Seven different ways how HF21 could backfiresteemCreated with Sketch.

in #hardfork5 years ago (edited)

In this post, @justineh eloquently described what HF21 is all about and what, if all goes well, HF21 should bring us. For me as both a tech geek (data engineering, cryptography, computer forensics, statistics, software engineering, information security) and a fiction author (alt history, mythpunk, sci-fi), explaining highly technical subjects and writing for a broad audience have always very much been separate worlds, and I greatly admire people who can truly combine the two without either dumbing things down too much or getting overly verbose.

While I'm not sure I can succeed, in this post I want to try and follow @justineh's luminous example in this post.
@justineh wrote about what HF21 is supposed to do. What it is designed to do. What the happy path would look like if all goes as projected.

But what if it doesn't? What if the happy path doesn't play out and parts of HF21 backfire? What could be the consequences? In this post that I hope will complement @justineh's post by looking at just that.

Seven ways how HF21 could backfire.

1) Killing social interaction

One place where the separate parts of the EIP really stack up is the place where much of the platforms social interaction lives. Comments and comment up-votes. There are many smaller accounts whose main contribution to the platform is valuable social interaction and for who the $0.02 .. $1.00 up-votes by minnows and dolphins are the sole source of income. Adding up the effects of SPS funding, the 50/50 split and the convergent linear reward curve and keeping in mind that there is a so called dust-level threshold below what payout does not occur, the sum of it all is quite a blow to the commenting realm. A $0.10 up-vote on a comment that used to bring in 0.075 to the commenter, will only bring in 0.021 after HF21. Only a fraction above the $0.02 dust level threshold. A $0.09 up-vote on a comment that currently would bring in $0.067, after HF21 will yield no reward at all. Will the 71% cut in revenues from comments currently making $0.10 and up, and the dustification of the up-votes on all comments currently making $0.02..0.09, dissuade a large portion of the current social interaction on the platform? It could. Both commenting and up-voting of comments are very much disincentified by the combination of SPS funding, the 50/50 reward split, the convergent linear reward curve and the $0.02 dust level threshold.

2) Hindering onboarding and driving away smaller fish

It is not just social interaction that could suffer from the stacking up of the combination of SPS funding, the 50/50 reward split, the convergent linear reward curve and the $0.02 dust level threshold. New accounts as well should find it harder to grow from redfish accounts with zero followers and a mere dozen SP delegated to it to minnow level and above. Many smaller accounts will see their revenues go down. If we consider that STEEM competes with add revenue sharing platforms for content creators, it could be expected that HF21 not only would hinder the potential for onboarding content creators currently making money from ad revenue sharing, but could also drive newer and smaller content creators to add revenue sharing platforms.

3) A flourishing bid-bot economy

A third way in what HF21 could backfire would be if it ended up promoting the bid-bot economy beyond its current shape and size. There are several ways how the bid-bot economy could end up being stimulated by HF21.

  • The convergent lineair reward curve could stimulate the use of bid-bots in order to beat the curve.
  • The 50/50 reduces ROI for bid bot users and increases it for bid-bot owners. This could result in a downward price correction for bids, what in turn frees up funds currently being pumped around through bid bots and create demand for more bid bot delegated SP.
  • The down-vote pool with free down votes makes self votes an easy and fully designating target for policing initiatives, this could drive passive stake holders to delegate more of their stake to bid bots.

The combination of these incentives could potentially create a boost to the bid-bot economy and as such increase rather than decrease the level of false curation on the platform.

4) Passive stake holders selling

The EIP assumes passive stake holders can be incentified to become active stake holders. But what if they can't and all the EIP does is incentify passive stake holders to move their stake to other investments. A massive sell off could result and this could potentially dramatically drive down the price of STEEM.

5) Escalating flag wars

A fifth way that HF21 could backfire would be escalating would be if the new free down-votes would end up to escalating flag wars. Today a sizable percentage of down votes can be attributed to retribution. It is hoped the pool will create incentive to act against undesired behavior, but what if larger stake holders who have invested in a specific less desirable business model won't go down without a fight? The flag wars could quickly escalate and valuable accounts could get decimated in the process.

6) Bad behavior shifting up-curve

The convergent linear reward curve is aimed at specific types of bad behavior taking place at a specific range of post value. But as the curve doesn't scale incentive across the post value spectrum, it is quite possible for poor behavior to simply shift up the curve. So while the curve might dissuade a dolphin from acting badly at scale X, because it doesn't do the same for an orca at scale Y, it might end up just shifting bad behavior to next fish size accounts.

7) Traders losing faith

Finally, with the EIP claiming to address the economic system of the platform, and with exactly zero measures working against inflation, the one property of the STEEM blockchain that really touches the economics of the world traders live in, the seventh way how HF21 could backfire would be by making traders lose faith in a coin that fails to address what they believe to be the prime economic shortcomings the STEEM coin. If this happens, this could potentially result in a huge drop in price of STEEM.

Push EIP to HF21 or have an EIP kill-switch?

Please note that this is meant to be an objective iteration of seven possible ways in what HF21 in its proposed form could potentially backfire. I am not saying HF21 will backfire in any of the above ways. I am saying it might backfire in one or more of the ways above and possibly in ways that I didn't think of. The important thing to realize is, I think, that any of the above scenarios is a viable way in what things could play out after HF21 happens, and we should probably consider what should happen if it does.

One way to deal with the potential of the above seven scenarios would be putting in a kill switch into the hard fork. Just roll out HF21 with proposed features and put in a kill switch for the witnesses to use at their discretion if things end up going south in any of the above ways.

Alternatively, the EIP could be delayed to HF22 while ways are examined to minimize the probability of any of the above doom scenarios playing out.


Sort:  

Number 2 is already a reality sort of for me. I rarely vote on a comment anymore. At 50% Vote power my 1600SP is only valued at about 0.018. Dust level knock off. There are a lot of post I like to comment on, there are a lot of post I like to vote on. I do not have the Vote Power value to spread my vote as far and wide as I would like, and have it still be a meaningful vote.

Unfortunately same is for me, 50% of my vote is less than dust threshold. I still vote comments though, even if that doesn't make me or the commenter any good on the reward side (it actually hurts me, because I drain my VP and don't reward the commenter with real rewards either). Probably at these prices I would need around 75% upvote to go above the threshold.

My probability estimates on each:

  1. 15%. A lot of people feel a compulsion to speak their mind... and that is unlikely to change. There will be a lot loss commenting just to get rewards, though. I could live with a little less 'great post brother, please follow back my resteem'... although RC cost of commenting has made most of that disappear after HF20.
  2. 30% Not one I had put a lot of thought into, but I know quite a few onboarding initiatives that are gearing up and have solid game plans in place. Things could still falter depending on how CLRC plays out, but I suspect their 'losses' will be offset by the reward pool gains from downvote wars.
  3. 99.99% This is a near certainty, especially as bid-bots either fully monetize their downvote pools or use them for retaliatory strikes to protect their customers from 'bid-bot abuse' downvotes. Another potential racket is protection money. Average customer payout = x; any post payout > x receives randomly distributed downvotes... I guess that's an 8th item for your list, though?
  4. 5% High probability of bid-bots flourishing make this one a low probability... passive investors will continue to take some risks on the underlying STEEM price so long as passive delegation rewards remain high. Prices have been falling on dlease.io lately, but that's because a lot of active capital is piling into SCOT tokens instead of projects with high lease demands. I expect 50/50 curation to restore demand for delegation leases alongside spurring bid-bot rewards to delegation.
  5. 95% Another near-certainty (especially in the short-run). I'm told that there is 10MM worth of SP waiting for active downvotes... human nature being what it is, there will be lots of flag-tantrums and retaliatory downvoting happening. Social norms need to shift massively from perceiving downvotes as 'stealing my rewards' to the more accurate 'establishing consensus' and without that shift most downvotes will trigger retaliation.
  6. 100%. Almost all bad behaviors on Steem are profit-maximizing. If $5 buys you $2.60 of upvotes, but $10 buys you $10.40 of upvotes, which do you pick? (note that this was an intended effect so at least something will work as advertised - their argument is that moving bad-behavior up-curve will make it easier to identify and deal with using downvote curves)
  7. 0%. In my opinion this one confuses traders with investors. If your holding period is less than 13 weeks, whether inflation is 8.5% of 2.5% doesn't have a material impact. Investors can easily offset inflation with passive income... traders don't actually care. Really... you can't tell me that traders cared about fundamentals like inflation rate when SBD went to $12... (though my 0% is a little irrelevant here... traders have mostly abandoned STEEM/SBD already. Why do you think they keep getting delisted?)

Obviously my probabilities are subjective... if I had a quantitative model that could measure such probabilities accurately I would be making bank in coins with volume, not piddling around on Steem with a $25k passion project.

@josephsavage,

Intelligent commentary.

Proposal:

  1. Appoint a Council comprised of people surnamed "Savage."

  2. Delegate to said council dicatorial powers for 6 months. (I volunteer to handle the "disciplinary aspects" for the Council).

  3. Be astonished at how polite and well-behaved everyone instantly becomes.

The best argument in favor of centralization is watching the members of a decentralized blockchain try to solve even the simplest of problems. Why not just ban bidbots? How?

  1. Make it a Violation of Steemit's Terms of Service to RUN and/or USE bidbots.

  2. Make it a Violation of Steemit's Terms of Service to use any other Front End that does not similarly ban bidbots.

This would collapse bidbot usage instantly. Instead, we have these Rube Goldberg proposals ... trying to discourage bad behavior by incentivizing good behavior.

Enough carrots ... it's time for sticks.

Quill

Bidbots are a symptom of larger problems. Any solution that disempowers them disenfranchises significant parts of the investor base. At this point, killing off bidbots via mandate (instead of structuring the appropriate incentives that people stop delegating to them and using them on their own) would destroy investor trust in the blockchain. I very seriously doubt that the patient would survive the 'cure'.

Geez, it is like the perfect storm. All of the above!! I can tell you one thing, ever since I got my comments were annihilated here it does make me want to go down fighting! Flag wars a coming for sure! More rampant bidbot abuse coming, people will start bid botting their comments just to make the status-quo!
Screen Shot 2019-06-23 at 2.13.50 PM.png

Screen Shot 2019-06-23 at 2.14.59 PM.png
I did vote for this author a little because I thought she was over flagged that's all. This shows that when people are pushed into a corner on steem they will turn to the bots to remedy!
That is what the people in charge want. The highest people are more tech savvy, have more steem, set up more bot systems, and make more money! Now they are greedy and want even more so they drive the price down and will probably buy more up on this dip too!