Because of its money-based voting system, Yours.org already has the potential to be vastly superior to forums like reddit where vote manipulation and sock-puppeting are rampant. Communities built on this type of system are still vulnerable, however, to individuals who have enough funds to drown out the rest of the community's votes. (The smaller the community, the more vulnerable it is, and even the biggest potential communities could probably be out-funded by large institutions like governments or large corporations.)
Another layer of curation is required on top of the post-specific voting system in order to protect communities from manipulation by well funded meddlers. What's needed is a system that provides curation of curators. Such a system could be implemented on top of the current Yours.org platform with little to no change to the current interface and experience, where all actions on the site simply affect both the curation of content, and the curation of curators.
When a user votes on an article, that vote provides valuable data in addition to that user's opinion of the post itself. That vote potentially links the voter with the other users with whom he or she agrees about that post, at least as far as whether it belongs at the top of their feeds. Similarly to how one might find a movie critic with whom one agrees, and then weigh their reviews over other critics' reviews, that coinciding voting data can be used to build up a personalized vote weight system, one which is different for every user, where users' votes are weighed differently depending on coinciding voting history.
Given that premise, a preliminary question that might come to mind is "how much more should coinciding users' votes be weighed?" Should voters with one coinciding vote get double the priority? 1.00001 times? A hundred times? The best option will probably involve a gradually increasing weight, where the extra weight given to coinciding voters starts somewhere close to zero (though the best starting values are beyond both the scope of this article and my knowledge), and then the extra weight given to users with whom one's votes have coincided would increase as the user votes more (since, in doing so, the user would be providing more data on their preferences, allowing the system to more likely do an accurate job predicting their preferences based on voters with similar preferences). The weight could asymptotically approach a given max, or approach infinity.
The arbitrary nature of the specific default values for the system is, in my opinion, an opportunity to give the user control over their experience. A user might, at times, want the "top" of their feed to be limited to posts that the weighting system considers extremely likely to be relevant to them, only showing posts voted on by voters with whom the user's votes have coincided in the past. (That is to say, the user wants the system to weigh coinciding voters' votes infinitely higher than other votes.) At other times, however, they may want to open up the top of their feed more to generally highly voted content, perhaps to be exposed to potentially quality posts that just happen to be voted on by users with whom they've not yet had a chance to build a coinciding vote reputation. (That is to say, they may want to weigh all votes equally.) A "weight slider" could allow a user to adjust the degree to which the system weighs coinciding voters' votes. The recommended weighting, given the amount of voting already done, could be indicated by a tick on the slider, giving the user a reference point for where they started, and where they might want to return for general purpose browsing.
A middle ground between doing everything "under the hood" and giving the user a "weight slider" might be to allow the turning on and off of the personal reputation system, without giving users control over the weight that the reputation system gives to other coinciding voters while it's on.
Such a personalized vote weighting scheme would help to thwart users who might try to vote specific posts to the top for personal gain. Trying to do so without building up coinciding votes with many users by consistently voting on content would have limited value, given that the highly voted-on post would actually not make it very high in many users' feeds, or would require disproportionately more votes (aka money) to get there. Having to build up coinciding votes first would also increase the amount of time and money required to "force" a post to the top of other people's feeds and, given the implementation of a "downvote" button with a similar, negative effect on the weight system, such a user would, upon heavily upvoting their own personally-vested post, immediately lose the "reputation" they'd built up with other users when those other users downvote the obviously self-serving post, likely dooming it to the bottom of other users' feeds (which are affected by the downvoters' preferences) before it gets seen by many.
In the paid section are some fun possible additions to such a system.
 

$3.25
25.0¢

Reviews
3 of 3 reviewers say it's worth paying for

0 of 3 reviewers say it's not worth paying for
Comments
  earned 0.0¢
I like the idea, reminds me of the Google PageRank algorithm
0.0¢
   8mo ago
  spent 10.0¢
“When a user votes on an article, that vote provides valuable data in addition to that user's opinion of the post itself.” The vote doesn’t technically provide data about a user’s opinion. The fact is that no one can truly claim to know exactly WHY any particular person votes for a given post. It could be because they are ideologically aligned with the author of the post, or because they want to encourage a timid friend they know from outside the platform, or because they are impressed by it, despite the fact that it challenges them ideologically, or because they want to be included among all those who have already weighed in in favor of it, and not be left out. Or any combination of these. Or other possibilities. You have already made an unwarranted assumption. And you’ve only gotten started.
“Given that premise, a preliminary question that might come to mind is ‘how much more should coinciding users' votes be weighed?’” That’s only a preliminary question that might come to mind if we are already on board. As I’m not already sold on the premise, it isn’t given, and so will have to be argued for. You’re making the case for adding a complex new system for fixing a problem that may not even exist. As you yourself admit in the opening of this post: “Because of its money-based voting system, Yours.org already has the potential to be vastly superior to forums like reddit where vote manipulation and sock-puppeting are rampant.” You are essentially admitting that Yours is doing something right. Don’t forget that Reddit also began and was designed with the intent to make online content more fair and transparent. But things went wrong over time. I believe the road to hell is paved with good intentions, and so calling for needless layers of complexity can UNintentionally make things worse.
TO BE CLEAR: If there is an ACTUAL PROBLEM with the way Yours is STRUCTURED CURRENTLY which raises your concerns and urges you to craft a solution, that’s one thing. But if you are SUPPOSING that there MAY BE a problem in the FUTURE, one which, by all accounts, has not now materialized, and so you are advocating a solution to this potential, but as yet unrealized problem, then that is another thing entirely. Realize that when you propose actual, concrete solutions to combat potentially existing problems, you run a risk of creating more problems than you solve. I say, propose solutions to currently manifesting problems, not to potential or non-evident ones.
You speak of vulnerability of communities. Is it fair to ask, “from what aspect of the platform does this vulnerability emerge?” What would be your answer? Surely you know the answer, because you’re advocating solutions. What if the vulnerability is lower with a more closely connected community, and higher with a less closely connected community? Is that factor worth pondering? If so, then let me ask you this: How does your vote weighting system enhance the INTERCONNECTEDNESS of the community? Because it sounds to me that you’re advocating a layer of complexity specifically for the purpose of DOING THE WORK OF THE COMMUNITY, so that THEY DON’T HAVE TO DO IT, or perhaps, BECAUSE YOU DON’T TRUST THEM TO DO IT. Well, let me go on record as saying that I DON’T TRUST YOU to design a SUBSTITUTE MECHANISM that functions better than THE COMMUNITY ITSELF. I see your proposal as just that, then: a substitute for community participation. In other words, something that gets in the way.
What I would propose instead is a community driven solution, one which allows people to express more fully what they liked about something they voted for. So, for example, when they vote, they may have the option to go into a bit more detail about why they voted. Why do I propose this? Because a vote can be motivated by so many different things, making it hard to get a good signal from voting actions as they currently stand. So among the voters displayed at the bottom, there could be an associated emoticon displayed along with their name to symbolize the answer they gave to the following question when voting: What did you like about this post? Options could be: I agree with the author’s opinion, or I want to support the author financially, or I want to see their content promoted, or I was challenged by the author in a way that surprised me, etc. etc. These would be optional, of course, but they could provide extra information to those who are reviewing the post.
Could people lie about these responses? Sure. But I don’t think they’d likely lie as I really don’t see how that would benefit any voters. I think it most likely that they just wouldn’t bother to add these details when voting, which would still be an option. Also, by being publically displayed, it helps discourage recurring patterns of abuse. And besides, if we’re really that worried about lies and manipulation, then we have a larger problem elsewhere, with the cohesion of the community itself. And no solution conceived along purely technical lines can ever hope to fix that.
I’m not all that passionate about my alternative proposal, but I offer it only for the purposes of illustration, of how simplicity can work, and how designs should take the need for community participation into account and not do anything to squelch it. My proposal seeks to be simple AND still operate on the assumption that the community itself will be the engine which drives the platform, instead of some lifeless formula.
When you talk about well-funded meddlers, as if this somehow grave problem for this platform, I wonder if you think you may be on Steemit, where this really is a problem, and worse, is a constant source of drama, intrigue and whining. Here, everyone pays to post, out of their own pockets, which solves most of your worries right off the bat. As for some having potentially more influence by having more money, and that enabling them to, say, vote more than others for the same content: Is this causing a lot of trouble so far, in your opinion? And if its your view that it is, why is your solution not to simply limit every user to one vote per post? Why are you taking that aspect of Yours completely for granted as you proceed to propose extra layers of complication instead?
What you are proposing, knowingly or not, is the needless algorythmization of this platform, which is already what platforms like Facebook are plagued with. WE DON’T NEED ANOTHER FACEBOOK! WE NEED COMMUNITY!
I’m actually disappointed to see how much money you made on this post. That’s probably because of the paywall, for which I noticed almost half of the buyers neglected to leave any review. This observation may sound snide or petty, so I want to clearly state that if Yours were structured more to my liking, then I wouldn’t even know how much you made. I advocated some time ago for the earnings of authors to be a purely PRIVATE MATTER, known only to the authors themselves, instead of being auto-flaunted, even when that would be against the author’s wishes. It’s NONE OF MY BUSINESS what you earn. But, because I know about it, I can now be vexed by it, and so I feel compelled to respond.
At least it motivated me to write, I suppose.
For better or worse.
@Steve Patterson
0.0¢
   7mo ago
10.0¢
  earned 0.0¢
@Steve Patterson
Thanks for you in depth response!
>The vote doesn’t technically provide data about a user’s opinion. The fact is that no one can truly claim to know exactly WHY any particular person votes for a given post. It could be because they are ideologically aligned with the author of the post, or because they want to encourage a timid friend they know from outside the platform, or because they are impressed by it, despite the fact that it challenges them ideologically, or because they want to be included among all those who have already weighed in in favor of it, and not be left out. Or any combination of these. Or other possibilities. You have already made an unwarranted assumption. And you’ve only gotten started.
The fact that we don't know the exact reason why people vote on articles doesn't, I don't think, affect the fact that, by voting on an article, they have decided (for *whatever reason*) that it deserves to be higher in everyone else's feed. I make this point in the article when I qualify the statement you quoted in the very next sentence:
>That vote potentially links the voter with the other users with whom he or she agrees about that post, *at least as far as whether it belongs at the top of their feeds*.
If someone is voting on an article but doesn't think it deserves to be higher in other people's feeds, then they are not using the yours system as intended. And all my article assumes is that the voting user has decided that the article deserves to be higher in other people's feeds. (Again, for *whatever reason*.)
>“Given that premise, a preliminary question that might come to mind is ‘how much more should coinciding users' votes be weighed?’” That’s only a preliminary question that might come to mind if we are already on board. As I’m not already sold on the premise, it isn’t given, and so will have to be argued for. You’re making the case for adding a complex new system for fixing a problem that may not even exist. As you yourself admit in the opening of this post: “Because of its money-based voting system, Yours.org already has the potential to be vastly superior to forums like reddit where vote manipulation and sock-puppeting are rampant.” You are essentially admitting that Yours is doing something right. Don’t forget that Reddit also began and was designed with the intent to make online content more fair and transparent. But things went wrong over time. I believe the road to hell is paved with good intentions, and so calling for needless layers of complexity can UNintentionally make things worse.
Why did reddit go wrong over time? Because it was easy to game the system and, once it got big enough that it was profitable to game, it happened. Yours is inherently harder to "game", but this article highlights a *potential* vulnerability that I see left un-addressed. The fact that it's not yet an obvious problem does not imply that it shouldn't be protected against, the same way that the fact that reddit didn't always have the problems it has now doesn't imply it was a good idea to not prepare for vote manipulation and sock puppeting.
>TO BE CLEAR: If there is an ACTUAL PROBLEM with the way Yours is STRUCTURED CURRENTLY which raises your concerns and urges you to craft a solution, that’s one thing. But if you are SUPPOSING that there MAY BE a problem in the FUTURE, one which, by all accounts, has not now materialized, and so you are advocating a solution to this potential, but as yet unrealized problem, then that is another thing entirely. Realize that when you propose actual, concrete solutions to combat potentially existing problems, you run a risk of creating more problems than you solve. I say, propose solutions to currently manifesting problems, not to potential or non-evident ones.
As I said above, I recognize that this is a *potential* problem. But that doesn't mean we shouldn't prepare for it. If Yours.org becomes as popular as Reddit, I'm *extremely confident* that we'd run into this issue. The fact that solutions have the potential to introduce new problems isn't necessarily a reason to not implement them. It is a reason to implement them cautiously. And I'm not suggesting that Yours.org implement this solution tomorrow without any other thought. I'm suggesting that they strongly consider such a system and, in doing so, of course also scrutinize it for possible problems it may introduce. And even though you can't possibly be sure you've predicted all possible problems, you can say that the cost/benefit analysis leads you to believe it is or isn't worth the risk. I'm of the opinion that, unless someone can come up with a specific way that the system I propose would create more problems than it solves, it should be strongly considered.
>You speak of vulnerability of communities. Is it fair to ask, “from what aspect of the platform does this vulnerability emerge?”
Sure it's fair to ask, but I think the article addresses that question doesn't it? The aspect of the platform from which the vulnerability emerges is the fact that posts' positions in your feed are determined purely by monetary voting which, in spite of being superior to Sybil-able sites like Reddit, is still manipulable by someone with more money than the general voting body.
>What if the vulnerability is lower with a more closely connected community, and higher with a less closely connected community?
I don't understand your "what if". How could the issue I'm talking about be affected by the closeness or connectedness of the community? I'm talking strictly about the amount of money people have to throw around, and the fact that, if you have more money than someone, then you have more say on this website, which gives you more power. This fact could be countered by a user-specific vote weight system, so that someone with more money than everyone else who tried to vote a post to the top of everyone else's feed for personal gain would have barriers stopping them from doing that. And again, just to be clear, the vulnerability doesn't have to be being exploited now to notice that it could be in the future, and I don't just think that it "could" be. If Yours.org gets popular, it almost certainly would be.
>How does your vote weighting system enhance the INTERCONNECTEDNESS of the community? Because it sounds to me that you’re advocating a layer of complexity specifically for the purpose of DOING THE WORK OF THE COMMUNITY, so that THEY DON’T HAVE TO DO IT, or perhaps, BECAUSE YOU DON’T TRUST THEM TO DO IT.
It's actually the exact opposite. What this system would do is connect you with people with whom you've voted in agreement, and let you trust them more than the people with whom you've not voted in agreement. That way, even a post that got only a couple votes would make it to the top of your feed, as long as the couple votes came from voters whom the vote weight system determines you're more likely to trust, again, at least as far as what you both think should be at the top of your feeds.
>Well, let me go on record as saying that I DON’T TRUST YOU to design a SUBSTITUTE MECHANISM that functions better than THE COMMUNITY ITSELF.
What I'm suggesting would not be a "substitute mechanism" to the community itself, the same way that going from a standard time-sorted forum to a vote-based site (like reddit) does not mean you are "replacing" the community with the voting mechanism. Instead, the intent with reddit was to *enhance* the community via the ability to vote. And similarly here, my intent is to enhance the community by adding another layer on top of the voting mechanism which I think will allow the communities to function *even better*.
>I see your proposal as just that, then: a substitute for community participation. In other words, something that gets in the way.
Maybe, if it turns out to not work the way I think it would, that would be the case. But I'm pretty sure that that's not what would happen. I'm pretty sure that it would enhance the community, not get in its way. And unless you can explicitly how you think my system would "get in the way", I'm inclined to think you're just reacting negatively to the idea that an "algorithm" would affect your feed. But the algorithm isn't doing anything other than connecting you with other voters with similar taste. It's not relying on an algorithm so much as it is relying on the human element of shared tastes/preferences.
>What I would propose instead is a community driven solution, one which allows people to express more fully what they liked about something they voted for.
There are several problems with this approach, and I'll go into them below:
>So, for example, when they vote, they may have the option to go into a bit more detail about why they voted.
This is a problem becuase it requires extra effort on the part of both the voter (they have to write about why they voted) and the people who would take advantage of such a system (because they'd have to read what other people wrote.) That could probably work for small communities but, once they get big enough, then everyone's comments are drowned out by everyone else's. It's like when there are 10,000 reviews on Amazon for a product. There's no way you can actually get a relevant sampling of the general sentiment without investing a disproportionate amount of time compared to what it's worth.
>Why do I propose this? Because a vote can be motivated by so many different things, making it hard to get a good signal from voting actions as they currently stand. So among the voters displayed at the bottom, there could be an associated emoticon displayed along with their name to symbolize the answer they gave to the following question when voting: What did you like about this post? Options could be: I agree with the author’s opinion, or I want to support the author financially, or I want to see their content promoted, or I was challenged by the author in a way that surprised me, etc. etc. These would be optional, of course, but they could provide extra information to those who are reviewing the post.
This has the same problem as above: extra effort for the voter and reader which would eventually get drowned out by the large volume of voting. And how do you know you'll agree with the other voters' opinions? Maybe 90% of other voters have completely different taste than you?
>Could people lie about these responses? Sure. But I don’t think they’d likely lie as I really don’t see how that would benefit any voters.
Lying could definitely benefit the voters if the voters are personally invested in getting the post to the top! (Maybe they're all employees of a company that's featured in the background, for instance.)
>I think it most likely that they just wouldn’t bother to add these details when voting, which would still be an option. Also, by being publically displayed, it helps discourage recurring patterns of abuse.
How would their being publicly displayed help discourage recurring patterns of abuse? How do you expect to verify that someone is lying about their details, or recognize any abuse at all from regular use?
>And besides, if we’re really that worried about lies and manipulation, then we have a larger problem elsewhere, with the cohesion of the community itself. And no solution conceived along purely technical lines can ever hope to fix that.
That's a pessimistic view with which I disagree. Lies and manipulation will always exist, and especially when it is profitable. So the question is: can we make it not profitable? The intent of this article is to suggest a way which might help make lying and manipulation not profitable, or at least less so.
>I’m not all that passionate about my alternative proposal, but I offer it only for the purposes of illustration, of how simplicity can work, and how designs should take the need for community participation into account and not do anything to squelch it.
The problem with needing "community participation" as you call it is that it can just as easily be gamed as voting on reddit is. You can't trust it because you can't verify it, and anyone can create a bunch of fake accounts to fake the "community" itself.
>My proposal seeks to be simple AND still operate on the assumption that the community itself will be the engine which drives the platform, instead of some lifeless formula.
Again, this "life-less formula" seeks to *connect* people, and let them reach content that they are *more likely* to respond to and interact with. Why you think that wouldn't be an effect of this type of system is not clear to me.
>When you talk about well-funded meddlers, as if this somehow grave problem for this platform, I wonder if you think you may be on Steemit, where this really is a problem, and worse, is a constant source of drama, intrigue and whining. Here, everyone pays to post, out of their own pockets, which solves most of your worries right off the bat.
As I mentioned above, the fact that it's not *yet* a grave problem for this platform doesn't mean that it won't be one in the future if we don't do something to prevent it. And paying to post does not solve all my worries, because my worries are about those who have enough money to invest in getting their posts seen, in spite of the community.
>As for some having potentially more influence by having more money, and that enabling them to, say, vote more than others for the same content: Is this causing a lot of trouble so far, in your opinion?
Same as above, not being a problem yet doesn't mean we shouldn't prepare for it.
>And if its your view that it is, why is your solution not to simply limit every user to one vote per post?
Because you can just make more than one account and then vote as many times as you want. That's the exact problem Reddit has.
>Why are you taking that aspect of Yours completely for granted as you proceed to propose extra layers of complication instead?
Like you mentioned above, I'm actually giving Yours.org lots of credit for already having the potential to be better than other forums. I'm just *also* considering ways that it might not have solved *all* the problems that other forums have.
>What you are proposing, knowingly or not, is the needless algorythmization of this platform, which is already what platforms like Facebook are plagued with. WE DON’T NEED ANOTHER FACEBOOK! WE NEED COMMUNITY!
You say it's a "needless algorithmization", but I'd argue it's a useful one! One that, like I said above, is personalized so that it strengthens the community, not works against it, because it's not really much of an "algorithm"... it's pretty simple really. It just takes note when your vote coincides with another person's vote, and takes into account the fact that that might mean you will be more likely to want to interact with the other posts that that user votes for.
I'd love to hear your thoughts on this response.
0.0¢
   7mo ago