what an amazing splash into the world of utopian. Amazing work. Its a little late here now and I am trying to get my head around your math. So when the funds are returned to the rewards pool, when and to whom are the redistributed?
Thank you for doing this, a lot of work and time go into these analysis posts.
That is the funny part about all this - when these funds return to the reward pool, what that really means is it increases the relative value of every other vote on every post/comment. That is why I showed the % of the total reward pool that is awarded to each SP band. I don't think a lot of people realize that > 70% of the reward pool goes to accounts with < 5000 SP, which is why there is so much confusion of the "HF20 will just make the rich get richer!" type.
Ultimately the actual impact on the bigger accounts is probably even more exaggerated compared to the smaller accounts. Minnows don't (usually) have a bunch of votes automatically coming to them from a fanbase autovoter. An interesting avenue for future analysis would be to look at what % of the payout value for posts comes from votes before / after 15 minutes. I am willing to bet good money that this also follows the same general pattern, with larger accounts receiving more early votes and smaller accounts receiving next to none. Think about it this way - under HF20 any minnow / new user who writes a post which then gets found by a curator / established user with more SP and receives a vote down the road is going to get a bigger share of the pool, funded by the larger users above playing games with optimal self vote time :)
thank you for clearing that up for me. From what I am aware ( and could be wrong) the rewards pool works on a 30 day average, so there will be an initial 30 day settling period for the pool to balance out one this comes into play
One other thing that I should probably edit into the post is I just (for simplicity's sake) used the current percent of the reward pool commanded by each SP band. Those numbers are definitely inflated for the highest SP users as seen by the greater % of author rewards vs. curator rewards they receive - the actual % of the reward pool commanded by the highest SP users should decrease quite a bit once the loophole that allows > 100% ROI on a self vote is closed.
Looking at the big picture, there cannot help but be a reduction in income for the largest accounts under HF20, and this reduction in income goes to the reward pool at large where the smallest accounts will be receiving a larger % of it than they do now.
so if you dont self vote, and you are a smaller account ( like me) then you stand to have a better post payout?
For the individual account it will depend on how many votes you previously were receiving prior to 30 minutes and even knowing that we can't tell for sure if it will be net positive or negative until we know the actual amounts returned to the reward pool under HF20. Then we can calculate what the rshares received on your posts would have actually been worth when your posts paid out. It is not enough to just say you would have lost X amount from votes that came in before 30 minutes; that is only part of the equation.
Even though you are technically a "small" account it wouldn't surprise me if you personally will lose out at least a small portion under HF20, as you are definitely punching above your SP weight in terms of your average post reward.
But you also will not be one of the big losers - accounts that previously were using the 0 minute self vote to maximum effect were sometimes getting up to 800% ROI on the self vote, and the best they will hope to get under HF20 is 100% (as @tcpolymath points out in the comments here, 100% ROI at least is easily attainable by self-voting comments). That is still a huge difference of course - the accounts that combined large voting stake with use of the 0 minute self vote were previously commanding a hugely disproportionate share of the total reward pool.
The share of the reward pool dominated by 50K+ SP accounts will decrease - that is a net positive for everyone else.