You are viewing a single comment's thread from:

RE: Why I Advise Against Linear Reward

in #steem6 years ago

When you upvote you have a choice of:

  1. Upvote yourself (some people do this with literally empty comments; others are more subtle)
  2. Upvote some content at random or for some other non-monetary, non-value reason.
  3. Upvote some content on merit/opinion.

When you downvote, your choices are:

  1. Downvote yourself.
  2. Downvote some content at random or for some other non-monetary, non-value reason.
  3. Downvote some content on merit/opinion.

Review these lists carefully and then tell me whether you think it is more likely that people upvote or downvote based on merit/opinion.

Sort:  

Everyone has different motives. Users such as @v4vapid and myself tend to vote more to promote content - whereas many other users vote purely for financial gain. Steem presents a mixture of key selling points that both include financial gain and the ability to post transparently (uncensored) and in a fair way. We have here, therefore, a mixture of capitalism and the desire for free communication both competing for the same space. What combines all of these together is liberty, in a sense - or something approaching anarchy.

I personally upvote my own posts and then mostly only upvote posts based on content/opinion. I don't downvote due to disagreement (unless in extreme circumstances) and I don't downvote due to use of bots. While I might feel motivated to downvote due to the use of bots - I prefer solving the problem in a more uniform way instead of some bot users getting downvoted while others aren't.

In terms of the behavior of other users, I don't have accurate data on that - but I do know that there is a split between people who think that bid bots are 'valuable businesses' (lol) and those who vote based on opinion/content.

I don't think that really got at the point. Of course there are always users with different motivations. The common threads that runs through all are the incentives that are built into the system. People will respond differently to them but all respond in some manner. It's almost impossible, if not actually impossible, to participate in the system at all without doing so.

With that in mind I think should be clear from the above why downvotes are naturally much more effective than upvotes in getting voters to express an opinon, considering incentives across a range of personalities, motivations, etc. When you look at the current system where a very large amount of vote power (and a correspondingly large amount of the reward pool) is deployed in a manner that does not particularly (if at all) involve expressing opinion aka proof-of-brain, it is clear this outcome is tied more to upvoting than downvoting (aside from the overall scarcity of the latter).

I'll admit it, you have totally lost me here.. lol
I'm finding it hard to track your points - maybe I am missing something. From my perspective, the issue I want to see solved - and the issue that most people I speak to want to see solved, is that actual proof of brain is being replaced with 'proof of wallet'. So we have a situation akin to corporate mainstream media instead of 'the voice of the people' - which is, really by definitions, what people actually want/need. I don't see how downvotes can really achieve that without a coherent effort from the community to continually downvote - to the point of it being quite time consuming. The financial reward for those doing the upvoting (of themselves and co-conspirators) will motivate them to spend their time doing that.. but for downvoters, there is no guaranteed reward, yet they are using their time to 'police' things. From a 'business' perspective, the motivation and reward for abusing the situation remains higher than it is for trying to fix the situation with downvotes. When it is considered that downvotes can ALSO be used to aid in abuse of the situation, I find it hard to think they will be helpful overall.

On the other hand, by empowering users to be able to take direct and permanent action to remove abusers from their experience, we at least give users the opportunity to clean up their own feed and trending pages. This won't automatically stop the real abusers from still abusing the system, but it is easy to imagine a situation where abusers realise that their posts literally aren't being seen by anyone and thus the only benefit they get from posting and buying votes is whatever small return they get from the bidbots and nothing else. The system could also display the accounts that are being vote muted the most across the network - inspiring others to join in. I think this is a much more powerful option than is being recognised here.

I am all for just doing what works - if super charging downvotes works, then do that - but I feel it brings a disturbance in the force ;)

Let me distill it down to the essence. The goal of proof-of-brain is to get people to vote based on opinion. Downvotes can do that just as well as upvotes. Downvoting what you think is low-value is just the flip side of upvoting what you think is high value.

The difference is that upvotes can be turned into a cheat mode where you just vote to reward yourself, downvotes can't. The bottom line is that whatever else you think about downvotes, they are a much better way to end up with proof-of-brain that doesn't collapse into proof-of-self-enrichment.

But downvotes return rewards to the reward pool, which then increase your own payout. So even if you don't really disagree with people's posts, you might be inspired to use the 'cheat mode' of downvotes to negate other people's posts to increase your own payout. up/down votes are just different paths to the same outcome. the effect of downvotes is more diffused, since the rewards that are returned to the pool are distributed by all users and don't just go to the downvoter, but the effect on a logical level is the same (less payout for someone and more for the voter).. With sufficient SP held by a single account, downvoting could make a significant difference to the amount paid out to that account. This could be prevented by having rewards that are taken from posts via downvoting then being redistributed to somewhere else, such as a charity pool or a 'project budget' pool or something like that - instead of going back to the general rewards pool..

There are two differences:

  1. The effect is vastly smaller. Let's put some numbers on it. The largest account (excluding steemit) has about 3%. Which means (assuming 3% stakeholder is efficient in capturing 3% of the reward pool; not a given at all) 97% of a downvote by that account goes to others and 3% goes to the downvoter. With self-upvotes, 100% goes to the voter, so the effect is 30x stronger for upvotes. For everyone else the disparity is even greater. If we can fix 97% of a problem we should not dismiss that just because 3% might remain.
  2. Even downvoting with that up to 3% incentive in mind, it costs you nothing to also express an opinion in choosing which content to downvote. Unlike upvotes, the economic incentives do not require you to choose between your own economic benefit and expressing an opinion. You can still do so at no extra cost. (If you don't, you don't, we certainly can't force people but at least we're not paying them extra to undermine proof-of-brain.)

We could consider other aspects of it such as the fact that your own content might also get downvoted if others have a low opinion of it relative to its payout so you can't assume that downvoting other content necessarily increases the net payouts on yours (or that it even makes sense for you to post at all). That's not necessary though, since even without these additional considerations, we can already see that expressing an opinion via downvotes has much better incentives than doing so with upvotes, per #1 and #2 above.

redistributed to somewhere else, such as a charity pool or a 'project budget' pool or something like that - instead of going back to the general rewards pool..

That's vaguely plausible but I'm not convinced it is necessary. Stakeholders can use upvotes and downvotes to push rewards to charity payouts if that's where they want them to go. Expressing opinions via downvotes makes it far more likely this actually happens if people want it to.

Yes, the amounts involved in the gains from downvoting are smaller than from upvoting, but there are counter points here:

  1. I know from the interviews done by Tim Cliff that the biggest account holders here tend to have multiple accounts and some of their multiple accounts are large! So while the largest account may hold 3%, we don't know how many of those large accounts are held by any one user. They could have 10 - making 30%! How would we know? I'm not sure that we do - as long as accounts are anonymous. It would be interesting to see a breakdown of the largest accounts and which ones are anonymous or not proven to be held be a 'real' person.

  2. So, following on from point 1 - enhanced downvotes increases the gap between the rich and the poor, unless the rich are somehow going to become the super responsible content police and we have no reason to think that would happen... They don't even have enough time to do that. The amount of content involved means that this can ONLY practically be done by a decentralised method - such as with voter muting or something similar.

  3. In cases where users are posting a certain type of content, they may only have 2/3 competitors - so if they are able to downvote and remove their competitors consistently, due to having higher SP - then they very well could do that and the result would be a monopoly on that niche and also then their own gains would be signficantly higher than just the small gains made by returning rewards to the pool (As mentioned above).

  4. As already highlighted, there is more involved here than pure money. The issue of post discovery is involved and that comes in to the monopoly and free speech issue again. If new users come into this network expecting free speech (since it is one of the selling points) and find that their own viewpoint is unpopular among the prevailing SP whales of the day, then now their voice can be easily squashed and they will likely leave. At least when downvotes are tied to SP and there is a consideration of personal financial cost for downvoting, there is less motivation to act to silence the opposition. We already have a huge issue with user retention here and imo that is significantly due to the bid bots (and ninja mining etc.) - if our attempt to solve this does not take fully into consideration that the solution must not cause the problems for new users to just change form, then the attempts will probably fail. In this case, the problem of being 'out-gunned' and denied by those with the most SP (negating proof of brain) could, for many voices, just change form from not being able to be discovered due to the bid bot wars to literally not being seen by hardly anyone due to downvote domination. It would be nice to think that users would be smart enough to recognise the value of bringing in new users and having the network grow - thus learning and allowing respect for others users.. However, I know from 15 years of social networking and forums (including steem) that many people simply will not do that and will be anti-social regardless of the overall effect on the network. In fact, this is a potential attack vector by competing networks.

That's vaguely plausible but I'm not convinced it is necessary. Stakeholders can use upvotes and downvotes to push rewards to charity payouts if that's where they want them to go. Expressing opinions via downvotes makes it far more likely this actually happens if people want it to.

The point is that by forcing the changes introduced by downvotes to go to a third party or to projects that are meant to benefit steem, the actual downvoter does not personally benefit in a way that is out of balance with everyone else - the point is to demotivate them from gaming the downvote process.