A unique perspective on Hard Fork 25

in LeoFinance3 years ago (edited)

image.png

A lot of new changes are coming with Hard Fork 25, but the one that most people will likely be talking about is the changes to curation rewards.

Hard Fork 21 brought some of the most significant changes with the EIP (Economic Improvement Proposal). The EIP was a collection of three core changes to the rewards system.

  • New author reward curve (convergent curve)
  • Reduce author rewards to 50%
  • Introduce downvote mana pool

Many users were happy with these changes, many were not. One thing everyone can agree on is the changes removed all incentives to purchase votes. Free downvotes, although most people never used them, discouraged the use of bid bots to buy votes on posts. There are still some edge cases of this behavior but it has mostly been eliminated. As someone who ran a bid bot, I still supported this change.

What I do want to talk about is the author curve, this is something that was very misunderstood and highly controversial.



Source

When the curve was announced, it was to solve one main problem but it came with multiple side effects. The primary problem the curve was introduced to solve was low value spam, frequently on comments. A stake holder can easily hide a good amount of self votes (on their account directly or an alt that makes it more difficult to notice) on comments by upvoting in small amounts (typically under $0.50). These small votes look natural and do not draw attention to them. In large quantity even a whale could use most of their stake on themselves without most people noticing. If they were to upvote 10 shit posts a day with all their stake they were bound to be noticed.

The curve was designed to reduce the rewards you can get with smaller votes until a post or comment reached a certain threshold where the new curve reached a similar reward as the old curve. At the time this was about 16 Steem, but changes rapidly. The end result meant posts and comments would require $0.80 worth of votes to reach $0.65. Even smaller rewards the curve was even more harsh, a $0.02 comment or post would require almost $0.04 in votes to get there.

This curve was heavily misunderstood and many stake holders claimed their votes were worthless now as they only had a 2 cent or 4 cent vote. In reality, this wasn't really true as the curve affected post and comment payouts not individual votes. So even though you may have had a $0.02 vote, your vote would be worth just as much if the content you voted on was above a certain threshold (usually $1-$5). In fact, another side effect of the curve was at the high end where you would receive even more rewards than pre-EIP. The largest posts would receive a "bonus" over the old curve and would mean posts that are popular would receive more rewards. This was actually part of the plan from the beginning, but I don't think many people really focused on this.

Anyway, I am not writing here to tell you about how things were. I just wanted to make sure you had enough understand to understand where we are going and some unique opportunities it presents.

With Hard Fork 25 there will be no convergent curve, and the reverse auction is being eliminated.

I explained how the new curve worked, without it you will no longer be penalized for smaller rewarded posts. The curve won't be linear, but it will be close to linear, and nearly exactly linear within the first 24 hours. Everyone who votes within the first 24 hour (this time window may change) of a post/comment will receive the same portion of the curation pool based on their stake. No one within this time frame will have a unique advantage. This will disincentivize front running large votes and voting quickly automatically to get in first to get the most rewards.

The reverse auction is the five-minute window where you sacrifice a percentage of your curation rewards to get a better place in line and ultimately gaining more curation on all the votes that come after you. This window is being removed completely and there will be no penalty to voting early. Typically curation rewards only accounted for 25-30% of post rewards as a good portion of them were lost to the reverse auction.

These changes should allow stake holders to vote organically without feeling rushed. It will also eliminate the ultra competition to get a vote in within the first 3 minutes of a post, typically two minutes earlier than the reverse auction window ends. Manual curators won't be competing with bots, and large stake holders won't be forced to automate much of their voting or drop massive votes to use their influence.

I said this was about a unique perspective of Hard Fork 25 and so far I haven't done that, I just explained what is coming in as clearly as I can.

If you have been paying attention to communities on Hive, specifically tribes you would have noticed a pattern forming. More of the tribes are moving towards Linear rewards for the sole reason of removing competition for curation rewards. Leo Finance lead the charge being the first to change and my own tribe STEMGeeks followed suit to do the same shortly after. I have noticed newer tribes like CineTV and Proof of Brain are opting for linear rewards.

While the main purpose of linear rewards is to remove competition to vote first and fast, there is something extra that comes with near linear rewards that may not be completely obvious. As the reverse auction window shrunk from 30 minutes to 15 minutes, and then 5 minutes the idea was to allow people to vote more natural and hopefully right after they read something so they don't have to wait until it is a good time to vote. This usually isn't a big deal as most posts take at least a minute to read and you likely didn't see it the second it was created, but comments are far faster pace, especially if you use notification services like Gina and F.r.i.d.a.y.

It is very possible you will be notified, read, respond, and vote on a comment within the first 60 seconds of it's existence. I may even go so far to say comments are a bigger part of the conversation than the original post, especially when you have an author who engages their audience. The curve drastically reduced the incentive to reward comments, paired with the reverse auction it was downright disappointing to vote on comments unless you waited it out. By removing the reverse auction and curve penalty on relatively low value comments, more authors will be more inclined to reward quality engaging comments on their posts. In the past the reverse auction gave the burned rewards to the account being voted if you voted early, this was changed to return them to the reward pool. This means if you wanted to give someone a $1 vote early on (say 30 seconds in), you would be giving them $0.50 but a large portion of your curation reward would be returned to the pool and not to the author you voted on. This also discouraged people to vote comments as it isn't uncommon to respond to a comment earlier than the five-minute window.

You can see this play out already on Leo Finance where Leo stake holders frequently reward quality comments with votes. Tribes do not have a reverse auction window, so there is no reason to wait to reward a comment ultimately making most authors to just not bother.

I think this is an interesting and potentially exciting change that may help increase and encourage quality engagement. It doesn't come without a cost though. The main reason for the curve is due to the fact comments are the easiest way to hide self votes and vote trading with little risk. You also frequently can self vote or even vote others with little to no curation reward competition. This means we will need to be diligent to watch for users who may take advantage of this and attempt to exploit the system.

Sort:  

Hopefully, the comment farming won’t be as bad as hypothesised.

Would love to see more people upvote comments.

Hopefully it wont be, but we all know that scammers gonna scam and farmers gonna farm.

Whack a mole it is.

:point_up:

Posted Using LeoFinance Beta

I'd also like to see way more votes and interactions in the comment sections.
...BUT... the native HIVE vote is maybe not the right solution for mass comment voting. I could easily see a layer2 coming in handy here. Where 'Votes' don't actually vote, but are likes. Your dedicated voting power for comments could be split alongside all your comment-likes after a certain period of time, maybe every two hours. Comment-votes with less than 2h left until payout would be ignored. The UI could then just show the number of likes and the payout.

How well would that work? Who knows... but is it better than setting comment voting default power to 1%? For 'smaller' accounts a lot! :D

One day, my dude. One day.

have an upvote! ;)

One thing not mentioned here (at least not explicitly and clearly broken out) is that one of the purposes of the auction was actually to reduce curation rewards on "dumb curation" (authors that everyone knows are going to trend every time) and allocate that budget to authors (and also non-dumb curation) instead. As you mentioned in the post, with HF24 curation rewards are typically 25-30%. With the auction going away, fully 50% will go to curation, including "dumb curation" which means roughly 30% lower author rewards, but higher curation rewards even to less optimizing curators. Good? Bad? I don't know, it is hard to say, but definitely different.

Maybe I am misunderstanding this portion, would this mean that curators will get more at the expense of the authors?

In the aggregate, since the portion of curation that was previously returned to the pool (and then goes to increase author rewards on other content) won't be. Curators will literally get 50% now and authors 50%. Which is entirely as advertised, so I guess no one should be surprised by it, but the effect of the previous system was somewhat different.

Thanks for the explanation. Are you a fan of 50/50 or would you like to return to 75/25? From my point of view I liked 75/25 because overtime any whales share would be diluted(they are giving out more rewards than they take in). The new direction seems to be based on allowing people to get more rewards for staking.

I liked 75/25 because overtime any whales share would be diluted(they are giving out more rewards than they take in).

This wasn't what actually happened though.

Massive delegations to bid bots and paid votes meant that whales got more like 90% (combo of curation and the bids) and legitimate authors got almost nothing (sure you got 75% as an author on paper, but to get any real votes you had to pay for them in practice, giving most of that back).

The change to 50/50 was part of a package meant to discourage vote selling (including bid bots) and encourage more merit-based voting. And it somewhat worked (there is still some vote selling, but it just isn't as rampant). That included downvotes, the curve, and switch to 50/50. Perhaps only downvotes is enough, but perhaps not. I'd rather change one thing at a time (in this case the curve) and see how that goes first rather than rip out all of the changes that actually worked and bet it all on downvotes alone being enough to keep vote selling from growing back into a bigger problem.

Staking might be a factor too. I mean if staking is a big draw and it keeps the price of HIVE higher, then authors benefit that way too. Personally, staking isn't really something that gets me excited, but I recognize that others feel differently.

I agree that bidbots did ruin the 75/25. I will defintely be keeping an eye on how things turn out a year from now.

The way I was seeing it play out long term is that bidbots were kind of being handled somewhat by the free downvotes thing. I do agree that the curve changing is a good change.

How about 55/45 or 60/40? I realize 50/50 and 75/25 is easier to comprehend due to multiples of 25% and most people can easily thinkg of that.

I don't really have an opinion on the number but I do have an opinion that how well it works overall (and by "works" I mean more broadly including growth, not just the narrow notion of "rewarding authors") is more important than the number.

Interesting to see how this will work out in practice, but I look forward to these changes which make reward distribution in my opinion fairer and easier to understand and use.

All I can say is that on LEO life is much more easy-going, and I am quite sure this has a lot to do with the reward curve. Hope it has this impact on Hive as well!
btw.: Great picture, makes me somehow sad (in Austria there is stupid full on lockdown).

and with the high rewards overall :p

not so sure about that; more and more people abuse the leo tag...

Appreciate the explanation. This might be the best I've seen it explained. Like the fact that there is more time to read, comment, etc and not be worried about any kind of a penalty for the time factor.

A lot of new changes are coming with Hard Fork 25, but the one that most people will likely be talking about is the changes to curation rewards.

It's true that most people will likely talk about this change because it's the one that will affect them 10 times a day. But I don't think it's the most interesting change.

I think it's this one:

https://gitlab.syncad.com/hive/hive/-/issues/129

But maybe that's for a different post.

Posted Using LeoFinance Beta

Interesting maybe, but most do not seem to look long term and are hyper focused on now and their rewards.

Posted Using LeoFinance Beta

Right. It depends on how much stake they have to throw around and how much stake is thrown at them ... or both.

is that implemented with the idea to make the peg to 1$ easier? i am stupid to understand how will that work :) (maybe i did not made the effort)

also seen RC delegations mentioned few times but not sure will they be ready for HF 25. That would also be interesting if we manage to get some more people on the chain.

Yeah, it's the upper peg logic. And I'm happy to see they intend to implement it with collateral, so I'm optimistic.

Regarding RC delegations, that doesn't have to come in HF25, but in theory, it could. It likely won't need a hardfork to implement.

It is supposed to improve the peg but it also creates another use case for HIVE since having HIVE allows you to create HBD when you can sell the HBD for more than $1 (plus a fee of $0.05, so actually $1.05).

So perhaps this will have interesting effects on the demand for HIVE, perhaps not. We'll see I guess.

i missed the $0.05 thing. Thanks

RC Delegations are being worked on and are a big priority.

Posted Using LeoFinance Beta

i listened to blocktrades when he said it is worked on but he was not sure about the progress because he is not working on it. Nice to hear that it is a priority.
More things in the HF is great (if everything works :D )

Good balanced analysis.

Informative! I am in learning phase at the platform and It is clear to understand.
Thanks for Sharing😍

Thats interesting. I didn't realize that the original change to a non-linear curve was to combat disguised self votes. Tricksy people. Makes sense. Hopefully most of those people have been flushed out and back into Steem. My behavior has never really been altered during any of these changes but I like the fact that I may notice a difference going forward. I used to make like 1 Hive (/Steem) per day just commenting and engaging with people, but that went down to like 0.5 Hive per week after the curve changed lol.

I suspect these new changes will result in far more comment voting and hopefully more engagement. What I would hate to see is more "nice post" and "good photography" comments that plagued Steem for years.

Posted Using LeoFinance Beta

Hive doesn't seem to be as bad as Steem was for that but it still happens. People will always do what they can / find new ways to game the system, unfortunately.

What I would hate to see is more "nice post" and "good photography" comments that plagued Steem for years.

One reason why this plague existed is that new users thought it is an easy way to draw attention onto their account. Should not be happening again. This culture of commenting is dead.

I really look forward to the linear curve, and I am glad that the reverse auction window is finally eliminated, I would even support removing or extending the 24h window to give manual curators more time to discover good posts. Otherwise we would have an incentive to not vote on posts older than 24h, which is too short in my opinion. The 7 day window is already short enough. I think we should look at comment spam after HF25, I will not upvote self-voted comments or even downvote such cases.

uh oh....im guessing we will be talking about hf26 real soon to take care of that comment farming :-P

We already have downvotes for that.

Posted Using LeoFinance Beta

I guess if there are ways to monitor at scale and enough people monitoring well it will be enough

This is very interesting! I have to admit, I don't even bother looking at the 1 min old posts because I'll forget to vote on them if they're good because of the "reverse aution" I like that this will incentivize me to read more posts that are fresh off the presses, or older than 6 hours.

I have always been a content creator, I post slowly as I simmer my post very very slowly but I don't get that much feedback as I used to. I have never self-voted myself. I'm not sure how this will affect me as a creator profile as I try to be active but I only engage (being honest) on the days I'm around cos I'm posting around here, which is tops once a week (the rate I take to make a post, same in all networks. But I trust if it's been done this way it's because it's best for all, so I'll see on the go.

Decentralization is very nasty when trying to please everyone. Let's see where hive will go in next 6 months.

2021 is a massive year for our HIVE.

HF25 & Few billionaire HIVE investors will take HIVE to $10

Posted Using LeoFinance Beta

Thank you.

This is my year. This is my decade.

This is our decade. This is our decade.

Let's go HIVE community

Time to advertise good Downvote Trails again, my DV Mana is always in action.

Guess its bad for me, im getting only smal upvotes theese days... But i dont know, in the past i reject some forks who make everything finally works better. Hope is for the best, but anyways i will stay on hive.

I just love the projects and the fertile ecosystem (myself mostly interested on games)

Posted Using LeoFinance Beta

this is the first hardfork I read about that I'm completely satisfied with from the get go

I think these changes are going to be a good adjustment to the experiment. This addresses many things I have observed over the last 5 years.

You are too stupid to do any useful fork.