You are viewing a single comment's thread from:

RE: Obstacles to Steem adoption

in #steem5 years ago

There are different reasons for lack of adoption versus those who sign up and then leave, and to be successful we will need to address all of these issues, but first they need to be acknowledged, and the current STEEM culture tends to reward a certain type of unquestioning optimism which results in self-righteous believers that cheer-lead about the current state of affairs, and so there is very little constructive criticism. Whenever someone brings up an issue, we can expect the status quo to loudly bark, "You're just doing it wrong!" As a wiser poster than me once said, software updates aimed at changing user behavior rarely work, but instead of listening to a multitude of users offering constructive criticism in the past, we have ignored them and continued to offer new users difficult signups, awkward user interfaces, user-unfriendly designs, a confusingly unfair reward algorithm, little rewards or interactions, and then we enabled more self-dealing developers and a higher ratio of whale shitpostings with the latest fork. Currently STEEM's optics look like it is designed to lure some sucker whales into a ponzi scheme, legitimized and funded by the labor of people producing blog content for pennies.

After waiting for a complicated signup process, new users are confronted with complicated mechanics to learn such as the SBD/STEEM currency system, several different passwords, unexplained voting power which is invisible on steemit.com, and primitive user interface options compared to twitter/reddit. Once they engage with the community, they will learn the in-group behavior, for example social cliques and whale tribes, result in echo chambers and circular voting. They will also eventually discover the unfair allocation of rewards by self-dealing developers and unnecessarily complex curation and total rewards algorithms that encourage automatic robot voting and small circular-voting whale tribes.

The only way STEEM will succeed is if an average person can have fun and make money. What is preventing an average person from having fun and making money here?

Sort:  

To clarify at least one of my complaints and offer a solution regarding unnecessarily complex and unfair curation reward distributions, I think a more successful model can be formed that is designed around two core principles, simplicity and fairness.

The most important aspect of the model is fairness, that all curation should receive the exact same ratio of return per invested STEEM. Currently the curation rewards are gamed in two ways, timing and total eventual rewards. The exact timing to optimize reward is based on some obscure algorithm that favors automated robot voting services, and it is easy to predict which whale accounts will inevitably receive the most rewards, so optimizing curation rewards is accomplished by automatically front-running predictable whale votes with an upvoting service, which is much different than the stated goal of supporting "higher quality posts and manual curating" that was used to promote the last fork. Also, dust votes are negated, which limits the earning potential of our smallest wallets, which deters new users. I understand there were some issues with Sybil attacks and dust votes being used to farm, but maybe instead of trying to control so much of the rewards pool with large wallets micromanaging everything and putting in tons of unpaid labor hours, we should create a reward system that is simple and intuitive to use, doesn't require so much labor, and automatically rewards everyone fairly according to their invested STEEM and minimum standard of participation. A simple and fair model would look much different than micromanaging 10 upvotes and 2.5 downvotes per day, without any knowledge or explanation of voting power for new users on steemit.com, based on some obscure curation timing algorithm front-running predictable whale votes, and there are too many different levers to adjust on the reward pool to be overly attached to any particular configuration, but I would prefer a more simplistic and fair reward system for curation that allows users more control over their voting power. I would like the option to use 1 vote per day instead of 10, and create automatic curation lists of my favorite authors that would automatically use my voting power in ratio to the number of posts by those authors at the end of each day, and automatically receive a fair linear curation reward regardless of when I voted on the post.

There is no totally fair system and what we have could work, after it was already tweaked. A lot of the issues are to do with the people. I do my best to support those who need it, but some others just care about profit when they are probably not desperate for money. I want more people to have the possibility to earn.

The current system is notably unfair in the way that curation returns are disproportionately awarded to the quickest automated bot voters instead of manual curators, which is the opposite result of the stated intention to change the complex curation reward algorithm during the last major fork to reward manual curation.

I was not in favour of reducing the voting period. They should do what they can to encourage manual voting. I don't fish for curation rewards.