RE: Latest Curation Reward Solution
I think the whole issue of changing the curation rewards is a matter of integrity.
First off, greed is a necessary in order to understand the underlying mechanism that fuels the Steem incentive system. If rules are put in place, and they are followed, which they were, then from a game theoretic perspective, one would expect a rational agent to do everything possible to maximize rewards. All agents (be they early adopters, cryptoheads, people new to the scene, etc.,) based their actions on a pre-defined set of rules.
Now, it turns out that due to the beta nature of this 'game,' and over the course of time by performing a live test on the system, we understand the current algorithm for distributing rewards for a given set of actions (posting, commenting, upvoting, etc.) much better. We see that the current implementation doesn't achieve what was initially intended.
The fact of the matter is that there is nothing faulty in the code, per se, and the current implementation is not, what I would consider, buggy. Rather, this is an an algorithm that attempts a particular type of distribution for the rewards and does so in a poor manner. In this sense, the algorithm failed, but it produced correct results as implemented. In addition, it is important to note that all participants knew the rules of the game in that the rules were in readily accessible places displayed on the Steem blockchain and strewn throughout the internet or in the Steem white paper.
Since the rules of the game were known before any agent decided to participate, actions and behavior occurred according to said rules. To change the rules retroactively is to penalize behavior that was deemed appropriate within the scopes of the rules at that time.
Changing the payout scheme and applying it retroactively to past behavior is no different than a government charging you with a crime after a law has been passed. Ignoring beliefs as to the role of government / nations (if any), established precedent exists in that most nations have ex post facto laws so that laws cannot apply retroactively to behavior that is deemed legal at a certain point of time.
I do not believe that anyone has an issue with changing how the rewards are calculated moving forward or if a future date (even possibly prior to the 4th of July large payout) is decided as to when one algorithm is used for the reward system and then the new schema is adopted. However, the problem lies in penalizing legitimate behavior retroactively.
Disclaimer: I would benefit from a retroactive change.
But weren't the rules said to be open to change?
I never saw that, at least until the recent round of posts started being made about shifting things around. I read the white paper, web site, bitcointalk posts, and no I never saw it. People who were encouraged to participate on the basis of rewards being displayed on the site in response to their actions were always told that the rewards would change based on activity (being divided among more posts for example) but I do not recall any suggestion that the rules governing those reward would be retroactively rewritten in a substantial way until quite recently. In fact, if I recall correctly, some earlier changes and proposals were specifically put forward as only applying to future actions.
I'm not sure if this was explicitly state anywhere. I thought it was implied, however, and that (maybe?) common sense would suggest that behavior would be rewarded based upon agreed rules that were put in place at the time said behavior was performed. It's not like anyone can go back and retroactively change behavior (voting patterns, post patterns, etc.).
I expected that to be obvious, I don't see any other reason for the delay of the payouts until July 4th. The algorithm was live tested, and flaws can be resolved before it affects the intended goal of distribution.
There are at least two very good reasons to delay the payouts. One being the possibility of obvious bugs in new code. If, for example, it turned out that rewards were being generated at twice the specified rate, credited to the wrong account, etc. then it would make perfect sense to fix the bug before doing the payout.
The second reason is that the reward fund started accruing well before the web site was open for posting, and certainly before it had many users. If the payouts were made on the normal 24-hour schedule the first few users on the site the first day would have gotten huge rewards. It makes more sense for those initial rewards to be spread out over a few months of usage.
Yeah, makes sense.
Seems like we're a bit in the DAO dilemma - is it code or intention that counts, and if the latter, whose? I for my part expected a curve like the new one for the first payout, and wouldn't have participated at all if I could've forseen the old one. I don't fully understand the changes we see here, why I'm up so much and you and nextgen losses are so unproportionally high. But I wouldn't care when I'd be on the other side of the fence, because I want that wide distribution that has been announced at the beginning of beta to happen. That's what I bought into.
https://coinreport.net/conversation-ned-scott-ceo-steemit/
I agree with @pharesim that this is similar to a DAO dilemma (I would've posted directly under his post, but too many descendants from the parents' post).
However, what is different is not so much the intention-- I would argue that the DAO fiasco was due to a bug in the contract and someone understanding the nature of how that contract was written using his / her expertise as a programmer. But with this situation, what we have was a complete lack of data about **HOW** the algorithm would perform. I find it unjust to reward behavior based on rules that were not known while actions were being performed, even if the new rules achieve the intention of a much more even reward distribution after the data has come in.