Bots gaming the curation rewards system along with “bandwagon” voting have sparked a lively debate on Steemit lately.
Many creative suggestions on how the curation system can be enhanced have ensued as a result. I acknowledge the posts and comments that have contributed to my thoughts on the matter at this stage.
Executive summary
My two Satoshi’s worth it that reputation (author), post value ($) and number of votes should be hidden during a “curation window period”. This will allow for a fair and neutral assessment of quality of the content without preconceived bias.
All (true curation) votes in the blind window period receive maximized rewards. The early details of how such a system could work follow below.
Premise
It is easy for a bot to determine the reputation of a particular Author based on previous content votes and value.
Even without using any automation it doesn’t take much observation to figure out which authors are popular. When authors of “high reputation” post, it’s pretty much a forgone conclusion that the post will be a high earner before the quality of the actual content is accessed.
Reputation is not a bad thing as some authors work hard to build reputation by consistently posting good content and should be rewarded for that. Good content is what the platform wants and reputation will likely continue to play an important part into the future as the platform grows.
Suggestion
A way to eliminate reputation based voting is to mask the author of the post for a period of time while the content is being evaluated. Similarly early bandwagon and bot voting can be reduced by masking the details of the votes received so far and the current $ value of the post during this period. This could be considered as a “content curation window period”.
Basically once the post is created these details are hidden for a couple of hours (maybe as many as 12 or more to enable participation from all time zones) with only a countdown counter displayed in the place of the usual information during the curation window.
During this period all votes cast receive maximum weighting as if they were a first vote under the current system (or other suitable mechanism). After the window closes, the author, $ value and votes are revealed and vote weighting can then continue as normal with the usual reductions based on the number of votes already received.
Bots, bandwaggoneers and reputation voters can then continue voting as they would normally to increase the value of the post, but the greater portion of the curation rewards could potentially be weighted to the voters that “discovered” the content and risked their votes in the blind period. (There are many ways to do this and the details would need to be debated and worked out)
Conclusion
The masking must be from the user interface as well as at the deeper levels. I understand it may or may not be architecturally difficult to implement but, this could render early curation more rewarding, slightly more risky, fun and fair.
This also potentially has the “jackpot effect” in that authors and curators have to wait a few hours to see the results of their activities, thus building a little extra anticipation.
These are early and incomplete thoughts… so your comments will be appreciated. Admittedly bots can use term searches, topics, concepts etc. and these still need to be fleshed out.