RE: How will the HF20 dust vote changes affect the vote values?
There are some fairly interesting effects that this will have going forward that you suggest but don't explicitly call out:
Firstly, votes from big players get a magnitude enhancement because votes from small players are diminished by a flat amount. Anyone that can afford to give high SP votes gets that much more influence as a side effect.
Secondly, new/low SP players take a significant hit to their voting power, across-the-board, in ways which they can do nothing about unless they want to bring literal investment in from off the blockchain because, ironically enough, the voting value changes make it harder for them to gather value as a result of interacting with other low SP accounts.
Ultimate upshot: more whale fishing. The main side effect of these particular changes to the voting valuation is that whales are even more valuable and anything less than orca (and by that I mean the vast preponderance of minnows and krill) become less valuable at the sizes of audiences that you can expect to reach.
Personally, I'm not thinking that's a total win. In fact, I think it leads to behaviors which, coupled with the shortening of the curation window to 15 minutes, are going to lead to less diverse, more highly tailored and targeted content being produced. Bots get double the advantage because they have a perfect sense of timing and new users/focused audiences lose out.
The numbers are difficult to argue with.
I understand why @crokkon is apprehensive about asserting HF being an enhancement for the big players but my intuition is in line with yours. It seems like that the little guy is getting nerfed and whales are getting buffed. I really would like to understand the underlying rationale for this decision as it seems nonsensical to me but maybe I'm jumping the gun.
I know there had been issues with a proliferation of @steem faucet downvote accounts that may have contributed but that is speculative. They were more of a nuisance to minnows and plankton and really not a concern for whales especially the self voting variety. Even in the remote chance that was the reason, it would have been better to nerf the downvote and leave the upvote alone.
I'm still not 100% clear on all the tech details but from the impression I have gathered. I am not really a fan of HF 20 as it seems to give more power to the already powerful which IMHO is something that will encumber adoption in some sense.
The gain for big players is marginal (0.1% with the votes from that week of data) and assumes that the voting behavior remains the same. Any adoption of the curation-related votes will reduce that gain. Also, 0.1% is small against the variations of the sum of total rshares across days, not even including steem price changes. It's mathematically nice to see at which point the loss of the individual vote value is compensated by the reduced total sum of rshares. I wouldn't call this a real enhancement for big players before having seen HF20 data.
It is true, however, that new/low SP accounts get their vote value reduced and get less from interacting with low SP accounts. But are we in the position to complain? After all, it's (in most cases) Steemit's SP, given out for free. Is it an incentive to invest, or could it limit onboarding of new users? I don't know, actually...
I also agree that these changes are not necessarily a measure against bot votes. Bots have to include another parameter into their algorithms, but they'll get this done easily, maybe with slightly reduced returns. In combination with the 15 minute window, I could image that human curation (for profit) even gets harder in HF20.
When I referred to the "win" for big players, I was specifically thinking of the gain that they get relative to smaller players in the game.
That's not to say that the 0.1% is small given the variation of rshares by day, that's true – but that 0.1% is true every day, every week, every month, every year, and if there's one thing we know about values which compound in the long term, it's that the actual absolute value becomes significant.
No, from my perspective more interestingly is that low SP account hit, and especially when interacting with one another. That's murder. We know from looking at the numbers that there are vast and seething seas of low SP accounts, and for the most part low SP accounts only know one another. What this appears to suggest is that that activity is going to become less important to the fiscal transfers on the blockchain, shifting more of the focus to the activity of and interaction with significant whales.
In a sense, we are in the perfect position to complain. Users like you and I appear to be interested in the actual nature of content which is getting put onto the blockchain in terms of usability, how much it entertains – how much the system actually functions as a social network which promotes content which is of some level of useful quality.
From my perspective, a social media network which is successful at being a useful social media network is the primary driver here. As much as Steemit Inc.'s interests differ from that particular endpoint, I think they will inherently do less well, because the further they go from doing that one job well, the closer they go to just being one more cryptocurrency that isn't Bitcoin.
Given the changes to actually creating accounts, it looks like it will be effectively "cheaper" to create a new account in the new regime, but those individual accounts will start with even less effective voting power than the current ones. Combined with the new-accounts-as-a-class decrease in voting effectiveness across the board, and the continuing lack of impact for votes for a new users making a real difference to their experience – I get the feeling that the intent of the design was to bring more people on more quickly, but the actual impact of the design will be to bring people on more quickly and then have them leave just as quickly because nothing they do matters, including gathering the attention of other new players in the space.
That's my feeling as well. Under HF 20, the bot situation is simply that their voting timer becomes easier for them to hit and less easy for humans to hit, improving their advantage over human curators even more, and the relative change in voting power doesn't matter a bit to them. If anything, since most voting bots represent a high SP player, they gain an additional advantage because humans who might be seen as competitors in the same space have to get to the same level of SP to have the same kind of voting leverage and voting advantage, and one of the most obvious ways to do that is to use the voting bots to do so – which increases the SP of the voting bots. One ends up in a situation where you literally have to pay your enemy in order to possibly have the power to pay yourself in the future.
This is just an extension of a design problem they've had for a while.