Stopping Spam on Steemit without Hurting Minnows - Equilibrium Scores

I've been reflecting on this since reading @dragosroua's post about the upcoming transition from bandwidth to Resource Credits - RCs.

I noticed from reading @flauwy's post on SMT coins available now that SteemHunt had solved their user behavior problem by disincentivizing behavior that would undermine the platoform's ability to be fair (and accurately reflect value contributed).

In their case, the problem behavior to be prevented was voting for one's own product repeatedly. So for them they introduced a Diversity Score which looks at the distribution of votes and multiplies that by your otherwise earned scores, to increase or decrease your earning velocity according to how "pro community" your upvotes are.

I realized that Steemit couldn't imitate the exact same thing, simply because here we've got problems at both ends...

Meaning, we have both spammers who want to comment on lots and lots and lots of posts. And we have selfish whales who only want to upvote themselves or maybe a very small pool of other whales who can upvote them back powerfully. Both are influences that undermine the longterm health of the platform, so we want to discourage (or at least reduce the effectiveness of) both.

Unfortunately, these aren't just two poles on the same spectrum. That would be easy. We could just introduce a sort of Diversity Score that looked for a median instead of high or low rates of board/narrow voting.

But we have ranges to factor in that span both comment activity and upvote activity. We have the broad distribution of comments needing to be discouraged on one spectrum, and a narrow range of upvoting to be discouraged on another.

So this is a little messy as a solution, but I think still on the right track.

Let's come up with an Equilibrium Score (ES) that factors both an idealized mid-range band of activity when it comes to commenting AND a very high score when it comes to the diversity of accounts upvoted. This composite number is then multiplied by SP earned before actual SP is awarded.

Examples

So if you had a lot of comments you left on many different people's posts, say 500 per day, which I just don't believe anyone could read, then you'd have an ES of less than 1 unless you were really upvoting a lot of people. This pretty much rules out spammers, because they comment a lot but upvote very little.

We need to also eliminate giving % of a vote under 100% until your upvote is worth at least a full .02. We just need to make sure that the diversity score number for upvotes that is idealized, is within the reach of a plankton upvoting at 100% 10 times per day. I know the Steemit platform already doesn't allow it until 500 SP, but other STEEM publishing platforms/plugins do, and this introduces a vulnerability for spammers to simply start upvoting everyone they comment spam. So it needs to be prevented at the STEEM level.

As another example, let's say the person comments a reasonable amount, but only ever upvotes 5 people. Well this person would also wind up with an ES of less than 1, and have their SP reduced before being awarded. The amount of the reduction would depend on the severity of the out of equilibrium behavior.

In cases where someone is commenting on lots of people, but also upvoting a very diverse array of people, say splitting their 70 weekly 100% upvotes among more than 50 posters, which still let's them have a handful of people they adore that they are upvoting every single day, this person would not have their SP award reduced by their ES. In fact, depending on how much their commenting and voting promotes equilibrium on the platform, they might even have their SP award increased.

Benefits Above Other Approaches

Other approaches I've read all penalize plankton and minnow. Basically, if you don't have a lot of SP, you are really handicapped in ever getting there.

This isn't about creating quality content. I'm saying that even if you are consistently creating quality content, if you don't have enough SP from the start, you won't be able to differentiate yourself from spammers. So all the "remedies" the platform/coin introduces to try to fight spam, will kneecap you too.

But this equilibrium measure being added in allows a minnow/plankton to differentiate themselves from spammers, and so not be penalized with them. While simultaneously addressing the issue of excessive self/small-circle upvoting by large stakeholders, who really aren't reading and/or commenting on the platform.

What Do You Think?

Would you support the introduction of Equilibrium Scores?

Anyway, these are my random thoughts. I don't know how the code behind this blockchain works, so don't know if this is feasible. I also might have a flaw in my logic, which if you see, please let me know.

However perfect or imperfect this attempt may be, I do think we need to start thinking in terms of new metrics that can be introduced that give us the data we need to make Steemit a little spot of fairness within an unfair world.

(Photo Source: Pixabay)

Resteems always appreciated