You are viewing a single comment's thread from:

RE: Roadmap for Hive-related work by BlockTrades in the next 6 months

in HiveDevs5 months ago

I think we need to clarify on terminology first. We've never been time-linear in terms of rewards. Early voters always got more curation rewards than later voters. The planned change will make that not true during the window. I think some form of window here is valuable, because it makes some sense to reward early curators for doing the work of discovering content for the later votes.

However, you may be referring to a truly linear reward curve proportional to the amount of stake voted vs the change done by Vandeberg that nerfs the rewards for very small votes, making a quasi-linear curve. I've never investigated that issue much and therefore don't have much of an opinion on it , but I know it's been unpopular with some people. I'm not opposed to having a discussion about it (or maybe just a pointer to past posts on the subject), if this is the reversion you're suggesting. I suspect such a reversion wouldn't be difficult, so it's probably mostly a matter of weighing the pros and cons.

Sort:  

I am for a 2-24 hour window as you suggested in the post. At first I was at 24 hours because you can wake up and find new content everyday and be in the rush to vote. However, I do like the idea of 2 hours, as you said it creates a bit of competition and 2 hours is still plenty of time for manual curators. Maybe 2-3 hours can be game theorized and see if it makes much difference.

As for the reversion back from penalizing small votes, (comments got hurt the most by this) I am for this. I think we should try to boost engagement at all times and most users simply stopped voting comments because they can get better ROI (by a long shot) curating and voting bigger on post.

Why are we still trying to game theory the votes?? We have to accept there will always be side effects we can't predict before implementing the incentive change.

Let's make the window at least 24hours, I can't log in every 2 hours, it's insane.

A 2 hour window will force me to vote posts that were posted in the last 2 hours (even if I don't like them!), and will prevent me from upvoting posts that I enjoy but were posted 5 hours ago.

Hive is a global network and people are not awake at the same time. Content creators in different time zones will lose votes simply because of the time at which they post.

@blocktrades @theycallmedan

Why are we still trying to game theory the votes?

Because that's the smart thing to do when you create an economic system.

We have to accept there will always be side effects we can't predict before implementing the incentive change.

I don't think this is true. The possibility of making a mistake does not mean that mistakes are inevitable. And to the extent that the consequences are not always easy to predict, it still doesn't lead me to the conclusion that changes should not be made. IMO, this is clearly a change for the better.

You don't need to login every 2 hours. As mentioned elsewhere, you just need to find some good posts to vote on that have been posted in the last two hours. Given sufficient quality posts and reasonable tools for helping you find such posts, that shouldn't be too difficult, but to the extent that it is, the people doing such work deserve more rewards.

Content creators in different time zones will lose votes simply because of the time at which they post.

Auto-vote trailing of manual curators will likely negate any impact of that. In fact, it's reasonable to assume that some auto-vote bots would intentionally target posts made during "off hours" for more votes, because they would then benefit from manual voters finding the posts later after the two hour window.

Thanks for the input on the penalizing of small votes. What you said jibes with what I've heard in the past, although I wasn't aware this also reduced voting on comments.

If I recall correctly, the original motivation was to make it difficult to do some sort of large scale, low value self-voting. But I may have that wrong, and even if that's the reason it was introduced, I'm not sure it solved any real problem. I'll try to dig up the reasons, and see if there are any compelling reasons to stick with it. But my first guess is there aren't any.

I don't think changing the window does much to change human behavior to be honest. The bots just get set to whatever that window is in order to maximize. Getting rid of the window entirely and moving to straight linear is what LEO has done and it makes sense on a lot of levels. One of the biggest being that it keeps things simple for new users. One of the biggest problems with all of this is in terms of attracting and retaining users is how complex it all is for people outside of crypto.

That being said, it's never a good idea to make changes just because they are more "simple". This change also allows users to vote on what they actually like instead of worrying that they will be the last vote and lose curation etc, which is more or less what we are trying to accomplish with voting anyways, people voting on what they like. Also, agreeing with @theycallmedan above, yes the current voting system has killed comment voting as well as new/small users as they get 'penalized' until their vote is of sufficient size. I get the reasons behind some of these changes a while back in terms of trying to prevent abuse, but we have gone so far in trying to prevent abuse that we have killed on-boarding... which is the exact opposite of how/where we should be focused in my opinion...

The main change a longer flat window does is bring manual voters on par with auto voters. With the short time window, it was impossible to be a manual curator that earned as much as an auto voter. However, when all things are even, manual voters should earn more or at least more than they did before. The reason being if auto voters don't start following the manual curators (which dnst hurt the manual curator with a longer window) then they may start over rewarding content, which can lead to downvotes. How many times have you seen those embarrassing auto votes on a "This post has been deleted, Do not vote!" - ya well some people are willing to look bad for the increased rewards. So ya, all things equal, manual should earn more simply because they will be downvoted less because their votes will be strategic and realtime.

And curation isn't easy. I hire people to help with my curation, on top of spending hours a day doing it myself. When I see I earn on avg less than 10%, while auto voters earning upwards of 15%+, the investor inside me cringes. These people are doing no work at all, I'm coming out of pocket and spending countless hours and getting less. That is not an optimal system by any means.

maybe this would change with content + advertisement. So holder should only reward good content that other people from search engines also looking for ( for revenue).

Commercials with burn rate should profit every holder long term. It can be also flat 50/50 so the game is only looking for the best revenue content.

The change really did affect comment voting, although there are a few who 'waste' their VP rewarding comments as a way to try to boost engagement on their posts, and others.

Personally, I think that as long as there is a window, there will be 'curation rewards over content' automated voting. I still automate voting on the LEO tribe following the change to no window - I can follow the tribes curation account with being heavily penalized for following a large vote, 'late'. I can see people doing this with the base token if there was no window - following curation trails like curie, ocdb, etc.

There is a concern with regards to late self-voting though, but some of the tribes (LEO, STEM, SPORTS) have already done away with the window and at present are not suffering from late self-voting. The reasons for this are due to a) larger stakeholders knowing that self-voting is like crapping in their own kitchen and b) methods being introduced to allow more time to downvote than upvote. I worked with the SPORTS tribe after they introduced 3 days to upvote, and the full 7 days to downvote. It's not too difficult to run scripts to check who's been heavily (with stake or numbers) self-voting comments (either with same account or an alt), and having a few days after the last upvote can be cast allows time to check for 'abuse'.

Thanks as always for your work on Hive!

There will always be auto-voting as long as there are curation rewards, IMO, because not every stakeholder will want to curate. Accepting that premise, I think the proposed solution will at least direct that auto-vote towards more deserving posts.

I agree. Following a respected trail with an auto-vote knowing you wont be penalized feels more acceptable and better than the network over auto-voting an author who is known to collect decent rewards just because of the network they have. A removal of the any window could steer more to the former.

In my opinion, the "Rewarding Curve" is one of the most painful measures adopted in the past and the most discouraging in terms of engagement promotion. As @theycallmedan said to boost engagement is a MUST do.

ok some stupid question from me, what is with flat rewards 50/50 no curve? Or curve begins after day 3 so it's not last minute voting for rewards.