I mean, let's say I know you are going to vote anonymint with 80% chance, then my risk is too low. He'll get my upvote quickly to front-run "smooth the whale". Now repeat this with every other whale and their favorite authors, or even comment authors, and you are suddenly milking the curation through automation via a robotic circle-jerk. Humans are pretty "low entropy" in their behavior and it's not gonna get better anytime soon.
Having said that, we are far from being in a totally broken system.
This is why I vote people like heiditravels 30mins to an hr after she blogs, there's never a shortage of whales on those. It's amazing how many people do the same.
I agree with iamnotback that it's hard to form a community when whale votes are the only ones that really count. Every other vote is almost negligible
Although I agree with @smooth that the rewards system does find a way to pay out
some of the authors who can consistently deliver reasonable quality content, the problem for Steem is not whether the content is quality or not. IMO, this is entirely underappreciating that the value of social networks is not the content, but rather the connections in the network (and the impact that has on viral adoption and ecosystem network effects).
To be more precise, curation is only rewarded if the groupthink recognizes the blog posts. And so all the incentives in the system are designed around curating content that
will appeal widely, not specifically to any sub-group.
The rewards system is creating a rigor mortis of structure that forms a groupthink that discourages the formation of more connections in sub-groups. From my analysis,
I think this can't be fixed in any way through voting. It is fundamental and as I wrote way upthread weeks ago, they won't be able to fix this without entirely changing the reasons to own STEEM POWER (thus violating the current vested interests) and thus I think it is impossible for Steem to be fixed (which some people have noted in some aspects in their replies
when Dan mentioned removing curator rewards).
You may argue that multiple genres can be popular simultaneously, but the economic force is encouraging centralization and this can't be solved
without discarding the concept of voting from a shared debasement pool...
Yeah that is my point, it is benefiting me too much, because that huge burst of votes is getting me always into the attention zone where my blogs can at least make it to $300 with one minor whale or a few dolphins. And then reasonably good chance a more significant whale will vote because I am writing about important tech stuff they can appreciate.
So it creating a groupthink. They aren't making any curation decision. They are just trusting my skills are reasonably consistent, per my track record.
We can't have relevance without thought. And we can't have successful site without relevance.
My analysis is everything is out-of-whack on Steem and it will fail. I am now about 95% sure of it. Next stage is for me to decide if it can realistically be fixed.
I can't detail all my reasons. I made a short list today upthread. I am very sleepy right now so I can't post very coherently until after I awake.
The reason it can't be fixed through voting is because the only way to prevent voting from being gameable by collusion groups, is to have the whales there to stomp on any such attempts. But the whales can't be a spontaneous proliferation of sub-communities. I explained this in my recent blog:
https://steemit.com/steem/@anonymint/blog-rewards-can-t-be-widely-distributedAnd in my numerous comments on @dantheman's blog post:
https://steemit.com/steem/@dantheman/people-rank-using-page-rank-algorithm-for-better-curation-and-rewards#@anonymint/re-dantheman-people-rank-using-page-rank-algorithm-for-better-curation-and-rewards-20160811t031807328zAlso read the comment trail here:
https://steemit.com/steemit/@chitty/whale-s-dilemma#@dantheman/re-chitty-whale-s-dilemma-20160812t155654929zIf the content starts to suck, you will lose your audience and stop getting votes, with the same incentives working in reverse. The later votes will disappear and then the earlier votes (which exist because of the later votes) will unwind too.
This, frankly, is exactly what people want in a content source.
The content doesn't suck when the metric is overall groupthink approval, but it sucks because it wasn't targeted to any following. And that is absolutely essential. Make sure you understand this. Because this is my secret weapon that I will use to compete.
You can try creating a site where content is completely blind without any labeling of authorship or brand and instead every post has to be individually evaluated in a vacuum, but I'm virtually certain it will fail (except perhaps due to its novelty value as a sort of game).
Of course I am not going to tell anyone now the algorithms I have in mind, but let me just say I solved the Grassmannian problem
that @complexring spoke of.
The problem with the algorithm I had presented is that it had no reference point and everything was relative to everything else, so this superimposed it on a manifold of itself. This is why it isn't stable. We obfuscated with the relativity that we really have insufficient information to solve (all of externalities) the variables.
And I need to entirely get rid of voting. Everything will be entirely determined automatically from user actions (which can't be Sybil attacked in my formulation).
This is going to be deep. And sorry for talking about vaporware, but you said can't. And I think can. I'll try to speed up so we can publish specifics so you aren't debating againts something unspecified. At least you have my blog post as a rough idea of where I am headed.
But the economic incentives in my design will be much more attuned to the fact that 1/1000 users are good bloggers. I have an incentive for the other 999 which they can do.
Also my model is not specific to blogging but works with any content publishing and distribution, such as music promotion.
- No one is investing in creating communities (plural) because there is no monetary incentive to plant a homestead on Steem. Rather everyone is extracting from the groupthink. The exodus out of SP may have already started as I've seen some of the star ladies taking money out. Everybody is gaming the system, there is minimal investment into it. @smooth even you gamed it by mining it (no offense intended and thanks for sharing the love a bit)
I think a lot of that is the organization of the web site. There is nothng at all like a subreddit, personal or group home page, group feature with memberships, or other community features of any kind. The categories feature is underdeveloped and doesn't really work. In fact the whole web site seems like a afterthought that is poorly developed (though I wouldn't take anything away from their building a reasonably decent web wallet and blockchain explorer for the content; that was a significant effort).
I don't see any of the upcoming competing sites doing anything any better though.
I also agree with the later comments about voter apathy.
For as long as they have the voting feature, it won't matter much how they change the UI because the voting feature will just be noise relative to the sub-community groups. It has no relevance at all. And economically everyone will still be competing for the groupthink popular vote. As I argued to you several times upthread, but you just don't seem to get my point, is that 1 + 1 = -1, not 2 in this case. The monetary incentive is creating an ecosystem design to do one thing well and nothing else.
You can't readily see the -1 effect, because what we see is a lot of reasonable quality posts at the top of the rewards ranking. But the -1 effect is in all the lost sub-communities that are not spawning.
What I am trying to solve in terms of automatically ranking content well for relevance is something none of the social networks have solved, except Google seems to have done a good job with search. I will use the knowledge of the HIVE to solve it, but viewing the HIVE as a collection of sub-groups, not of one-mind (popularity). Scaling the algorithm required using decentralization of computation+storage to solve it, because there is no way it would fit into DRAM on a server.
Even Medium is trying to move away from the long-form format, because they realize only 1/1000 can contribute and engage. They found that most of the 25 million readers don't even login.