You are viewing a single comment's thread from:
RE: Benefits of Pure Linear Reward Distribution
it creates a huge incentive to vote for posts that are already the most popular.
With linear curation rewards plus linear content rewards, it's no incentive to vote for popular contents, nor disincentives though. You get same return no matter vote for what, the only factor is downvotes cast by others. Some people will downvote "overvalued" popular contents, so a better return is not guaranteed.
So the optimal strategy would be to vote for posts in the final seconds before they pay out, to minimize the chances of downvotes. That doesn't seem like something I'd want to incentivize.
Actually, the optimal strategy would be to cast a random downvote for every random upvote. That way you'd be pushing more rewards onto the posts that you're voting.
Good point. I have thought of this as well.
There is a review period in current system. During this period, post can only be downvoted. However, for better user experience (people who wants to upvote can upvote), it's only a few minutes now, IMHO it's too short. So I think we can set it longer, for example, one day, so good people have enough time to review and downvote overvalued contents. In regards to the user experience part, it's doable to just let people upvote, and only record the upvote, don't impact the payout value nor voting power.
I don't understand this. In my proposal downvotes will reduce total payout of a post. How can you get more rewards by doing so? If you cast a upvote and a same weighted downvote on same post, you won't increase total payout of the post, but will just inflate others' influence, so this strategy would be less profitable than upvoting good contents only.
Yeah, after thinking about it more, I now think I was wrong on that. :) You'd get more for your upvote somewhere else.
My thought was that I'd upvote Post A and downvote Post B to shift some of the rewards from B to A. But since there are many many posts in the world, the downvote wouldn't be an efficient use of voting power.
this is true.
The pure-linear approach work best in the situation where the community members are actively flagging those bad-quality posts. In a upvote-dominated system, where only a few minority doing policing job, then the flat reward-curve will just incentivize people to cast random votes.
so, linear-approach would work better if we could have a mechanism to compensate the down-voters. for e.g. maybe we could revise their voting power if a down-votes surpass some threshold (say, >50% of upvotes) so at least they wont be penalized on their voting power by doing good flagging job... it would be best if could come out with some rewarding scheme.
Good point. But keep in mind that downvotes can be abused as well, so hard to do the incentives right. As of now, I strongly tend to use an off-chain rewarding/compensating mechanism.
ok, i think i understand what youre saying now. and yeah youre right, with that you get a flat-fee curation reward (minus downvotes)... its exactly like always being the first to vote on something thats already after the 30 minute mark... you get a quarter of the value back that you added.
The only thing is, why bother to put any thought into your vote at all at that point.... why not just cast 40 random votes every 24 hours. If we're giving everyone the same reward regardless of the quality of their curation, why not simply build it into the system as SP incentives and have people vote for free.
It's still organic, right?
If you vote manually, at least you see the titles. When you saw the titles, you made judgement.
If you vote with a bot, either you follow authors or tags, or follow other voters, or judge by analyzing the contents, you've done your work/judgement.
If you make bad decisions, others will downvote, so you'll earn less.
Voting on popular contents and making them more popular is not fault. Actually it's why they're popular.
this is a good point
Seems like @abit won you over here! ☺️Is there anything you can think of that might be against this proposal? Just for completeness, I'm scratching my own head 🤔
nesting
Im kind of a cheap date when it comes to linearity. I'm not completely without reservation on making curation performance blind, but there doesnt seem to be an obvious GT problem with it...
There is the old argument that in a truly linear system, everyone will just vote for themselves, but IDK as i necessarily agree that this will happen. ANd as I point out in my 'opponent of the exponent' post, it seems likely that existing steemit institutions can handle potential abuse.
But if every vote pays out the same, why pay?
says the guy who makes more than everyone else
but yeah, this is my only beef with this idea. I think you need at least some front weighting to make people actually try.... though i think the current system has way too much.
"Making a content more popular" is the value later upvoters contributed, why not pay? They evaluate the results done by earlier voters, and confirm it by follow voting, done their work well.
Why can't there be an easy way to tell the community the best times to vote for good content? I have no idea how these whales are cheating the system and draining the pool.
Firstly, to be clear, whales are not cheating but playing by the (bad-designed) rule. Blame the rule, but not the people.
Secondly, for me, "the best time to vote" should be as simple as "when you read a post and think it's good". (Edit: the point is you shouldn't be punished when doing so)
theyre not really. And there isnt really a magic time.
Because author rewards are superlinear, and curation rewards are just a part of author rewards, that means that in order for there to be any significant curation rewards at all, a whale needs to vote for the post in question. And because there's a huge bonus for early voters over later voters, that means that to get a share you can't get in after the whale.
So who can vote on a post and be sure a whale will vote on the post? Well, a whale. For example, when abit votes on a post that has no votes yet (but has passed the 30 minute reverse auction period), he gets a 1.50 out of the six bucks his vote is worth, regardless of what else happens later. If more people vote for it, hell get more, but he can't get less. And even if he comes on to a post that already has a significant number of upvoters, just the weight he adds means that its probably rare for him to not get a significant piece of the pie for just voting.
In a way, what he proposes would make it a similar situation for anyone.
Now if you can predict the whales, then you could make money. The problem is that if youre predicting a whale vote from authors or topics the whale has already voted for, there are already bots doing that. and theyre taking most of the curation rewards (and giving them back to the author in the reverse auction)
Maybe I'm missing something but doesn't this mean that anyone, whale or not, can start upvoting random posts in order to just get the curation rewards?
As for people that will keep downvoting "overvalued" posts, I'm not so sure about that, I have a very specific "group of people" in my mind right now and I don't think that their agenda is for the good of steem.
Yes they can get the rewards, unless the posts they voted got downvoted by others.
I still believe we have more good people/stake than bad ones, so we can overcome that.
I fully agree with that one right now, but I'm not sure what will happen when we will start to attract the majority of "facebook users". I fear there will be a "free money" mentality and it will be in the shoulders of the few to correct the greed of the many.
Rewards are free money, but it is very diffcult to secure a result or predict it. Rewards really amount to the payment for the effort, in a subjective assessment by peers, of the results of the effort. A combination of diligence and luck. It's no less fair or entertaining than a game of Poker really. It even shares some features in common with poker in that the 'opponents' (your competitor curators and creator) decisions are only knowable after the transaction has been made. Predictable votes are easier to front-run.
It's a fair concern, but IMHO it's over-concerned right now.