The harm from abusive self-voting is self-limited: A stakeholder perspective

In this essay, I argue against an all-or-nothing approach to dealing with self votes, and assert that the potential for community harm from abusive self votes is constrained by market forces.

Introduction

Many steemit authors seem to be highly concerned about the harm that self-vote abuse will bring to the platform. It has been argued that self-voting is value-free and should, therefore, be banned. Accordingly, initiatives have been launched to automate the flagging of self-voted posts and comments and to propose block chain rules to punish and prevent self-voting and clique-voting.

It seems to me, however, that this widespread concern over self-votes is overblown. Setting aside the impossibility of actually banning self or clique voting, in the following sections, I will attempt to demonstrate that self-voting sometimes deserves to be rewarded; that current concerns about self-voting are often driven by an author-centric viewpoint - ignoring other important stakeholder groups; and that, when market forces are considered, the harm that self-voting can do to the Steem block chain is self-limited.

Image Source: pixabay.com, License: CC0, Public Domain

Section 1: Self voting is not value free

It has been argued that because of conflict of interest, all self-votes are a form of spam that contain no information, and therefore don't deserve to be rewarded. However, it is a basic rule of logic that if a counterexample exists, then a sweeping claim like this is false. So the question becomes, does a counterexample exist?

I would argue that all steemizens have particular areas of skill or expertise where their own self-vote contains far more information than the public at large. For example, @justtryme90 is a professional biologist, @lemouth is a particle physicist, @kenfinkel is a professor of history, @titin is a concert pianist, @morodiene is a professional opera singer, and @somtow is a professional conductor. Is it even plausible to argue that self-votes from these accounts are value-free in their own areas of highly specialized expertise? I think not. I also suspect that nearly everyone here has some area of particularized knowledge where their own self-vote is more valuable than votes from most other voters.

So the argument that self-votes contain no information is easily dispatched. In at least some cases, self-voting does provide information and value to the block chain, and it deserves to be rewarded.

Image Source: pixabay.com, License: CC0, Public Domain

Section 2: Concerns about self-voting often ignore non-author stakeholder groups.

Not surprisingly, the people arguing most vociferously against self-voting are steemit's authors. By and large, the most visible of these people are - almost by definition - good writers, so they make a persuasive case. Just like the self-voter, however, these rhetoricians are also acting from a position of self-interest. More self voting means less voting on the advocate's own posts. And remember, authors are not required to hold any substantial steempower investments! As a stakeholder group, authors are more free than most to view the platform exclusively from a short term perspective.

There are at least two other groups of stakeholders who don't have as much of a voice on the platform, but whose interests must also be protected. These are voters and investors.

From the perspective of the voter, self-voting represents a valuable tool. Self-voting one's own high-quality posts gets 100% of that account's rewards distribution plus 75% (or more) of what other voters allocate. If a voter truly believes that their own post merits an upvote, it is unrealistic and unfair to ask them to abstain. Even on low quality posts, self-voting one's own spammy posts may appear to be the best way to profit from one's investment. However, as we'll see in the next section and in the conclusion, that appearance may be wrong. At any rate, self-voting must be recognized as offering a powerful incentive for voters to purchase Steem and power it up.

From the perspective of the investor, things are trickier. On one hand, the value of Steem is believed to be tied to the quality of content on the block chain, so self-voting of spammy content is harmful to the investment. But on the other hand, voters buying Steem reduce its supply, which is good for the value of investment. What is clear is that by itself, self-voting is not the problem. The problem is voting for spammy content: whether self-voting or otherwise!

Image Source: pixabay.com, License: CC0, Public Domain

Section 3: Market forces will cause the harm from self-voting to be self-limited.

The reward mechanism from voting for content on the Steem blockchain can be imagined as a sort of table. One direction on the table is the quality of the item that's voted upon, and the other is what the voter (or author) does with their rewards. As such:

Quadrant A - Hold the rewards from high quality posts

Plainly, the best value for the Steem block chain is in quadrant A. If an author produces high quality items, or if a voter votes on high quality items, and they hold their rewards - that produces the most value in terms of social media content and in terms of market-supply.

Quadrant B - Hold the rewards from low quality posts

If an author or voter is holding the rewards that are earned from low quality posts, they're still providing a service for the block chain, by constraining the available supply of Steem in the marketplace. Additionally, the more Steem that the author or voter holds, the more incentive they will have to worry about its value and move their voting behavior into quadrant A.

Quadrant C - Dump the rewards from high quality posts

If an author or voter is dumping the rewards from a high quality post, they're providing a service for the block chain with their post selection quality, but they're simultaneously applying downward pressure on value by increasing the market supply of Steem. Notably, because the author or voter is dumping their Steem, the incentive to change their behavior is not growing like it does in quadrant B.

Quadrant D - Dump the rewards from low quality posts

These authors and voters are failing to improve the block chain with their post selections, and they're increasing the supply of Steem on the market. This is where we really don't want to see people. Fortunately, the market is working with us here. As time goes on people in quadrants A and B gain in holdings, but these folks don't. Thus, as time passes the quadrant C and D authors and voters will have less voting influence in relation to quadrants A and B.

Additionally, Steem provides the downvote mechanism for dealing with authors and voters who find themselves in quadrants B and D. Overall, I'm not a fan of the downvote, for reasons that I outlined here. I don't think anyone has demonstrated that the good that it does outweighs the harm that occurs from "flag-wars" and other abusive down votes, but if there's ever a case for it, quadrant D is it..

Image Source: pixabay.com, License: CC0, Public Domain

Equilibrium

Finally, note the interplay among these quadrants. If the percentage of voters in quadrants B and D is high enough to drive the price of Steem down, then voters in quadrant B will lose more in their wallets than they earn from rewards. In this case, they'll move to quadrant A and help drive the content quality up or move to quadrant D to cash out and slowly lose influence. This movement and loss of influence will make room for higher quality authors and voters to jump into quadrant A - and even C - to rescue the price.

Conclusion

I'm not - by any means - suggesting that we don't have a problem with excessive self votes on spammy content, but I am asking us to rethink the "my way or the highway" thinking that is demonstrated by so many suggested approaches towards solving the problem. Instead of taking the naive approach of trying to ban or attack all self votes, here are some steps that we, as a community, can consider to engage in more intelligent aggregated decision making about spammy content:

i. Produce reports on voting diversity and voting cliques that voters can use to intelligently guide their support for particular authors.

• Note that like self-voting, voting cliques also may not be malicious. For example, if father, mother, daughter, son, and grandparents all sign up for steemit, there would be nothing abusive about them voting more frequently for each other's posts than for posts by others. That's what social media is all about!

ii. Consider more sophisticated, but not punitive, reward distribution algorithms such as something modeled after a second-price auction or calculating from the median value (perhaps using median rshares instead of median price recommendation as suggested in that post).

iii. In analog fashion to the creative commons licenses, create a tiered set of voluntary self-vote policy standards that authors can adopt and voters can use to inform their voting (and muting!) decisions. (i.e. @neoxian) - and develop compliance reporting.

iv. Perhaps authors and developers should be searching for ways to increase the incentive to forego self-voting.

As an example of point [iv], I'd point to the @classical-music account and the Steemit's Best Classical Music Facebook page that @cmp2020 and I have launched. No one has really noticed this yet, but with our reward sharing program, we have accidentally taken steps that others can duplicate and improve upon for both the abusive self-voting problem and the problem that content becomes valueless after seven days. Namely, we distribute the liquid rewards from our posts to the authors of articles that we share.

With regards to self-voting, this creates a second path (and conceivably more!) for authors of quality content to earn rewards. With regards to post age, in our latest post, for example, we'll be distributing rewards for a steemit post that's four months old!

Venturing afield for one final rhetorical question before I close: If we can't trust voters to learn to be far-sighted with the power to vote on their own content, then how can we trust them with the power to downvote rewards on content by others?

Thank you for your time and attention. Here's a reward for anyone who made it this far:

Steve Palmer is an IT professional with three decades of professional experience in data communications and information systems. He holds a bachelor's degree in mathematics, a master's degree in computer science, and a master's degree in information systems and technology management. He has been awarded 3 US patents.

Follow: @remlaps

RSS for @remlaps, courtesy of streemian.com.