Overview of Vote distribution on Posts till Payout - Part 2 Daily and Hourly Charts

... Continued from Overview of Vote distribution on Posts till Payout - Part 1

In my previous article; we saw how an average post attracts votes over the seven days till its first payout. We observed that 69% of the total votes till the payout of an article was cast on the first day.

In this article we are going to look at the following:

• How an article gets votes on the first 24 hours of its publishing?
• Does it follow the Zipf's Law on the first day as in the previous article?
• How an article attracts votes in the first 60 minutes of publishing?
• We are also looking at how curators vote in the first hour when an article is published?

Since advent of the reverse-auction process in HF19; the first 30 minutes curation payout is given back to the author in a linear fashion. There will be minimal curation reward when you vote in that window.

Only the articles that were published in the last month are taken into account for this analysis.

Tools used:

• Blockchain data is sourced from steemsql.com by @arcange.
• The data is compiled and charted on Google Sheets - an excellent online alternative to MS Excel.

The Method:

• The category of the user is derived from the Accounts table.
• The post made by the user and it's payout date is derived from the Comments table.
• The votes made on the posts are taken from the TxVotes table. Only the upvotes are taken into account.
• We are getting the number of votes on a post and grouping it by hour or minutes as necessary.
• For post decay chart, the category is matched with the author of the post.
• For the curator chart, the category is matched with the voter of the post.
• There is a significant reduction in the curation of an article after 35 minutes of its publishing.

The Results:

Votes during the first day of publication:

• Planktons get 30% of their total votes in the first hour of publishing. This is way below the other categories who get around 52% of votes in this time.
• Overall, 49% of votes are made during the day are in the first haour of publishing.
• Voting declines to 8% in the second hour and deteriorates further in correlation with the Zipf's law.

Votes during the first hour of publication:

• There's a massive spike in voting during the first 10 minutes for a Plankton's
  votes indicating the voting bots' activity.
• Minutes 14 to 36 is the zone of major voting but the Planktons trail others suring this period.

Curation during the first hour of publication:

• The categories indicate the curators not authors as in the previous charts.
• The major presence of Planktons and Minnows in the first 5 minutes indicate that they vote as soon as a post is published probably unaware of the reverse auction process literally giving away their curation rewards.
• The major spike in Dolphins around the 30 minute mark indicates the work of expert curators.
• As seen in the previous chart, curators are busy voting posts that are within 14 to 36 minutes of publishing.
• This range could also indicate the automatic voting and trails.
• If you want to know more about how a curator votes, check out how a Top Curator votes in the first hour.

In short, the first hour of a post is it's most important one deciding its fate in this blockchain. Get more rewards in the first hour and end up in the trending posts or fail to get enough votes and end up in the clutter of articles that have very less or no viewership.

Conclusion:

The rewards from a post is all in the hands of the curators who vote during the first hour of it's publishing as around 35% of total votes on a post are made during this period setting up the article's future.

I am a part of the Steemit Business Intelligence community. We all post under the tag #BIsteemit. If you have an analysis you would like carried out on Steemit data, please do contact me or any of the #bisteemit team and we will do our best to help you...

Join the fabulous Steemit Blogger Discord community:
https://discord.gg/e5XFZKa

Cheers and happy steeming,
dbdecoy