‘User Authority’ - A Viable Replacement for the 'Broken' Reputation Score?
by @revisesociology | August 8, 2018
There’s been a lot of discussion on steemit in the last week about a new ‘metric’ called ‘User Authority’ (UA), which is currently being developed by @scipio as a potential replacement to what many believe to be the broken ‘reputation’ score.
To simplify to the extreme, the User Authority Metric calculates a score based on the number of followers an individual has, rather than the number of upvotes an individual receives from other users. Just as with the current reputation score, the User Authority metric calculation is weighted…. If you are followed by a user with 10K followers, then your own UA is boosted more than if you are followed by someone with no followers.
It seems that UA offers a much better way to provide an indication of how popular someone really is, and how engaged they are with others. The theory is that in order to maintain followers, then you need to actually engage with them to keep them on board.
UA was initially conceived as a tool to improve the Utopian voting bot, and to solve the problem of SPAM - people generally don’t follow bot accounts, and if lots of bot accounts follow each other but are followed by no one else outside of that circle, they all end up with a low UA score (for the math, see the first and second links below).
Find out more about User Authority…
- This original post by @scipio (it’s quite technical!): How-to solve SPAM and Democratize Steem: Introducing UserAuthority.
- This is a more recent post in which @scipio discusses UA in more layman’s terms: UserAuthority (UA): explanations, applications and implications.
- @Cryptoctopus - User Authority - A Better Reputation System With Interesting Applications - a well thought out post on the possible advantages of introduced UA to the platform.
- The post which alerted me to the whole UA proposal in the first place: @abh12345’s Some thoughts on User Authority - check out the comments for the relentless positive reception of the concept.
I use the term ‘discussion’ above rather than ‘debate’ informedly as practically every comment or post on the topic has been for the User Authority being implemented in a future hard fork, so it seems its introduction to the platform is a matter of ‘when’ rather than ‘if.
Why we need UA: two obvious problems with the reputation score:
One problem with the ‘reputation’ score is that it can be so very easily gamed. Because you gain a greater relative increase in reputation score when you are upvoted by other high reputation users, all you need to do is buy a few whale upvotes and your reputation score increases very rapidly.
It is in fact possible to set up a new account today, post trash twice a day, use bid bots to gain high whale upvotes, and you could easily be into the early 60s for reputation within a few months.
A further problem is that if you were a relatively early adopter and gained a high reputation for posting low quality content without anyone else with a high reputation noticing, then it’s highly unlikely that your reputation is going to go down, even if you carry on publishing low quality content, because flags from lower rep users do not reduce your reputation. Haejin is the classic example here: he has a reputation of 81. Thoroughly undeserved.
Strengths and Applications of the Proposed User Authority Metric
If UA is adopted, and applied, there are many potential benefits:
Democracy It seems that this gives everyday users much more day to day democratic control over influencing the score compared to the current reputation score: all you need to do is to unfollow people in order to reduce their UA score.
Engagement It should encourage more community engagement…. The way to increase your own UA score is to gain more followers, although of course there are limits to how engaged you can actually be!
Quality It should enable discovery of quality content: generally speaking the better the quality of content the more followers. Those who turn out trash are more likely to be unfollowed.
SPAM identification….. There is the potential to filter out users with a low UA score, which spammers will invariably have as they will have very few followers.
There are other advantages to adopting UA, covered in the posts above, which I recommend.
What is striking is that there seem to be no theoretical disadvantages to the adoption of UA over the current reputation score. Granted, there are the usual practical problems of how to get it implemented and there will be potential debates about how extensively it might be used, but in theory at least it seems like a very good idea.
Conclusions
it remains to be seen whether or not User Authority will be adopted in a future hard fork. If it is adopted, it also remains to be seen how it will be applied by communities and on the blockchain as a whole.
In theory at least, it seems that this metric could help mitigate many of steemit’s problems.
Picture Sources
UA picture from @scipio
