decrease the percentage of reputation allocated to certification

Service: Narrative

This is going to be an unpopular subject, I suspect (with the only Narrator who would probably upvote it not present on the community forum as it seems...) - but anyway, this is what I think.

It seems that it is not true that reputation can never be bought, can only be earned. As far as I can see it, you can buy A THIRD of it for 15 USD. A third makes a lot of difference and how is it fair that users get it merely for proving they're unique human beings, without any consideration for the quality of content they bring to the platform?

There's been quite a lengthy discussion about the unfairness of the 15 USD fee but I don't want to address it.

I'm interested in the problem (as I see it) of a high percentage of reputation being awarded for pretty much nothing and my opinion wouldn't change even if the certification was free.

I'm not against the certification per se - I understand it's needed for restricting 18+ content and financial operations (tipping, withdrawing a larger amount of many) to counteract potential money laundering and circulating tips among fake accounts. This is fine. But why it has to have such high weighting in the reputation score is beyond me.

Well, I guess it's because Narrative wants people to certify and the rep boost is a really good incentive but it goes so against the advertised premise of the reputation earned by quality, that I can't believe hardly anyone is questioning this policy.

So my suggestion is: 0% to max 5% of the reputation score for the certification component.

(An interesting discussion on this subject is happening under this post: https://www.narrative.org/post...nt=49451319039510109)

Original Post

Activity Stream

Totally agree that the hight weight of certification makes no sense on reputation.

I wouldn't go as far as suggesting a 0-5%, but maybe a 20-60-20 scheme would already be a great improvement. Right now, conduct weights 10%, and certification a whopping 3x. This means that your content can - as it happened - violate AUP, and you'll still keep a reputation around 90. Reputation should depend much more on conduct. And less on certification, since certification by itself guarantees nothing, and it's unfair. You can have the worst conduct on the planet, and still, add a 30% to your reputation (quality is already easy to get, at least up to a point...) by paying and sending a doc, while an exemplary member is capped at 70.

An excellent elaboration on the topic from Denis Wallez who is not active on the forum, copied from the post discussion mentioned above, with Denis' permission:

"I've been vocal against the $15 price combined the 30% weight in reputation… but not against the 'principle' of certification (not even against the price, IF the weight in rep is seriously lowered). 

I've been vocal against the fact that, with a 30% weight, some extra "high rep" rights can never be obtained by content alone in this "content economy"… but not against the 'principle' of certification (not even against the fact that certification might make the extra rights easier to obtain: I'm fine with 'easier' IF such 'easier' can be justified somehow).

What I'd really like is a 'good' reason for certification, though (AND for the record, I think it's possible to find some)…
And THEN a 'corresponding' effect WITHOUT extra effects:
• If  it's to protect teenagers against adults (as it has sometimes been presented), if it's about 18+, the effect should be in blocking 18+ content, without any additional reputational impact;
• if it's a regulatory requirement for "$600+ p.a." redemption (to avoid tax evasion), the effect should be in blocking redemptions above $600 p.a., without reputational impact;
• if it's to help decide who to vote for in elections, without having to screen all the various candidates for the various niches one follows, the effect should simply be a flag displayed next to your profile, without other reputational impact (like higher rep or voting weights);
• if it's to avoid bots, I don't see why the vote by e.g. 20 narrators couldn't 'certify' that someone interacts like a human (funkily, this wouldn't even equate content quality!)(feel free to pick your threshold if 20 isn't your cup of tea)… Of course, you'll say that votes from dummy accounts could bias things: feel free to add the "Medium rep' requirement on voters, that has been created to deal with spammers.

The certification by vote could point toward a solution to the « they could still relate to each other but perhaps in a way that solves the concern of not wanting to penalise people who can't afford to certify ». Because there are a few people on narrative already where I'm pretty confident I could up-vote their certification without them having to pay for that. And there could be a possibility to 'appeal' (if an account becomes suddenly spammy) or even simply leaving the votes open (so a spammer gets voted down quickly and thereby loses the certification: it would require a constant flood of dummy up-voting accounts to prevent the down-votes, but if each dummy account needs "Medium rep" before it can vote then frankly I think it works fine to protect us!)."

I agree with @Vico Biscotti.

20 - 60 - 20 feels more balanced.  It feels wrong that the maximum penalty for misconduct is to lose 10 rep, while $15 certification applies 30 rep.  We could even look at 25 conduct - 60 quality - 15 certification.

As I understand it, the only functional reason certification boosts reputation is so that voting power of verified individuals is boosted.  This is an important systemic measure.  It could be kept.  Certification could continue to boost voting power by 30, or even more if that is what it takes to counter the effect of bad actors.

The computation of voting power could take into account reputation (with certification contributing only 15-20 percent), but also have an additional certification component.

In essence, this would be a partial decoupling of voting power from reputation alone, adding additional importance for certification, but only for voting power.  For all the Community knows, voting power and reputation might already be partially decoupled behind the scenes.  For instance, as part of the anti-troll measures, it would make sense that voting power be reduced until a new account reaches a certain age, giving time for it to reveal its conduct.

Thanks for opening the suggestion, @Gosia Rokicka - the discussion will almost certainly help the reputation system improve.  Thank you also to Denis Wallez, who has often raised the issue of certification since before the Beta launched.

 

Certification by vote is an interesting consideration, but it has one immediate drawback.  It would be very ineffective at detecting dummy accounts created by  a human. 

As far as I can tell, the only way to defend Narrative's democratic aspects from such 'people bots' or non-unique accounts, is to boost the voting power of accounts verified as unique.

I think certification should be a significant part of reputation...  but I strongly believe it should be free.  The $15 certification fee really soured me on Narrative.

"Narrative...  where reputation is for sale!"

If it's buying reputation due to the fee, is it getting free reputation, if there isn't one? Just sayin'. I like the ideas of replacing at least some of the boost certification gives rep with ways that demonstrate good reasons for increased rep, whether there's a fee or not.

Colleen Ryer posted:

If it's buying reputation due to the fee, is it getting free reputation, if there isn't one? Just sayin'. I like the ideas of replacing at least some of the boost certification gives rep with ways that demonstrate good reasons for increased rep, whether there's a fee or not.

Well said - and perhaps more convincingly than I have been trying to say it.  I've always felt that the 'buying reputation' angle was the wrong thing to focus on.  A fee for processing clearly isn't the same thing as buying reputation, even if it is annoying and bad from an accessibility standpoint.

Gosia Rokicka posted:

An excellent elaboration on the topic from Denis Wallez who is not active on the forum, copied from the post discussion mentioned above, with Denis' permission:

"I've been vocal against the $15 price combined the 30% weight in reputation… but not against the 'principle' of certification (not even against the price, IF the weight in rep is seriously lowered). 

I've been vocal against the fact that, with a 30% weight, some extra "high rep" rights can never be obtained by content alone in this "content economy"… but not against the 'principle' of certification (not even against the fact that certification might make the extra rights easier to obtain: I'm fine with 'easier' IF such 'easier' can be justified somehow).

What I'd really like is a 'good' reason for certification, though (AND for the record, I think it's possible to find some)…
And THEN a 'corresponding' effect WITHOUT extra effects:
• If  it's to protect teenagers against adults (as it has sometimes been presented), if it's about 18+, the effect should be in blocking 18+ content, without any additional reputational impact;
• if it's a regulatory requirement for "$600+ p.a." redemption (to avoid tax evasion), the effect should be in blocking redemptions above $600 p.a., without reputational impact;
• if it's to help decide who to vote for in elections, without having to screen all the various candidates for the various niches one follows, the effect should simply be a flag displayed next to your profile, without other reputational impact (like higher rep or voting weights);
• if it's to avoid bots, I don't see why the vote by e.g. 20 narrators couldn't 'certify' that someone interacts like a human (funkily, this wouldn't even equate content quality!)(feel free to pick your threshold if 20 isn't your cup of tea)… Of course, you'll say that votes from dummy accounts could bias things: feel free to add the "Medium rep' requirement on voters, that has been created to deal with spammers.

The certification by vote could point toward a solution to the « they could still relate to each other but perhaps in a way that solves the concern of not wanting to penalise people who can't afford to certify ». Because there are a few people on narrative already where I'm pretty confident I could up-vote their certification without them having to pay for that. And there could be a possibility to 'appeal' (if an account becomes suddenly spammy) or even simply leaving the votes open (so a spammer gets voted down quickly and thereby loses the certification: it would require a constant flood of dummy up-voting accounts to prevent the down-votes, but if each dummy account needs "Medium rep" before it can vote then frankly I think it works fine to protect us!)."

I'm wondering if certification by vote could be maintained fairly, especially once the platform is in full swing. There would have to be a lot of checks and balances.

A couple of ideas I think might be easier -  apologies if they've already been suggested -

A probation period  - can't certify until a certain rep level is hit (if fee remains).

or

Once certified, rep has to be maintained to a standard for a set time, or certification is revoked (if free). Member can reapply, but not until rep score has reached the standard again.

Or if left as is, maybe members could be awarded extra points over time up to the value of certification, conditional on good rep. 

 

Adding on this, we can also go in this way.

User should have to reach upto 50 score (medium reputation) by creating or curating the contents and only then he/she will be allowed to certify for free. Or else they have to pay the usual fees like we have currently. So this might can push them to work hard on their contents to earn their certification for free. A win win for all by such increased activities. But yes we need to place some policies here (like posting limits) so users don't spam just to earn free certification.

Add Reply

×
×
×
×