Tracking Votes - What Is Happening On Steemit?
Let's Take A Look At Votes Since August 1, 2017
I had to pick a date, so I picked August 1. For no particular reason other than it fit well in a graph. There are 3 lines that you'll notice on the following graphs.
- Number of Votes on a given Day. (Note: these values are based a 24 hour day in Eastern Time [e.g. New York, USA])
- Linear Trendline (first for All Dates between August 1, 2017 and January 7, 2018, and second for the date range mentioned). This is a "Best Fit" Line for the corresponding date range.
- 30-Day-Moving-Average (for all dates between August 1, 2017 and January 7, 2018).
August 1, 2017 through January 7, 2018:
Looking at voting trends, it's fairly clear that votes were mostly stable but increasing slowly and steadily up through
Now, let's take a closer look on a month-by-month basis:
October 1, 2017 through November 1, 2017:
Solely looking at the overall trendline, you'd expect that we were increasing the number of votes executed per month, however, if you look deeper, there's a different story to be told. This is hinted at by the 30-day moving average peaking around October 18 and subsequently falling.
We saw a significant decline in the number of votes during the month of October, as you can see from the declining Trendline in the second graph. The number of votes between October 1 and November 1 trended downward by more than 100,000 votes for the month. Things seem to be looking up on November 1, but we'll look deeper into actual trends for November 2017 in the next section.
The 30-day moving average indicates that we're seeing a mostly flat voting month. The line moves upward and back downward again. Early in the month there appeared to be more voting activity, but by the middle of the month, all of that voting activity retreated and went into a negative trend through the end of the month (at least until November 1).
November 1, 2017 through December 1, 2017:
Looking at the graph above, it appears that voting is once again mostly flat for the month of November, with the number of votes on November 1 being slightly higher than the number of votes on December 1. The voting activity in the middle of the month is what saved the overall trend for the month to lean upward.
Despite the apparent surge in voting during the middle of the month (around Thanksgiving-time in the USA), the significant decline in number of votes on the 25th and 26th of this month were fairly drastic. Those low activity dates, along with the decrease in number of votes between November 1 and December 1 likely caused the downward-leaning trendline.
The 30-day moving average for November provides a similar story. The increase in votes between the 12th and 24th pushed the average upward, but the rest of the month's poor voting activity flattened out the gains into December from the earlier 12-day surge.
December 1, 2017 through January 1, 2018:
Despite the overall trendline appearing mostly flat in the graph above, you can notice a more volatile 30-day moving average as we get closer to the end of the month. Let's take a closer look in the next two charts.
Look at that linear growth! Starting around 390k votes per day and ending with a trend around 700k votes per day, looking only at the current month, is a significant increase! This was also around the time that Steem and SBD were growing astronomically in value! Do you think that might have been a coincidence? (I would say there is a fairly strong correlation!).
Even looking at a longer period (i.e. the 30-day moving average), you can see that the first few days of the month were trending negative or flat. Then the early-to-middle part of the month provided a steady increase, but by the 18th and 19th, you start to see a sharp incline in the number of votes per day! Could this be due to more people having time off around the holidays? Possibly. But again, my money is on the fact that Steem and SBD were soaring in price.
Steem and SBD reached their peak prices around the 21st and 22nd and you can see that people were taking advantage of these values over the following week (through the 29th), while prices remained higher than prior days.
What Do You Think?
Was voting increasing due to "seasonal volume", or were people voting more with the hopes of earning more curation rewards at higher values?? Or was there another force at work?