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 = 932/(932+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.
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?
- Don't mix more than one big variable at a time.
- 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 :)
- 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.
- 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.
