Just as easy access to vote bots change the dynamics of how people act, easy access to highly profitable curation programs will also change how people act. Investors will go where the money is. Currently the incentives in place encourage voting with bots, regardless of the content quality. Incentivizing curation incentivizes valuing content that others will also value (i.e. "good" content).
You are viewing a single comment's thread from:
No, it also incentivizes voting with bots, just @nfc type bots. Even the curation increase with HF20 has done that. I've been playing around with some of my own. It's not about content quality, it's just about vote-racing, mostly in front of who Steemit is voting by proxy through the @misterdelegation recipients. Which makes sense, since curation has never effectively promoted content discovery.
Also, it's pretty clear that the peak of "quality content" on Steem is content that is perceived to promote the Steem price. If you reduce the diversity of content by concentrating votes you're just hastening the Bitconnect-ization of Steem.
The only thing in this system that has ever been about content quality is downvoting. That's why reducing the cost of downvoting is essential to getting the rewards where they do the most good.
This is a package of changes and looking at them piecemeal is misleading or in some cases deliberate spin:
Looking at the pieces separately doesn't show the whole picture.
This is all focusing on abuse rather than usability, which is one of the Steem development process' biggest flaws. Yes, fewer votes might go to crap, and if that's your goal, bravo.
But the cost is fewer votes going to new posters. Something that we didn't start with enough of and have been dropping like flies.
Our argument here seems to be about whether promoting user growth is more important than reducing abuse. I strongly believe that to be true.
It's not focusing on abuse. It is focusing on content-agnostic voting for the purpose of generating rewards vs. expressing an opinion. Downvoting doesn't generate/direct rewards so it can freely be used by people to express an opinion without huge economic distortions in favor of reward extraction. That's why it is so critical here. Once we get people expressing their opinions (if mostly via "I object!") without an easy option to say "To me!" then we can see rewards flow to where stakeholders as a community engaged in proof of brain actually believe they should flow.
Curation is about finding and surfacing these posts to both help the economic system work as well as make the platform more useful as content discovery. But (as you correctly point out) it doesn't work without anchoring based on actual opinions of the content itself.
There is nothing about this discussion that involves fewer votes going to new posters. Indeed by shifting rewards away from massive schemes that gobble up a huge portion of them it leaves MORE rewards for new posters.
I'm beginning to feel like nobody behind this scheme has done any observing of the accounts that are currently trying to maximize curation rewards. They don't vote on new posters. When I've been building curation bots I don't vote on new posters. Why would I? It's much easier to predict the value of votes going to established posters.
I think the downvoting part of this proposal makes a lot of sense, but I don't see why it makes any more sense at 50/50 than it does at 75/25 or even 100/0. All of the benefits of downvoting you predict would remain.
Yes but also much more crowded.
I ran major curation efforts early on. Our biggest wins (hundreds of STEEM of curation rewards) were literally new posters like neilstrauss and others like that. They were lightly voted or unvoted for hours because everyone was fighting over a few STEEM curating stellabelle or dan.
Curation is very, very underdeveloped because there isn't enough money in it, and it isn't seen as effective. Both issues are addressed here. Like you, most who worked on it earlier have given up.
I agree the benefits of downvoting remain but there has to be some sort of carrot to go with the stick otherwise the benefits of selfvoting are still alluring (some chance you won't get downvoted). We need some incentive to shift the use of stake toward useful curation and away from extraction. So for that reason I don't think 100/0 is very good. The other numbers could work. There is a lot of support behind the idea that 25% is just too low, and the gap of 4-1 between curating and self-voting is too high. But you are right the 'best' number isn't clear at all.
Maybe it's because there's nobody in the space, and figuring out the new evp is taking people some time, but right now curation is massively more profitable than delegating to vote-selling bots. I have one very simple bot that's returning 5x par. I wouldn't say I've given up, I'm just not particularly interested in maximizing my Steem holdings versus participating in a growing and diverse content system that empowers large numbers of people. That's where all of my SP goes - I'm leasing more for that than I'm using for curation experiments - and that's the mission that is going backwards with this change.
If there's no extrinsic value, it really doesn't make sense for me to spend my time here versus doing something that earns USD, regardless of how my rewards are distributed. 50% would definitely make Steem more profitable for me, but not really in a way that makes it make economic sense to spend my time here. What makes it make sense to be here is the mission of trying to build that diverse community of stakeholders, and the long-term benefit that may accrue from it.
And this is the problem we're trying to fix by incentivizing more curation so that it can effectively promote content discovery.
If we assume more people move to curation bots, would the quality go up, got down, or stay the same? Curation may not be perfect, but I do think it's better than what we have now.
"Quality content" is certainly skewed towards those with the most voting power and it makes sense they want to increase the value of their holdings. One important way to do that is to show the Steem blockchain via Steemit, Busy, Steempeak, Steemmonsters, etc working effectively. That includes demonstrating an effective curation system bringing good content to the surface. By that, I mean it's not just "Hurray STEEM!" articles that will, in the long run, have a positive impact on token price. It's also important (for those who understand and consider long-term value) to promote and demonstrate the effectiveness of the stated (and valuable to investor) goals of the Steem blockchain.
"It doesn't work, so let's increase its influence."
If you could come up with a curation algorithm that actually worked for content discovery, there might be some argument for increasing rewards. But the current one demonstrably does not do that.
Can you clarify what you mean by curation algorithm?
The argument behind this change is that the current curation system does not give enough reward to curators, therefore it is ineffective. Increasing the amount of the rewards pool given to curators is, the argument goes, the change to make it work.
The method currently in place to distribute rewards from voting is an algorithm which rewards predicting what other accounts will vote on, not one which rewards discovering content.
If the goal is driving content discovery, it's ineffective because it fundamentally doesn't work.
How do we measured valuable discovered content and who discovered it first? Even discovering is not enough as it has to be discovered and promoted by an influencer and even then it may not get much traction. Sometimes bad content goes further simply because of the platform built by the author. These are not simple problems to solve. If you have a better algorithm, is it published somewhere for review?
Nope. That's the Holy Grail; if we could actually do that Steem would gain immensely more value, and you can bet that if I had one I would never shut up about it.
But I don't drive faster on a flat tire just because I don't have a spare.
Higher curation, cheaper downvotes, and even mild/moderate superlinear (superlinear I don't support) are all part of changing the algorithm. They all shift economic incentives and the incentives are very much part of the algorithm. Without input from external actors (votes) the algorithm does nothing and those external actors act based on the incentives. You can't separate one from the other.
Most obviously cheaper downvotes are a big change because they dramatically change curation incentives toward voting for content that has a lower risk of being downvoted. This means the curator must consider factors predictive of downvotes and/or net votes (including valuation and perceived value) and not just factors affecting upvotes (visibility, misterdelegation vote targets, paid votes, etc.)
You've made a pretty strong argument for an independent downvoting system, something I already thought was a good idea but a low priority; you've convinced me it's more important than I was giving it credit for.
But nothing you've argued has made me think that its benefits won't accrue just as well under 75/25 as they would under 50/50.
Fair enough. I agreed that outside of 100/0 we can't really make any strong absolute claims about what are the right numbers. We can make relative claims about shifting incentives in a direction (which is the reason for most of the support behind 75/25 -> 50/50), but that's not quite the same thing.
So given that and the concerns about "omnibus bill" forking, why not try independent flagging first and see how it works?
Curation bots is too generic to reach any conclusion. It depends how those bots are programmed, what inputs they receive, etc.
We can reason based on economic incentives and those incentives apply equally to humans and bots (created by humans based on the incentives). We can say with near certainty that changing the incentives will change the behavior. It is more difficult to predict exactly how the behavior will change, but we can still make some educated guesses (and accept that despite doing so we will sometimes be wrong).
Agreed. If more people move to curation for the rewards then it's feasible more people will move to curation bots. There are a lot of people in financially struggling areas posting multiple times a day, trying to scrape some rewards together. It's feasible to consider a good portion of them will instead curate content for rewards and, hopefully, hook up with SP holders to help curate for them. Maybe tools like https://wise.vote/ will help as well. I did an interview with Noisy a while back about Wise. More tools like that could bring the perfect balance of human automation and incentive / reward sharing.
Was not aware of wise but think you for pointing out it. That looks great and there can be all sorts of tools for different curation models including delegation, collaboration, and human-machine hybrids. Excellent points.