The Steemit Tragedy of The Commons
To quote wikipedia (HEEL @cheetah), a Tragedy of The Commons is defined as follows:
The tragedy of the commons is an economic theory of a situation within a shared-resource system where individual users acting independently according to their own self-interest behave contrary to the common good of all users by depleting that resource through their collective action.
It has been my experience over the past few weeks that Steemit is descending quite rapidly into a situation where the above definition is being applied almost verbatim.
There are a few major topics I want to touch on today. Unfortunately I do not have much in the way of solutions, but I want to raise questions so that the community can openly engage in discussion with their opinions as well. I fight actively for, and am invested in, the future of steemit (I pay for @cheetah bot to run), so I hope we can garner some improvements as a result of this post.
Curation, Voting, and Flagging Strategies
As independent actors, we are all trying to maximize our own rewards here at steemit. We are acting as individuals, rather than as a community. I am guilty of this myself, having toyed with a curation strategy (bot) to use my vote power to improve my own Steem Power (according to @trogdor I am doing fairly well too.), and this means using my vote power selfishly, rather than using my vote power on something that I enjoyed reading. Perhaps curation rewards should stop after a post is trending, reducing the pile-on effect?
The problem is that this incentivizes the individual, not the community as a whole. When all users start doing this, voting on a "sure shot", we only further centralize the reward pool on a small minority, and leave minnows with little reward for their efforts.
Unfortunately, I, like you all, are witnessing firsthand the effect that this selfish voting imparts. Our famous actors are becoming more famous, to the point where the few who realize have begun using their fame to promote lists of hidden gems and featured authors. But they still gain the steem power, getting fatter and fatter.
Furthermore, using a flag on low-quality posts is disincentivized due to the threat of retaliatory flagging across all posts. I am a victim of this constantly and am used to it, but for most minnows they do not have the power or reputation to risk this possibility. In this regard, the action of self-interest is to avoid calling out spam, plagiarism, or scams, an instead focus only on the self and their own posts or rewards. @liondani suggested anonymous flagging, which may solve this issue, but potentially may lead to anonymous abuse, so I do not have an answer to that either. The question of flagging has become larger and larger, on the use cases and accountability of using flags, as well.
To quote @tuck-fheman:
It appears that Steem has become all about flagging, downvoting and opposing instead of the original idea of upvoting for rewards. :(
What is Fair?
A second issue is fairness, and the distribution of reward from the reward pool. I have strong opinions on content, enough to continuously butt heads with several whales (no names -- no need to brigade).
The fact is that whales have the ability to make a post trend instantly. This can be a good thing or a bad thing -- given a benevolent whale, a really strong minnow post can be put into the spotlight despite the user not having a large following. But if the quality of the post is low, we run into a very distinct issue: users view the front page as a metric to gauge successfulness of their own posts. Then it becomes a case of "monkey see, monkey do", where the average user, seeing plagiarism, copy paste, and low-quality posts on the front page, contemplates their risk to reward of each and every blog post.
Why would someone spend hours on a quality post, when they can spit out 4 low quality posts a day on multiple accounts, and randomly get chosen by a whale?
When a whale upvotes a low-quality post, and causes it to trend, it only further incentivizes this in the future. However, curation rewards are given mostly to the whale anyways, so they have selfish reasons to NOT remove their vote (and use vote power), and keep the resulting curation reward, contributing to the tragedy of the commons.
What is Right?
As one of the head combatants of abuse and plagiarism over the past weeks, I have had many discussions and many arguments over the direction of content on Steemit. While I believe that copy paste is a waste of my time, again, many users disagree, and mine and others' combative nature against what we feel is abuse has been more vigilante actions than anything else.
Personally, I feel that Steemit should NOT be a news aggregation source, unless the provider of the information IS a news source. In this case, I believe people who copy paste articles that are not their own, whether or not it contains a source, should NOT be encouraged. This is equivalent to the difference in a reddit post or facebook status. Do you copy paste the article to reddit? NO! You just have a link and promote (and provide) discussion. Do you make your facebook status a full article? NO! You just have a link and promote (and provide) discussion.
We can do so much better than that; Steemit is something new and original. Lets be original too.
But what is RIGHT? If the answer were unclear, that would be one thing, however there currently is no answer as far as I am aware. Neither @dan nor @ned have made a public stance on official or unofficial set of rules for the content on steemit.com. Also, note that I am saying steemit.com here, not the blockchain steem. I think in the future there can and perhaps should be avenues on the steem blockchain that can be independent from the company and website, but that is off this topic.
Are reposts okay? Is copy paste okay? What is right?
I don't actually know.
But what I do know, is that everyone is acting independently on what they feel is right. And this independent selfishness is again contributing to the tragedy of the commons.
Transparency of Leadership
Here's another topic that I am likely to recieve a lot of hate for bringing up, but I feel the need for this to be discussed. This likely will ruin my chances as a back up witness, as this will not put me on the good side of some of the whales/employees but I feel that this needs to be said.
There is an issue in the transparency of leadership and witness activities, as well as the development towards the future of steemit.
If you don't believe me, I have a few examples. Lets start with the latest witnesses exposed edition, here.
There are a few issues that come to mind to start. Not all witnesses felt the need to respond, or were discussing stuff not explicitly steemit related. Why is this? This is not an issue with the witnesses themselves, but more with the transparency of actions -- the witnesses do A LOT of work, and they do not have the time or willpower to justify all their actions. Their updates are only a small part of the picture: they do a lot behind the scenes.
Behind. The. Scenes. That is the issue.
Furthermore, there is a private slack that only the witnesses and employees are invited to. Who is invited to these discussions? It does not appear to be a community decision.
Why are back up witnesses not invited to the private slack? Do they have no say? Is it only the bitshares witnesses, whom are very similar to the steemit witnesses, that are invited? I don't actually know -- it is not transparent.
As a second example, I will start with an issue that has arisen just today. If you are aware of the history of bot wars on steemit, you are also probably aware of @williambanks and myself continuously in (reasonable/gentleman's) debate.
But what do you know, we are on the same page today.
There is an open issue on the steemit github here.
This issue is less than a day old and already is one of the most discussed issues. The code has already been implemented and is on track to be included in the next hardfork. There has also been a MASSIVE discussion on this in steemit.chat, and we have been debating the potential issues that this change could impart.
There is already a very forward response to the issue here, and while I think the tone of the post is a bit exaggerated, it shows that there are people already speaking out on the potential for censorship of users.
Could this be another case of acting in what only one believes in, contributing to the tragedy of the commons?
To Be Clear...
This post is meant to incite discussion, on the following points (summarized):
- How can we prevent the tragedy of the commons by inciting communal behaviour rather than selfish behaviour?
- What is right, or fair, and are there (unofficial) rules to steemit?
- How can we improve transparency with the witnesses and leadership?
I hope to use what voice I have as "the cheetah bot creator" to get some some attention on these matters. As I am a believer in Steemit (I'm not powering down), I hope we can discuss and solve these issues, so let's try and be civil in the comment sections, and think of strategies to improve the current situation.
This is a generalization. Some witnesses do a lot of work. Some others don't do anything and post excuses like "I have a day job folks" or "see last week update" where last week update was already "see last week update" or don't post an update at all.
As a reminder, witnesses are paid in average 1500 STEEM per day, which at current rate corresponds to north of 60k USD per month. As I explained in an earlier post in this thread, this is the fixed salary a successful quant trader or portfolio manager makes on Wallstreet, or the salary of 6 senior developers in the US or 20~40 experienced developers in developing countries. And all that witnesses are effectively required to do is run a node on a decent dedicated server (cost $100 per month) and upgrade whenever there is a hard fork (a few hours of work per month) => source: I am running a node too.
The excuse (because yes this is an excuse) for witnesses to receive that crazy pay is that witness pay is in theory supposed to be a "hired by the blockchain" kind of deal to pay for developments, marketing etc. Although this was how it worked on Bitshares, this is completely untrue on Steem. Steem comes with a builtin mechanism to raise funds for specific projects: blogging, and this mechanism is far superior to the old Bitshares witness model because it forces the developer who wishes to raise funds for a project to make a convincing post to present his project, give updates etc. We have seen it at work: so far the blogging and voting system has helped funding many apps, projects, websites, helpful bots, marketing efforts, reachout efforts, evangelization efforts etc and Steemians are rushing at the gate everyday with new ideas and new projects all more awesome than ever. It works just fine, and all these people do orders of magnitude more work than what witnesses have shown they were doing so far (for the ones who do anything at all), and for a far more realistic pay that the community explicitly approved.
But the more ironic of it all is that, in spite of the fact witnesses are already ridiculously overpaid for the work they don't even necessarily do, the ones who end up actually doing something useful end up posting about it, and get showered with cash (like if what they receive as witnesses wasn't enough already). Example @arhag: that's a witness who does some work, but guess what, each time he posts about what he does he is showered with more cash than what he would have received at a normal firm for doing the job. Just look at his posting history. Nothing wrong with that, except for the fact he (like every other witness, not pointing at @arhag in particular) is already receiving 60k USD worth of Steem Power per month for doing precisely that!
Conclusion, witness pay is redundant. It can be trivially and obviously determined by looking at witnesses blogs that witnesses, like everyone else who does useful work for Steem in the ecosystem, are already being generously rewarded by the posting reward system. All that witness pay effectively covers that the posting reward doesn't cover is the cost of running a node on a dedicated server, keeping updated with release news, and upgrading steemd timely. That's certainly not worth 60k USD / month.
Bottom line: witness pay is the most absurd money grab in the whole system, and before trying to "balance" better the rewards between bloggers who for the large majority do actual real work, we should focus first on putting an end to this Steem Power hemorrhage that sees more than 1M USD worth of Steem getting pocketed every month by a handful of users whose only real added value is to run and upgrade steemd nodes.
Now how could we improve that? It's very simple. This is Steem and not Bitshares. All and every initiative including developments, marketing, tools etc. should be financed using the posting and voting mechanism. Under that assumption, there is no more reason to give such an extravagant salary to witnesses, and the funds can be reallocated to content rewards and the block reward reduced so that witnesses would receive a more reasonable pay for the job of running and maintaining a node. The witnesses who actually did stuff can now use the same fund raising mechanism as every other Steemian, and their "fame" as witness will surely help them receive generous rewards for their work.
This is, indeed, a huge reward. Perhaps we should consider shuffling the witness list every week or so? There are a lot of people, more than just the top 19, doing extremely valuable work for steemit who I feel should be rewarded for their efforts.
Or perhaps, as you say, we should consider reducing the pay and leaving the support to the reward pool, as witnesses discuss what they do, and people can decide on the value of their actions.
Anyway, thanks for the thoughts on this @recursive, valuable opinions as always. I'm saving my upvote for your comment for past 1st payout, to make it go on top again.
I think it would be optimal if the probability to produce a block , was related only with the % of votes each witness has! Like it happens for the witnesses after position 19 !
...read more here:
https://steemit.com/steem/@arhag/how-the-new-witness-scheduling-algorithm-works
Wo, I need to read more about witnesses.
Very Good
Elinor Ostrom won the nobel prize for her work on 'common goods' and how they can be set up to benefit all (or conversely how people can create a 'tragedy of the commons' that you indicate above). Perhaps it's worth the Steemit community reflecting on the principles she identified (and won the nobel prize for) and see how strong we feel these elements are represented in Steemit currently, to help prevent our own tragedy of the commons.
There are 8 principles which groups adhere to when the whole process 'works'
As you point out, there seem to be gaps between how the Steemit community currently works and how it needs to work in the future to navigate into a successful common goods platform - which essentially what Steemit needs to become.
https://en.wikipedia.org/wiki/Elinor_Ostrom
thanks for the article @anyx
I'm a fan of Ostrom, but I'm not sure how well her principles will work on the blockchain. It's really hard to build a longlasting and strong community when people can't see themselves face to face. Most users stay more or less anonymous. When people are anonymous, it's probably better to rely more on economic incentives than Ostrom's principles.
Perhaps using Ostrom's principles to inform economic incentives might help? You also made me think of Ernst Fehr work on 'Altruistic punishment' - where they tested economic principles to govern selfish behaviour - but using altruistic principles.
Hey bud. While they may or may not translate directly, it is valuable to give them the once over to see what might be able to fit. There isn't a foolproof blueprint for what is being done here and there needs to be give and take and adjustment.
We can see that economic incentives are not working as well as we would hope and I'm desperate to see Steemit succeed, so don't leave any rocks left unturned while looking for helpful solutions.
Another point came to my mind... Effective way of handling the tragedy of the commons with Ostrom's principles needs to have quite stable community. That is not the case with Steem. New users are coming in and some old users are leaving, so the community is under constant change. In a case like this, it's really hard to enforce the culture that is needed to counter the problems from tragedy of the commons.
Excellent principles. The key I think is going to be 7, since Steemit is going to be about self governance.
Thank you for the feedback!
this is the most interesting comment that I see so far.
here have true ideas to create a policy governing communities
You covered a lot! Spot on for your concerns. Even as a noob, I see it too.
Okay, you wanted input and a discussion: Here are my thoughts of the 3 problems you outlined.
I do believe many of these problems can be solved. The peer-review community and collective voting aspects are huge assets if employed correctly. But currently they are not. A million new minnows can outweigh a whale. Minnows don’t have any weight or reward for curation. Down-votes are not tied to expertise or authority. There are some problems which can be solved.
I wrote up a post with some of my crazy ideas Steemit Proposal for Developer and Community Evaluation https://steemit.com/steemit/@mrosenquist/steemit-proposal-for-developer-and-community-evaluation
Rules of Steemit. Without rules society de-evolves into chaos….then a strange thing happens. It tends to reinvent itself where individuals and groups begin to organize and defend themselves. This then forms a governing system and norms are established. I think the same thing is happening to Steemit. We are in the Wild West at the moment. The community is looking after itself and in the process establishing boundaries which must be tested. Plagerism a problem? Create @cheetah (good kitty). People not following the tagging rules, downvote. Some of these rules will be tested, debated, and even abolished. Eventually, if @Dan and @Ned don’t establish a formal set of laws, the members will create informal ones with creative consequences. BUT the key here is each person must have a voice. See my comments above about the need for every minnow to have some influence and reward for their curation work.
Transparency. I am on the fence about this one. To be transparent (yes, I did that) I do not have sufficient knowledge of the challenges, opportunities, and risks to form a respectable opinion. So I will do the noble thing and just remain quiet until such time I can contribute something worthwhile.
This is a huge problem. Furthermore, the reward pool doesn't pay out anything less than 2 cents it seems. So if you get a cent on a comment, you don't actually get it, how crappy is that?
And indeed, we are in the wild west and that is exactly why I created @cheetah.
https://steemit.com/steemit-ideas/@liondani/don-t-remove-curation-rewards-just-stop-them-when-the-post-reach-the-trending-page
This is a good idea! Alternatively to once it reaches the trending page, we could also cut off the reward after a certain reward value.
Then mention the idea on your post maybe ... It will be nice it get noticed from steemit key developers!
Added! This is also now a trending comment, so it should get noticed! :)
Yes indeed! But it must be a dynamic number, nothing static!
Private ownership is usually great answer to the tragedy of the commons. In a blockchain world the best solution is to change the incentives (what blockchain is encouraging to be done by paying for it) and UI (users don't necessarily want to make great effort to do something even when it rewards them more).
The blockchain reward system is quite good right now. I wouldn't do any drastic changes to it.
The UI is the thing what should be adjusted. The reason why some posts get a lot of votes is because they can be seen in the front page. If they were not on the frontpage/trending, they would still get a lot of votes but less than they get now.
Frontpage is creating too strong positive feedback loop for some posts that get a little bit of initial attention from the whales. When a post gets into frontpage, it's pretty much guaranteed to get a lot more votes.
My solution would be to individualize the frontpage for every user (maybe a combination of their feed and recommended posts that they may like). Trending page should be a little bit harder to see (a couple of mouse clicks) so that users wouldn't spend so much time in there (the feedback loop has to be made weaker).
There is a downside to this. Posts with large(r) payouts need to be seen so abuse or risks of abuse can be detected. Yesterday I saw the post for the iOS app and it had already reached over $10000 after several hours with no verification whatsoever. I only saw this post because it was on the front page.
It was not an author or topic I would have followed (the author was brand new anyway). After I posted a comment about the lack of verification, the post dropped as low as $1000. In this case, the author of the post then provided convincing verification and the post ended up climbing back to around $10000. But it could just as easily been the case that the post was a fraud (and there have been some high profile frauds on Steemit) and was only caught before a $10000 payout because it was highly visible.
A counterargument might be that if the post isn't visible then it wouldn't get such high rewards in the first place, but I question whether that is a good thing either. Once the post was verified, people found it worthwhile and voted it back up. Having worthwhile posts not get rewarded for lack of visibility also seems bad.
I don't know the answers here, just that there isn't one easy answer to all these issues.
Yeah, you are absolutely right that trending page is an important tool. I'm just trying to say that it shouldn't be something that users will always see first.
There should be more pages that show potential problems. For example, a page for posts/users that are gaining lots of flags.
Smooth, that could be as simple as you (and anyone with a hammer as large as yours who need to see things like this) setting your bookmark to the "trending" page.
Meanwhile, the default page of the site is changed to naturally land on "new".
Great feedback! The feedback loop on the tending page is very real.
I agree that the UI needs a lot of work, especially with search and organization in order to help seek out hidden gem posts.
I'm assuming that at first glance the "Trending" front page was meant to be a draw for new users. They would see the high dollar posts and think, "Wow! That could have been me!" The funny thing about social media is plenty of things go viral without a monetary incentive. Some people actually do engage with social media for the enjoyment factor.
Facebook has close to 1 billion active accounts and doesn't pay their users a dime. It filled a need in a space where there was a want. No one signed up to strike it rich on Reddit, Facebook, or YouTube, and yet some people's claim to fame came from just those very places.
Hi @samupaha, whilst not disagreeing with you, it would be great to hear why 'Private ownership is usually great answer to the tragedy of the commons' - thanks
The basic idea is to eliminate the commons. If you have a piece of land that nobody owns, the incentives can be for everyone to exploit it in every manner and to a degree that the land is ruined. If someone owns it, the owner won't want the land ruined and will most likely define rules and limits governing its use.
Not all commons can realistically be converted to private ownership (example: air), so this approach isn't always a useful one.
I see thanks. As we know Ostrom’s achievement was to effectively answer that private property isn't the only means of protecting finite resources from ruin or depletion. So I guess the one main question here is if any of Steem is commons or not (including social elements like trust in the system itself). And if it is, we'd perhaps be wise to learn from these kind of principles in action and if not, have an effective transparent model that we can all believe in. Thanks for finding the time to response @smooth.
Creating voting etiquette will always only go so far, i think solutions have to be in the structure of the economy.
Right now i think the economics of post reward rounds are a bit flip flopped. The initial phase has too much weight, too much power. It should be a test phase that doesn't consume so much of the networks inflation resources. This should be a test phase for posts that leads to a longer more thought out payout period.
I think that the first curation round on posts should be slower not in time, but in monetary reward. Meaning there should be a lower percentage share given for the initial payout round of any post, and that the 30 day round should have a greater potential payout, or a larger share from an upvote compared to the initial round.
This would slow down the hype train that happens around voting, a race that rewards selfishness rather than quality. If more is allocated after the initial test for any blog post, this would allow for more long term attention to high quality posts that take time to be digested, and responded to and more carefully curated.
Very interesting idea. This could disincentivize bot voting in the initial round, as the reward is much less, and perhaps make whales much more conscious about their votes.
Well Anyx, her is what I think:
I cannot comment the witness section (well i can, but my knowledge in the matter is just not enough).
As for the article section, I have the same opinion. I write ONLY original content myself. And sometime seeing news psots or posts that have been obviously came from the Wikepdia is upseting me.
For one, its useles, and causes more minnows to get less rewards
For two, its unfair when by copying a post, you suddenly get reward (it happens), and an original writer sometimes gets very little.
Im not against all the pictures of the kittens sleeping and so on, as long as its the authors pet that is been posted. Nothing wrong with that, after all it is a blog.
Though to make this a rubbish bin for copy paste is not an option
I definitely agree. Every time someone is rewarded for a garbage copy paste post, this is disingenuous to someone who spent hours on making something original, and will turn them off from the site -- leaving only garbage posters to post in a garbage bin.
Just as we speak, i realize for one more type of posts, i am personally agianst.
They are written by the author, but its things like "happy birthday to kylie minogue" or "compare the childhood pictures of ant and dec" and so on.
Once again, its just almost copied (as the information in took out from a source), and once again it makes people not wanting to post for no reward.
How do you think we can influence this?
People are doing this because they see it was rewarded highly in the past. We need to influence people, to be careful with content curation, and have whales follow in the footsteps of @smooth with manual curation to find gem posts.
Alone, we don't have much influence, but if the community shouts loud enough people will pay attention... I hope.
The real issue is that not much curation is actually happening.
Minnows vote - but mostly because they want a reciprocal vote, it is a way of getting attention. They can't deliver any monetary value with their votes. The whales vote for about 20 authors all the time, and no-one else. And the growing dolphin class don't bother at all. They could deliver a few cents to writers, and it would add up if lots of them did it - but they feel they don't get much curation award from it, so why bother - they are better off spending all their time writing for the whale-bots.
Ideally you would have the dolphin class do a lot of curation just to ensure the health of the system - but maybe they think it will fail anyway, so why bother, focus on writing as much as possible and reaping the awards.
Right, there are simply not enough dolphins, and there is simply not a good enough incentive for dolphins to 'properly' curate content.
Why should a dolphin upvote some article, when they could instead use the vote power to upvote an article they know whales will upvote, and thus give a massive curation reward?
Then indeed, they move on to trying to reap rewards... causing a selfish behaviour.. and perpetuating the tragedy of the commons.
Exactly... I feel that it is a game and i am losing it right from the start. I am a lousy gamer.
I think the curation rewards will be largely competed away and alternative voting strategies will surface, notably ones that reward finding new content.
I've been looking deeper into curation recently and can see that trend developing soon. It's far too easy to "front-run" when you have perfect information that those rewards will tighten until It's more profitable to find new quality content.
One way to improve the situation around selfish voting is to make the computation of rewards cyclical, and apply to the total curation reward for the cycle a weight that grows super-linearly as a function of the total number of votes, exactly like it's already done to encourage concentrating stake on the same account. That way, instead of concentrating their firepower to vote on only a few dozen "bluechip" authors, bots will have an incentive to spread voting large and wide to increase the reward weight. This can also be implemented without a fixed cycle: the system can keep paying the normal reward as post payouts happen (without bonus), and pay a bonus once daily based on the number of posts that have been upvoted in the last 24h and the total payout received for these posts.