Decline of busy.org posts - Need power up for newbies!

Repository

https://github.com/busyorg/busy

Introduction

I think most users have ever used https://busy.org to get voting. I love Busy and also made some contributions to the source (see the end of the post).

As you may know, there are two voting accounts: @busy.org and @busy.pay. The voting weight is determined by your "influence", i.e., proportionate to the sum of your followers' (organic) SP. If the sum is above a certain threshold, you'll get voting from @busy.pay (about 500K SP), but from @busy.org (about 650 SP) otherwise.

In fact, many votes of busy.org are actually dust votes, which mean that they actually don't count at all, literally zero value! This might be too harsh, considering that users need to use 'busy' tag and busy.org site as well.

(If you're not familiar with dust vote, see my recent Utopian post [Analysis] May need vote dust threshold again for an introduction and recent trend.)

This analysis shows that the number of posts for busy.org is decreasing (of course relatively to the decrease of the entire posts). I also make some suggestions.

Scope

I analyze all posts from 9-28-2018 to 12-19-2018 that received voting from either busy.org or busy.pay. To show the relative trend, I also counted the number of all posts.

Results

Dust vote ratio of busy.org

num_post: number of posts that received voting from busy.org (Actual number can be higher, see Tools and Scripts section.)

1. Dust vote ratio is quite high (~8%).
2. Dust vote ratio is decreasing
3. Number of busy.org posts is also decreasing

• The initial glitch is due to HF20.

Actually 1) and 2) are a little bit unusual if you're not familiar with recent dust vote trend. For entire posts, dust vote ratio is in fact increasing as you can see below (blue).

[Analysis] May need vote dust threshold again by @blockchainstudio

One plausible reason behind this is that users don't like dust vote (who likes?), and 3) they stopped using it. (This can be actually shown by examining tag usage, but I think the current analysis is already enough to say the main message, so further investigation is left as an exercise :) Thus only users who can receive non-negligible voting keep using it, and therefore, 2) dust vote ratio is decreasing. But still 1) it (~8%) is much higher than average (~1%).

Decline of busy.org post

You may think previous "3) Number of busy.org posts is also decreasing" is just due to the decline of the entire posts. So I analyzed the data further to show that it is not.

• The declines of busy.pay posts and entire posts show pretty much same trend.
• But the decline of busy.org posts is much steeper than the others.

Fortunately, I'm getting voting from busy.pay. Thanks a lot. (I deserve it, or probably more :) I made 4 pull requests in the busy github and one was already merged (edit old posts). I have a couple of more bug fix/new features pending, but I don't wanna PR spamming :)

Many people who are eligible for busy.pay voting use it. And that's why the trend of busy.pay posts are pretty much same as the entire posts. But for busy.org, clearly the usage is declining probably due to too small vote, 8% of them are even zero! This is also not good for busy.org's curation reward.

Busy.org has 4,500 liquid STEEM in their account for a long time

As shown in the first figure, busy.org has 4,500 STEEM for a long time. 4,500 STEEM isn't big at all for entire Busy, so I think they may have even forgotten. If there is no plan to use it soon, why not powering up to give some actual reward for precious newbies these days. They're our future :) With more than 5K SP, at least there is no dust vote anymore.

Conclusion

Busy (https://busy.org) provides generous voting so that many people still have some incentive to write on Steemit. But some users (mostly newbies) receive too small votings, sometimes even zero worth. So busy.org postings are decreasing. If Busy powers busy.org up a little bit, many newbies will benefit from it.

Tools and Scripts

• SQL query (Note that I only used finalized voting data by using created < last_payout, so the actual number of posts can be more, as explained in [Analysis] May need vote dust threshold again This provides one good natural random sampling.)

SELECT created, TxVotes.author, TxVotes.permlink, TxVotes.weight, Comments.active_votes
FROM TxVotes JOIN Comments ON (TxVotes.author = Comments.author AND TxVotes.permlink = Comments.permlink)
WHERE TxVotes.voter = 'busy.org'
    AND DEPTH = 0
    AND created < last_payout
    AND created BETWEEN '2018-9-27' AND '2018-12-20'
ORDER BY created ASC

• Data process
  See [Analysis] May need vote dust threshold again by @blockchainstudio.

• Python script for graphs

fig, ax = plt.subplots(figsize=(14,7))
ax2 = ax.twinx()
rspine = ax2.spines['right']
rspine.set_position(('axes', 1.1))
ax2.set_frame_on(True)
ax2.patch.set_visible(False)
fig.subplots_adjust(right=0.7)

df['num_post'].plot(ax=ax, style='C1-')
df['num_paypost'].plot(ax=ax, style='C0-', secondary_y=True)
df['num_allpost'].plot(ax=ax2, style='C2-')

ax2.legend([ax.get_lines()[0], ax.right_ax.get_lines()[0], ax2.get_lines()[0]],
           ['busy.org','busy.pay','all'], bbox_to_anchor=(0.98, 0.98))
ax.set_xlabel('')
ax.set_ylabel('busy.org')
ax.right_ax.set_ylabel('busy.pay')
ax2.set_ylabel('all')

Relevant Links and Resources

• 3 axis plot: https://stackoverflow.com/questions/11640243/pandas-plot-multiple-y-axes)
• [Analysis] May need vote dust threshold again by @blockchainstudio

Here is my recent Utopian contribution for Busy projects

• busy feed/blog/replies/follow bugs due to API no longer supported
• [Bug Fix - Merged and Live!] Finally, Busy can edit posts older than 7 days!
• [Bug fix] Incorrect vote value due to incorrect voting power

and 3 pull requests, waiting for review. (My Github: https://github.com/economicstudio)

ps. Special thanks to Busy team: @ekitcho, @sekhmet, @jm90mm (let me know if there are more, I found you guys through github)