RE: A new approach to Content Reward Allocation
It sounds too complicated.
My suggestion is to take a step back and simply let community dynamics play out for a couple of weeks - groups tend to alter behavior automatically to resolve issues without the "system" having to be changed.
Here is a comment I wrote on another thread which was complaining about Ned upvoting noobs:
It is better that the rewards get spread about rather than just concentrated on a few.
Suppose the voting bots are designed to look at your blog posts, assess the average earnings and then compile a list of those with an average above a certain amount. They then automatically vote for any new posts people on the list make.
If the "usual suspects" are no longer getting the rewards and the new people are, then the lists start to change. In fact if the whales spread their love far and wide enough, it will become impossible for a bot to tell from the previous average, which will be the new high paying posts. The owners will be forced to log in and read - because if the rewards are genuinely going to those with great content you should be able to find this out only by reading...
And lo and behold Steemit's problems are solved!
Agreed. 5 star for me complicated too. Most people vote 5 or 1. But if you should thing that average would be then 4 would be good guess if bot voting based on reputation.
We are dealing with unorganic behavior driven by bots. Its a question of incentives that rewards people to upvite before the bots...even if they didnt read the articles.
You miss the point - the bots only upvote because they believe the whales will. I saw a post today that had 31 votes but it only had a value of $4, because the whales were concentrating on noobs. So any bot voting that post in expectation it would perform the way previous ones by that writer did, would have been disappointed.
If the distribution is spread out far and wide, it becomes harder for a bot to predict, and will require the owner of the account to actually log in and read, to predict the value.
A agree with @alyssas but I'd add that it is all part of system dynamics. As the user base grows beyond the extreme early adopters, whales, and steemit insiders that have dominated it so far, the results will change, including the behavior and success rate of bots.
@dantheman's proposal here is good in the sense that it tries to take a fresh look at the problem and improve the algorithm overall, rather than being as clearly a knee-jerk reaction to current dynamics as was the earlier proposal to simply remove curation rewards. Still I think there is a lot of value in letting the system grow and seeing how things actually work as it grows, rather than jumping to hasty conclusions.
I want to see @dan make at least 1 shitpost and laugh at the bots upvoting him $6k. :)
It will reduce number bot that of votes for him. It will correlate with his reputation.
You say that like it's a bad thing. ;)
I see bots blindly voting for Dan the same way I see people who praise every word that comes out of a president's mouth.
President : "We tortured some folks".
Sycophants : "UPVOTE!!!! OMG!!! He's so cool!!!"
Except here it's even worse in my opinion, because the sycophant follows up with, "how much money did I just make?!".
;)
It is good thing that bots will stop voting if author with reputation start posting junk. Such author just trading his reputation for the money directly.
I just don't see a big problem with bots. They are measure instrument for reputation and individual network capacity of individual.
This only happens because everyone in the community knows Dan. It's like how Mark Zuckerberg at one time had us all on his friend list.
I think you might be right. Maybe a wait and see approach is best. It's still very early and not enough time is being given to the current algorithm to find out if it works or not when it scales. With enough people it might work better but with 10 people or 1000 people it might not work well.