
Good catch :) updated to # of authors so that it is not confusing :)
Oh I get it... that is probably harder...
I was thinking of building a dashboard that shows what accounts the top curators are voting for and then show the KE for those accounts...
That sounds close enough for me.
Interesting, but not surprising to me. With such a small userbase, the top 600 is quite a large group. I'm on the 380th spot when you look at staked Hive.
The top10 does play a big role though...
What is also interesting is that not all stakeholders vote frequently enough or efficiently enough to be ranked curation rewards wise similar to their staked Hive ranking.
If you scroll through the hivebuzz.me/ranking list you also see a lot of inactive accounts with a lot of HP. I always wonder what happened to those users. Would they know that they still have that amount of Hive staked they could set to use?
https://peakd.com/utopian-io/@paulag/the-impact-of-unused-steempower-on-the-rewards-pool-blockchain-business-intelligence
Begging for whale votes is cutting your own throat.
And, the throats of everybody that that whale's vote takes value from.
When whales vote everybody else looses money.
That is an interesting article...
I've been in this for a while now.
A 1000mv soft cap may not solve this issue, but it will certainly mitigate against it.
It levels the playing field for the late comers.
I have been trying to build a historical dashboard for top 100 curators and where their votes go and I can see why nobody built it yet... :) It is a lot of data mining and really intensive all around. Particularly enriching the data with KE scores of all the authors that were voted for...
This fella said the same thing.
I was thinking that haf would solve this.
https://peakd.com/steemit/@bitgeek/payout-stats-report-for-15th-december-2017--part-i
Well, I published the preview, this is still a work in progress and will take a while to complete... So far I was able to calculate top 300 most voted for authors, so you will see many rows with 0 values meaning that KE value is not enriched yet. I am running it now to enrich for the top 1000 voted for authors:
https://seattlea.z5.web.core.windows.net/curator-voting-history-dashboard.html
The k/e on the right is the consolidated score of who was voted?
Almost :) The higher the Weighted Author KE is for a curator is the more that curator is allocating their curation rewards towards accounts with higher KE ratios. Basically this measures how much a curator is encouraging extracting behaviors...
So a high number is bad?
Yes it is similar to KE but measures a different thing. The higher the number the more this curator has fed those that produced a lot of posts in the past five years and received a lot of author rewards and don't have much staked now.
A hover over popup for the column headings to explain what exactly the numbers are representing?
And, a donation button so folks can donate.
I have the column heading sorting and tool tips in the works as well :)
I am ok without a donation button, I just want upvotes and witness votes, this is not really a money motivated effort, I like dashboards and data :)
You are doing wonderful.
BTW, I added some of the tool tips sorting and now have KE for top 1000 authors up from 300 authors, so data is more complete...
https://peakd.com/hive/@khrom/why-downvotes-contradict-decentralization-a-mathematical-perspective
This one hits pretty hard, too.
Hmmm. I didn't know this. I still don't know how the basics of this blockchain work 😀
Probably should look this one over, too.
https://peakd.com/hive/@khrom/why-downvotes-contradict-decentralization-a-mathematical-perspective
Hi @seattlea,
Is it possible to view results from previous days?
I’m looking for last Sunday’s results. Unfortunately, I didn’t take a screenshot, but I have the “Most Efficient Curators” list. At that time, I was ranked 46th. I needed a screenshot to illustrate a post I’m writing, following up on this one, based on your Top 600 Curator Analytics tool.
Thanks 🤝
Well that particular chart was not designed to be anything but a snapshot of the last seven days updated every 4 hours. I will put into development pipeline. Making curation data from previous days would mean that I have to persist the data in JSON like I did for the massive curator history dataset. It is doable, but will probably have to be built as a going forward daily thing with a limit on data retention so that JSON doesn't grow too big.
You might be interested in a brand new dashboard that shows information on the top 100 curators for the past 5 years that I have just built and deployed today: https://seattlea.z5.web.core.windows.net/curator-voting-history-dashboard.html
It is kind of a beta version still...
Hmm... Okay. I see—it’ll be hard to get the image or chart for that exact day. No problem.
Since the results are organic and change every 4 hours, it might make it back onto the Top 52 list 👨🔧.
But for my post, the list I have is enough. That was the order on that day; I just don’t have the numbers for the other users, but that’s not relevant to what I want.
In that future post, I’ll tag you to send you an invitation or suggestion regarding your tool. In this excellent work you’re doing using the analytics tools, you’re only focusing on the Proof of Stake protocol (Earned, Reward, APR %, etc.). And I want to draw a parallel between yours and those from ManteCurated Live (Mantequila Soft), which incorporates the Proof of Brain protocol. After all, these two protocols are the essence—the narrative—of the HIVE blockchain.
As for the Top 100 list, I’m looking forward to seeing it. Although I’ve only been on HIVE for 2 years.
Thanks a lot 🤝