Repository
https://github.com/steemit/steem
I actually checked every steemian's voting data to see when people would vote. Considering the fact that the data is vast and the voting rules have changed since the hard fork 20, it is necessary to set clear criteria, so the following criteria are applied.
It is targeted for voting from October 3, 18, to November 14, 18, which began to be influenced by hardfork20. It will also only use the votes cast in posts prior to the payout.
So, we exclude votes that were made after a payout, or that were made in comments, not posts.
Aim of Analysis
The percentage of votes received within one hour of posting during the payout period
The percentage of votes received within 24 hours of posting during the payout period
Percentage of votes received per day after posting during the payout period
Conclusion
1. The percentage of votes received within one hour of posting
How many votes will be made in an hour after posting?
As you can see in the chart above, there was a lot of voting within 1 hour after posting, including about 9.3% (based on number of votes) between 15 minutes and 20 minutes.
The accumulation is as follows.
Within 15 minutes, 18.9% of all voters have already voted, and the proportion of cumulative votes within one hour is a whopping 52.3% of the total number of votes in seven days.
In other words, if you only look at the number of votes, about 52.3% of the total will be done within one hour of posting.
The above table shows the number of votes and the proportion of each segment.
2. The percentage of votes received within 24 hours of posting
It is somewhat surprising that the number of votes cast within one hour of posting is a whopping 52.3% of the total votes cast to payout.
So what is the percentage of voting on the first posting day?
Compared to the number of intense votes in the first hour, it looks quite reduced.
However, the votes between 1 hour and 24 hours after posting seem to be distributed evenly.
The accumulation is as follows.
On the aspect of number of votes, 83.6% of the votes were cast within 24 hours.
So, between 1 hour and 24 hours after posting, about 31.3% voted.
Detailed figures are shown in the table below.
3. Percentage of votes received per day after posting
Let's check the voting pattern for seven days after posting.
As you can see in the chart above, 83.6% are already voted on the first day of posting and 6.7% are voted on the second day. Approximately 90.3% of votes were completed on the first day and the second day after posting.
3.5% on the third day and 2.3% on the fourth day.
From the 5th to the 7th day, only a very small number of votes were made at 1.2 ~ 1.4% level.
The accumulation is as follows.
Detailed figures are shown in the table below.
4. Conclusion
Within one hour after posting, about 52.3% of the votes were concentrated, and about 31.3% of votes were made until 24 hours after that.
If you look only at the number of votes, you can see that 83.6% of voting is done on the first day of posting, and most of the postings on this standard are close to one day.
Of course, because the number of curation bots who vote very little early may be affected, I think voting concentration will be more relaxed by the value of the vote, not the number of votes.
However, the number of votes in 15 minutes is only 18.9%, so the influence of curation bots will not be large.
Let's look at each day.
83.6% on the first day after posting, 90.3% cumulative by day 2, and 96.1% by day 4, which means that all votes are almost complete.
Do you feel that the overall voting data is similar to what you have experienced?
I've seen occasional discussions to reduce the payout period. In fact, the voting power recharge time, the period to recall delegation is 5 days instead of 7 days.
What do you think about the payout period?
The Data and Queries
I did this analysis by connecting to the @steemsql db with MSSQL client(Microsoft SQL server management studio)
Refer to My Github