Shaved Whale Balls and MEGA CIRCLE JERKS- The economics of HF 19; a witness update.

in #hf198 years ago (edited)

This is an economics post trying to calculate the slightly positive effects of linear rewards on the reward pool distribution (potential 5% positive effect toward equitable distribution) followed by a discussion of the likely intensely enhanced clustering around limited numbers of posts. If you thought circle jerking was bad now wait until you see HF19!

The Basics of HF 19

Steemitblog annouced HF 19 and entitled it Linear Rewards.

So, "Linear Rewards" has gotten the most press.

https://steemit.com/steem/@steemitblog/2hfxx1-pre-release-hf19-linear-rewards

Linear Rewards vs Squared Rewards

The post doesn't detail it terribly well. So, here's my best explanation I can offer. When calculating post rewards there is a term called Reward Shares. I'm still digging into it, but my current assumption is that it's a function dominated by Steem Power. Right now when calculating your actual reward the backbone takes your amount of reward shares and squares it. In the new update it will take your Reward Shares unaltered.

Here's the coding changes.

What's the effect?

Well, distribution is such that 93% of the Steem rests in the hands of 1% of the users (excluding the steemit account, which would increase the number if it was included). So, whales basically dictate the rewards pool. Under Linear Rewards they will get 93% of the say because they have 93% of the steem.

That's an improvement over current conditions. Assuming they get 93% of say and I haven't totally misunderstood rewards shares and the calculated effect is squared they end up dictating 99% percent of the reward pool currently. (I think the math = 93^2/(93^2+7^2)= 99%). This doesn't happen right now because of "the non-voting whales experiment," but that's what could be happening.

Assuming 240M steem, 10% inflation, and $2.16/steem the reward pool is ~$50M/yr which works out to$137k/day. 85% of the rewards pool goes to authors and curators so, here's the net effect.

Whales control $115k/day now.

Whales control $108k/day after HF 19.

That's a modest change, so I'm happy to say it's a "shave" to their current influence on the platform now (aka whale balls). If all the whales come back voting hard though the shave won't be enough to account for the increased voting activity.

Overall, in theory it looks like a change of 6% of the say of how the reward pool is spent and that works out to about $7k/day towards better distribution control by smaller parties.

Keep in mind tho' that not all whales vote because of the experiment... so, what happens in reality is a little different than this. It's just meant to be a high level thought that suggests Linear Rewards is mildly beneficial for the common minnow in theory, but will probably have a negative impact overall if whales come back voting hard and don't stay experimental.

Vote Impact

The changes to the impact of a 100% vote are more material. A 100% vote will be 4 times more powerful once HF19 is completed. As it stands now a single 100% vote uses only 0.5% of your remaining voting power (voting power regenerates fully over a 5 day period). That means real users would need to vote 40 times a day at full power to use all of their voting power. This leaves users who are less active unable to fully leverage their voting power. After this change, a single 100% vote will use 2% of your remaining voting power, meaning that 10 full-power votes a day will now exhaust the majority of your voting power. Of course, if one does not wish to use this much voting power in a single vote, anyone is still free to lower the power-percentage of their votes accordingly.

I think the thought behind this is awesome. Minnows aren't curating a lot so they are missing out on curation rewards. Let's lower the number of votes needed a day to curate effectively. That's awesome... but I don't thinka it's gonna do what they thinka it's gonna do.

Right now curation is dominated by a handful of posts with huge rewards. It's often the same people making the same huge rewards. It seems like the big guys are all helping out the same people and so those posts get huge. It turns out that's not only good for the author, but it also helps maximize curation gains for the whales. (Don't hate the big authors hate the game)

Right now whales have to distribute those big votes over 40 posts. HF 19 they can distribute them just over 10. Instead of a whale dropping a max of say $50 on one post now it's going to be closer to $200... you know what's that's going to do to post values? TO THE MOON!!!! but only for like 10-20 posts a day.

Remember when we talked about only doing 1 major change at a time?

As a scientist I know not to mix more than one variable in an experiment at a time. Voting and linear rewards should have been separated. It'll be hard to isolate what is causing what if you haven't read this post yet. I fully believe this plan was implemented to help minnows. I think the fanfare over linear rewards dwarfed attention to this second component, which I actually think will have drastically more impact on voting rewards distribution and be extremely negative towards equitable distribution of reward pool funds.

Terrified and calm

So, here's what I think is going to happen. You're going to see your rewards drop like crazy unless you're a top poster. People will abandon the platform and the price of steem will drop. Maybe a few other witnesses will understand what just happened and go along with some of my proposed fixes below. There will be manual adjustments through flagging again that will be seen as unfair that actually help the other 138,980 users on the platform. Then there's a relatively quick fix.

After that fix we see an increase in steem price and people come back. It also brings awareness to how better distribution of the out going rewards pool is important to this place, and we overcompensate in that direction. I suspect a roller coaster. I also think people here have good intentions and it'll be fixed.

What to do?

  1. Don't mix more than one big variable at a time.
  2. Go the opposite direction with vote count. Give people more votes (80-100) so it's hard to keep only supporting the top 20 authors every day and you have to dig for more people. Support steemit.com based curation strategies that are fast to implement for new users so they can curate too (ie 1 click implementation of following the Minnow Support Project's curation trail). It ends up looking like a curation mutual fund :)
  3. Make vote power loss dependent on the relationship of SP of the voter to the author. If they have the same SP it costs normal. If you're voting on a minnow with much less SP it should cost a whale much less of their voting pool to upvote their content. That changes the incentive to upvote minnows.
  4. Double flagging efforts until its resolved.

BTFD

Should be an easy fix. So, buy the dip. This is also all theoretical because there are way too many programmed variables and human interactions to predict what we see in reality.

I've altered my witness to HF 19 to verify my hypothesis. I think the blowback in a month will help minnows. Hold onto your hats.

In other news

My witness campaign is going great. I'm at Witness 51 after 2 weeks. I witness ~20 blocks per day. About 250 people have upvoted my witness.

My witness central project is going great. The Minnow Support Project has 300+ people in the Discord channel from multiple countries helping one another out with votes, resteems, follows, and writing suggestions.

I'm connecting with multiple countries that could use centralized support to growth hack in depressed economies. They need steem so they can have a dependable deflationary currency. Should be fun.


Sort:  

This is not an economic analysis. It's just fear, uncertainty, and doubt. I appreciate the concern, but that is the extent of the analysis here.

The number of rshares cast on a daily basis after HF19 goes live will be higher. But the reward pool will be the same. The concern is that the rshares will be concentrated. But there's no basis to support this assumption.

The number of rshares cast by a single account will still be regulated by the off-chain influence cap like it is now (by the whale truce). The only way to cast more rshares will be to spread votes over more posts (just like now).

The net effect is more posts will get more votes. And since the daily reward pool will remain unchanged, this will just pull down rewards from the higher earners and spread it out to the lower earners.

I hope you are correct in that there will be more of a distribution of the total rewards pool across the whole of Steemit and not more concentration of a small percentage.

I tend to agree. I'm pretty sure HF19 will bring more equality.

The main focus should be to make the platform seem more lucrative for new users in order to adopt more members. So the focus of the whales should be to increase the wealth distribution down, in order to have higher volume coming in.

This means less paid for votes at the top, but more votes.

The problem would be how to justify this, as this basically would mean whales would have to redistribute a small allocation of their gold mine. Why would they do that? For the community to grow? Things are sweet as they are.

But this is needed for the community to grow. Not many people would take the hassle to join on upvote-bot program in order to earn 1 buck per post. Its only computer dudes like you and me, who already joined this betaversion.

There need to be a just balance or we'll all end up losing.

I believe you are correct in some perspective.

As it stands now a single 100% vote uses only 0.5% of your remaining voting power ... .That means real users would need to vote 40 times a day at full power to use all of their voting power.

That is not accurate.

First, if you were losing 0.5% for each vote at full power, after 40 votes you would lose only 20%, not "all your voting power".

Second, after each vote at full power, you lose only 0.5% of your current voting power. So, after 40 votes at full power, your voting power is 99.5%40, that is 81.83%.

I am voting more than 40 times a day at full power and my voting power has never been less than 30%.

What is true is that if your voting power is 80%, after 24 hours of not voting at all you will get back a voting power of 100%.

After HF19, starting at 100% voting power, voting 10 times at full power, your voting power will be down to 81.71%.
Going from 80% to 100%, with no voting at all, will still take 24 hours.

That section is a direct quote from the steemitblog post. Take it up with them :)

@steemitblog is not God. They obviously made a mistake here.

It is obvious to me, and it should be obvious to you, that starting from 100%, losing 0.5% forty times cannot result in 0%, but at 80%.

I agree. @steemitblog didn't express themselves correctly. We know what they meant but what they meant isn't what they wrote.

you are correct. and only slightly modify the description and they can be right too.

That means real users would need to vote more than 40 times a day at full power to use all of their voting power given in one day.

Thanks for putting your points across in this post they are interesting to hear. It's very hard to know what exactly will happen but I really hope it helps newbies more as that's what will really grow the numbers here!

Thanks for the analysis, this is interesting. I like all of the suggestions you've presented. I think giving more votes rather than less will allow for curators to look for more content (or whale authors will just create more content that can voted on amongst peers). And making it so less vote power is used on minnows to encourage votes is good too.

What about increasing the effect a vote has on a minnow if it comes from a whale? Further incentivizing curators to look for good content from the lesser knowns, and giving them a better chance to be seen.

That's why I want less voting power used when voting on a minnow. I think it's similar to increasing rewards when the SP spread is large.

What if you did both? It would be a boon for minnows. Your suggestion gives more votes for the whales (a good thing for them) and mine gives a better chance to be seen for the minnows (good for them).

Whales have more opportunity to go find good content to curate, minnows get the juice they need when a whale backs them :) win/win

I think it'd be awesome. They should still run them one at a time tho.

Oh for sure, agree with you on that. Thanks for all the hard work you're putting towards the minnow population

This is an excellent point and would greatly increase the incentive to vote for minnows. Let's be honest, the future of Steemit being able to scale relies on the minnows, and lots of them. If Steemit is to become what we all know and hope it can be, something needs to be in place to promote whales and dolphins alike to search for good content to curate in the minnow pool.

Make vote power loss dependent on the relationship of SP of the voter to the author. If they have the same SP it costs normal. If you're voting on a minnow with much less SP it should cost a whale much less of their voting pool to upvote their content. That changes the incentive to upvote minnows.

This really activated my almonds. I think this would be an effective adjustment. Good write up.

Also, concur on not mixing more than one major variable at a time. Easier to assess changes, UNLESS a solid argument could be made that the efficacy of the two changes are interrelated/dependent on each other for an optimal outcome hypothesis.

Activated my almonds LOL

Ah finally a post that actually makes me understand the upcoming hardfork. Seems like a good attempt and while I am quite active now I cannot allways be, like many people I think. Then it is nice you can just in a few minutes do your curation.

For some reason, only thought of whale balls and miley cyrus

Looks delicious.

That is funny.

I'm going to re-read this several times as it's a lot for a minnow to take in one bite, I may have more questions as I work my way down through it again but:

"Right now when calculating your actual reward the backbone takes your amount of reward shares and squares it."

So right now, my 0.01 reward share is actually 0.0001?

and

" In the new update it will take your Reward Shares unaltered."

So my 0.01 reward share will actually be 0.01!!!!!!!!!!!!

HOLY FUCK THAT'S AWESOME!!!

Resteemed and upvoted (back down to 72% voting power)

It's something like that. Reward share isn't exactly SP though so it's more complicated.

It's more like your RS of 2 = 4.

Their reward share of 10,000 = 100,000,000.

Now it will be 2, 10,000. Which is fair.

So the linear reward share system sound good to us minnows, but frankly, if the RS is still not decouple with the difference between SPs, a whale's vote will probably make minnows getting practically nothing.
I think to help minnows, there should be some way to boost them up based on their comments' quality.
Well... maybe for next HF.

I get it, but my background was 20+ years watching GM self destruct from inside.

I'm not even at a $2 RS yet, I'd be peeing myself with $75 or $100 posts.

I'm not some starry eyed fresh college grad dreaming of that corner office and $500k salary and supercars.

But unless there's an infinite supply of former Facebook or Reddit users with 4,000 followers coming here you need a constant infusion of fresh blood to keep adding value and and develop into dependable content creators.

It's hard to keep talented people when their measly 0.01 turns into 0.0001 for the 1st couple of months.

Or am I just completely missing something?

Honest question, I'm not trying to be a dick.

Oh, I agree with all four of your "what to do" suggestions.

Crap.

Perhaps a UI change can counter some of the negative effects of the new 10 x 100% vote (if your predictions come true).

For instance, the upvote arrow could be split into multiple buttons for 25/50/100% power.

Kinda messy, but I might actually prefer that over the 3 actions I currently need to do for a custom vote strength.

Relative to now those buttons represent 100, 200, 400 post HF 19. I think that's bad news bears.