You are viewing a single comment's thread from:

RE: Some Thoughts on Steemit

in #steemit7 years ago

Good article. I agree that the voting system could indeed be more equitable, something it seems will be addressed in HF19. As to earnings, it seems to me that most of the big earners are the ones longest on the platform and producing consistent quality content, which is fair enough. As @kafkanarchy pointed out in his article today, he was earning only pennies when he began a year ago. I'm sure a year from now, if you keep going as you are, you'll be up there with the big boys.

Sort:  

Thanks for your substantive comment, and kind words. I expect I'll have to reiterate in every reply that I am not personally concerned with emuneration. I don't care much about money, and reckon I'll manage well enough to die old.

What I am discussing is that the concentration of wealth isn't self-correcting, which is why HF19, and that dramatic divergence is seemingly based more on gaming the platform to milk rewards than divergence in quality of content based on superior ability and experience.

It isn't mathematically sound to claim the divergence in rewards reflects divergence in ability. Many Steemers discuss this, from all wealth perspectives, and find it unfair. The perception of fairness is actually a lot more important than actual unfairness, because people will act, not based on reality, but rather their perceptions.

There are other pertinent factors, such as time in, which significantly affects following, and thus exposure, but following is also a commonly gamed feature, although not one I address.

All that being said, everybody finds more money welcome. I'm no different, just not very motivated thereby.

I wouldn't claim that 'divergence in rewards reflects the divergence in ability'. What I meant to suggest was not that all ‘able’ posts were well rewarded, but rather that, for the most part, the well-rewarded posts were ‘able’.

I guess it’s like life, you can have all the ability in the world but remain unnoticed and unrewarded.

I don’t know or understand HF19 well enough to debate it but got the idea that it would reduce the influence of the whales and increases that of the minnows. As you said, it is the perception of fairness which is important, as it is upon this that people act.

You are right. Quality alone isn't what makes a thing popular, nor all that adds value. Neither is value the only criteria that determines rewards, as the reach of the author plays a huge part, as well as other factors.

Some of these aren't problems I have tried to solve, or think can be solved.

But, in the final analysis, you are right. It is perceptions that inform belief, and belief impels our acts.

This is the crux, and I replied to your well thought out post in Jerry Banfield's thread concerning this sort of exploitation.

The concept that we should recognize and reward Quality posts is a morally good idea imo.
The premise that we can identify and properly evaluate Quality is flawed.

Quality is subjective, not objective. If we want to try and define Quality objectively, we're immediately treading on a slippery slope of conformity, which is precisely what a block chain media outlet should never be. The very idea of sucking up to whale's should be a clear indicator of this phenomenon.

I think if we're going to try and build an algorithm to represent Quality in a universal way, then we had better make it very scalable on many factors that are not easily exploitable in unison, nor heavily weighted individually. For instance, if we want to say that high VP should be to some degree related to high SP, then we take a tiny % of that and apply it to our Quality algorithm, rather than holistically represent it in this manner. It also needs to be on a logarithmic curve with a soft cap, much like many RPG XP systems. This is becoming verbal diarrhea, so I'll try to express it in a simple example formula:

Quality = %TotalActiveMembers * log(#Comments/(%TotalViews) * (UpVoteValuation/DownVoteValuation);

Where UpVoteValuation & DownVoteValuation = Sum(Vote * (log(SP)) * (log(Reputation)) )i to n; where i to n is individual votes
*** This is a very rough example, but I think you get the idea

Despite my personal vow to let my VP recharge, I was unable to not toss a vote for your substantive and EXTREMELY informative comment.

Thanks!

Although I don't write code, I expect I could if I set my hand to try. I used to write websites on notepad.

Before XHTML and Javascript required me to actually crack a book to figger out how to do it anymore =p

But I think I follow the example you gave, and i reckon it would be a good way to consider both SP and rep in weighting VP, in principle. Depending on the language (none of which I am familiar with) the syntax may vary, however, the essence of the formula seems sound.

Calibrae, a fork of Steemit I have had a look at, briefly, seems to be doing exactly that, and some other things, like eliminating SBD, changing the powerdown method and timeline, etc...

So, clearly your thoughts on the matter aren't so beyond the pale that others aren't thinking the very same thing. Also, your simple formula indicates that making such a change isn't very difficult.

So, again, Thanks!

Thanks for the vote but please don't feel the need to do so on my behalf. I'm not here for the money just like you (though it is always nice ofc). I'm here for the censorship free platform, and Steemit is still perfectly capable of this role. Your thoughts on this matter and reply means much more to me, so thanks for spending your time on this matter.

Calibrae, I don't remember exactly where, but I know it was you who mentioned Calibrae before and I intend to follow its progress on Github.

I wasn't even thinking of a language when I delivered that example. It is strictly a math formula at this point, and as I said it was rough, so likely not even a good formula to use. It is merely an example of the degree of complexity that would be required to represent Quality.

I'm sure you could get back into programming if you wanted, you've clearly got a rational mind and that's really all you need. The hardest part is getting started, but here is a really simple way to get off the ground:
https://www.tutorialspoint.com/codingground.htm

It is browser based IDE for a ton of diff languages. It avoids the messy starting setup for beginners, allows you to just dive right in. If you have any questions about code at any point, feel free to send them my way I'd be happy to help :).

Thanks! I shall have to consider that undertaking should opportunity arise. I am presently utterly overwhelmed with matters of pressing urgency.

Coin Marketplace

STEEM 0.16
TRX 0.17
JST 0.029
BTC 69212.15
ETH 2511.68
USDT 1.00
SBD 2.57