Steem: An attempt to show improved distribution to content authors over the past year
A look at reward distribution between authors and the participation in content production over the past year.
tl;dr: Good news, I think!
The initial idea of this data grab, based on numerous requests, was to take a look at pre and post hard fork author rewards to see if the EIP had changed the distribution in any way.
Before I get to the data, a reminder of what is stated in the Steem whitepaper:
The actual distribution will depend upon the voting patterns of users, but we suspect that the vast majority of the rewards will be distributed to the most popular content.
... if we order some large collection by size or popularity, the second element in the collection will be about half the measure of the first one, the third one will be about one-third the measure of the first one, and so on.
... if we have a million items, then the most popular 100 will contribute a third of the total value, the next 10,000 another third, and the remaining 989,900 the final third.
The quotes and chart above are all taken from page 13 of the Steem Whitepaper.
My short interpretation of the above is that it is expected that the vast majority of content producers will earn little compared to the few at the top. I've included the above to make clear that this is the game and it's unlikely that however we choose to slice the cake, a short head and long tail is still likely to exist.
Comparing author rewards over the past year
To take a shot at this, I've picked 5 the following data ranges and summed up each authors vesting rewards. The liquid rewards (and curation rewards) are ignored in this post.
01/25/2019 00:00:00 - 01/31/2019 23:59:59 -- last week of January 2019 04/24/2019 00:00:00 - 04/30/2019 23:59:59 -- last week of April 2019 07/25/2019 00:00:00 - 07/31/2019 23:59:59 -- last week of July 2019 10/25/2019 00:00:00 - 10/31/2019 23:59:59 -- last week of October 2019 01/25/2020 00:00:00 - 01/31/2020 23:59:59 -- last week of January 2020
First, a chart that has no right in any publication due to its lack of clarity (you may disagree!) - The vested author rewards for week ending 31st January 2019.
The label on the chart is the total vests for the week above for author 25 - who by my reckoning is somewhere around the end of the head and the start of the tail? The chart for week ending 31st January 2020 shows little difference to the naked eye, and so I've grouped authoring accounts into the buckets below to produce the next display.
Top 10, top 20, top 100, top 500.
Taking the data from Jan 2019 and Jan 2020, the percentage of total vesting rewards has reduced across all the buckets.
The most notable dip in percentage came in the first data range which followed the EIP fork.
The percentage of vests earned by the top 10 and top 20 authors has picked up again in the most recent data set.
Removing @burnpost
The reason the final statement above is true is likely due to the rise in popularity of @burnpost, which was easily the largest receiver of rewards at the end of January 2020.
All rewards received by @burnpost are eventually burned and so although these vests are not available for authors, I thought it was worth (you may think otherwise) running the numbers again without this account included.
w/b - without @burnpost
Excluding @burnpost, the percentage of vests going to the top 10 authors free to powerdown and cash-out vesting rewards has almost halved.
Whether the removal of @burnpost is meaningful or not is open to debate - the Vests are unavailable - would these vests have previously have been (evenly) spread across authors?
Back to the original data (with @burnpost) and the quote from the whitepaper...
... if we have a million items, then the most popular 100 will contribute a third of the total value, the next 10,000 another third, and the remaining 989,900 the final third.
We don't have a million accounts earning vesting rewards (contributing value?) and .0001% (100/1,000,000) of 6326 (accounts) is just over half an account, and so taking that route doesn't really tell much.
Looking at the total number of vests rewarded and the total number of unique authors receiving them does seem to show a wider distribution (more top authors sharing the first 3rd of value).
The top 165 accounts took (or shared) 1/3 of the author vests in the final week of January 2019, 48 accounts more than in the final week of January 2019.
The number of authors sharing the 'middle 3rd' of the VESTS in January 2020 was less than the number in January 2019, however as percentages, which accounts for the total number of unique authors in each data set receiving vests:
Based on the above, distribution to authors has improved a little?
Cheers
Asher
(10% to @burnpost for the name-dropping extravaganza)
