Why n^1.3,? That seems like an arbitrary magic number. Why not n^2? The only thing I know is Steem isn't working. It seems entirely broken now while in the past it seemed inefficient but not entirely broken. It used to be that once in a while I would discover really interesting content but this happens less and less often.
You are viewing a single comment's thread from:
It's arbitrary to an extent, maybe 1.2 or 1.25 is ok too, maybe we need to go higher
The benefits of superlinear in a nutshell: forces all profitable voting behavior into the light (wack a mole with 1 hole), disincnetivize profit based spamming and micro voting, make it more difficult for bid bots to accurate place a price on votes and hence increase the cost of content indifferent behavior, helps make good curation more competitive in terms of returns. It basically helps prevent certain abuse and effectively reflects the wisdom of the crowds.
But n^2 is far too extreme. There are detriments to these measures. If superlinear is too high it entices collusion between whales to pile onto a certain posts to exploit exponentially larger rewards. Someone with 10x your SP at n^2 means their vote counts for 100x more starting from 0, so it's highly unfair and favors larger voters too much.
At n^1.3 (maybe with a linear tail like the nike sign), someone with 10x your SP has a vote around 20x. It's sort of a balance that hopefully provides most of the benefits of n^2 at a fraction of the cost in inequality. Exploitation is further deterred by the 10% free downvotes (which of course itself can lead to toxic behavior too).
That's what I mean by using the least 'damaging' set of measures that is sufficient to break the current economic equilibrium of content indifferent voting. All these measures have downsides, we use as soft a touch as possible that is still strong enough to break down the 4x advantage that bad voting currently has over good voting.
What about creating a test system, using genetic programming, to find the best possible number ? Yes I know you'll think I'm crazy. Maybe I get a chance to explain at Steemfest3 if you'll be there
With @steemflagrewards supported well enough you can get a return on downvoting, no need to change anything except the non-use of the bot.
This keeps a human on the paid for downvoting switch, too.
The key phrase which stood out from everything you said. You should do a blog post just on this phrase. How can we find ways to increase the cost of content indifferent behavior?
Because you are good at communicating this, we need more blog posts so people can grasp what you're saying. I haven't studied this enough to have much ideas to contribute yet but I'm beginning to think about it.
I agree, @kevinwong and @trafalgar should start pushing this initiative forward together and gather support, I personally am willing to support only those witnesses that support these changes. This madness has been going on for far too long.
8 months ago I went on a tirade about how Steem needs to adapt to maximize utility. At the time I wrote the post I did not have a plan for how Steem could do it. I knew Steem was broken even back then.
Please note that I'm not an economist. I haven't studied economics. I know only the basics. But from what I do know about these basics the core of capitalism, of business, is the satisfaction of the customer. If the users of your product are not satisfied then you have to change your product.
So if the economics aren't working at all toward maximization of utility then the economics have to be changed. @trafalgar has presented a list of changes worth exploring. I think @kevinwong has also suggested a similar list of changes. I think we should consider these changes and what is there to lose by implementing them?
At this point we have a broken incentive structure. It's seemingly not working for anyone. Even if these new ideas make it slightly less broken it is still an improvement over completely broken.
"basics the core of capitalism, of business, is the satisfaction of the customer."
Yes, we have to ask ourselves who is the happy customer in our current ecosystem? Is it the content creators who are almost forced to buy votes to get noticed at all, curators who have to go through lot of promoted junk, just normal everyday users who want to find good content but are having trouble doing so or passive SP holders who just want to see their wealth increase with minimal effort possible.
I'd like to see ecosystem where I could actually spend time looking for great authors, reward and interact with them, without feeling that I'm losing out a big time on a rat race to increase our Steem Power or drain fiat out of the system.
Eventually even if we can't get this change through, I hope to see other frontends as well offer an option to not be blasted by promoted posts only on trending.Ned said frontends could do UI change, and that's actually something I've been doing with my own little project at https://steemliber.com (work in progress), where I hide all the promoted posts by default in trending feed.
With n2 rewards, downvoting a post with 25% of it's upvote weight would halve it's payout = cheap downvotes.
At superlinear, discounted downvoting is built in; which was the whole point in the first place.
50% curation, n2 rewards.
All the stuff that wasn't broken originally, but got fixed anyway.
We don't need to fix the incentives, we need to unbreak them.
I'm concerned that any superlinear reward scheme will shift the reward balance even more towards the biggest whales and bid-bot-runners, it also shifts the balance further towards the more popular authors and away from the less-known authors.
Even now I'm often finding myself in a curationists dilemma - should I participate in the Keynesian beauty contest and give my upvotes to the most popular authors (maximizing my curation rewards), or should I send some rewards to some less-known people that actually posts good-quality-content?
I do have people in my feed that earns significant on each post they make, no matter the quality of the post. I have unfollowed one of them because I found myself quite disgusted of the high reward/quality he got on his posts (and, of course also because I didn't see much quality in his posts). The best way to gain curation rewards is to place carefully timed auto-votes on the most popular authors.