RE: My take on Self-voting, Vote-buying and Reward Pool Rape
I wrote 'non-linear' because there was some discussion around other 'curves' - sigmoid for example. I think there never was anything wrong with n^2.
I think you know I like the idea of a sigmoid reward curve as it is getting more flat in the end than n^2 which avoids extreme rewards, but anyway: you are completely right that the linear reward curve makes self-voting just too attractive and should be changed to anything non-linear again!
Even if it is not the exact topic of your article I should add that also the restriction to four fully rewarded posts per day to prevent (self-upvoted) spam posts was very reasonable in my opinion. Who is really writing more than four good posts per day? At least I could never achieve that ... (and yes, yes, I know: if I can't that doesn't mean nobody can ... but I just doubt that many can do that. :-) )
As @timcliff emphasized correctly in one of his recent comments most well intended measures can be circumvented with some efforts but I still think at least it shouldn't be made that easy to exploit the system ...
Sigmoid looks great on a first glance, but at which point should the slope start to decline ? ...
That huge problem aside, it would counter the idea of getting more influence over the rewards, the more popular a post gets.
The 4 post a day limit never applied to comments, so it is not even a question how to circumvent such a limitation for abuse.
Yes, this had to be discussed. I think one could reach a consensus here ...
One could actually also let the sigmoid curve become linear (after x has exceeded a certain value)! You see I am far from having an elaborated solution, but am contributing so many highly interesting thoughts, haha. :-) I am just an ordinary Steemit-user giving the implementation into the hands of professionals who are getting paid for doing so. :-))) (Read that with a twinkle in your eye ...:)
At a certain point, yes. However one could select kind of a 'stretched' sigmoid function, so that one could receive 'quite some' upvotes before the flattening takes place.
One could set a limit for the number of fully rewarded comments as well. In addition one could limit the maximum percentage of voting power which can be used to upvote comments. What about a restriction to (let's say) 15 % of the own voting power for comments?
I fear all those complicated solutions will get in the way of SMTs and other possible applications of STEEM.
Keep it simple.