Topic: Trust System Upgrade - page 2. (Read 1155 times)

What is your point even? If they are smart enough to erase their trail completely then you aren't catching them anyway. Again, what evidence is available would be submitted for public review. Neutral ratings could serve as warnings with the thread referenced. I agree "untrustworthy behavior is incredibly subjective", which is why I am arguing for an OBJECTIVE standard of evidence before negative rating. Account sellers could again be neutral rated with a reference thread. This is not covering it up. This is defining a very objective line so that the community can focus on fraud, not absolutely everything being perceived as subjectively "untrustworthy". The negative trust factory farmers would just have to find a way to get their dopamine hits elsewhere rather than from dropping negative ratings on people assembly line style and fellating themselves over the moderate amount of authority it gives them.

standard of evidence of theft, violation of contractual agreement, or violation of applicable laws just means that people need to be smart enough to not leave certain bits of evidence behind while doing what they are doing. Not being able to leave someone feedback on a technicality doesn't seem fantastic either. In addition, untrustworthy behavior is incredibly subjective, so who will decide what warrants negative feedback?

If someone leaves a frivolous negative because they believe you are untrustworthy on the grounds that you bought your account for example. A fair number of people might find that significant and would rather it not be covered up because it isn't breaking any laws, agreements, and no theft was involved.

I have, several times. I will however explain again. Just as currently with any normal scam accusation a thread is posted. It is considered a general requirement to have some kind of substantiation when these claims are presented. What I am advocating for is that we follow this model and require AT LEAST a base standard of presenting SOME tangible evidence of theft, violation of contractual agreement, or violation of applicable laws. As is the standard now for scam accusations, the evidence would be submitted for public review. After the prerequisite of evidence of theft, violation of contractual agreement, or violation of applicable laws has been met, a general consensus would be reached much as it is now during public discussion whether negative ratings would be merited, and users would be free to make that decision individually.

Really it is very similar to the way it works now, only the standard of evidence of the objectively predefined offenses would serve not only as a limiting factor for issuing false ratings, it would again mandate accountability on those rating to substantiate their claims using objective standards of evidence. Anyone refusing to follow the standard of presenting evidence under objective standards would then find their own reputation suffering. This would, depending on the system in use result in exclusions, or negative ratings. If Theymos were to mandate this standard of evidence, it would then be a form a "terms of use" agreement, and those violating those terms would then be subject to negative ratings themselves. He nor the staff actually has to enforce anything, the simple act of setting the standard for the terms of use of the trust system would be sufficient, then the rest would be a matter of public review in the appropriate subforums.

Current system is working fine. Still now most of DT (new) members don't have proper knowledge about trust system. I don't think it should be more complicated. Perhaps we need to observe few more months about current trust system, then may be admin could picked few suggestions those are "good for forum". Likely theymos will not do anything now.

Forgive me if you've already answered this, but should this be implemented what would make people follow these rules? Would someone have to sift through every negative rating left and verify that each one is valid according to them?

While I agree with several of your conclusions here, and you bring up very good points about there being no limiting factor on frivolous negative ratings, I think your solution is lacking and is only adding more complication to an already over complicated system. The system needs to be simplified, not added to. My suggestion is to have a standard of evidence of theft, violation of contractual agreement, or violation of applicable laws before leaving a negative rating. This solves the primary issue of there being no limiting factor to false or frivolous negative ratings with minimal effort, creates a simpler system, and less counter productive side effects.

Well yea, you don't ever need to use any particular feature.


My understanding of the intention behind the current implementation of the trust system is it is intended to not be limited to those who you have personally traded with. If trust ratings were limited to trades, a scammer could simply never trade with someone on DT (or never scam someone on DT, but trade with them), and no warning would ever show, which on its face would be a bad thing.

This proposal is trying to solve the problem that a single person can effectively torpedo a person's reputation without any real recourse, even if many others on DT disagree. It effectively takes three people to override a single negative rating, and this number grows exponentially as each additional negative rating is given. There is a small group of people who frequently leave negative ratings together, that effectively cannot be overridden -- when 4 people leave a negative rating, it will take 17 positive ratings for for person to show as having not negative trust, and when 5 people leave a negative rating, 33 people will need to leave a positive rating.

The above group of people tends to leave very controversial ratings with impunity and without consequence, and this results in this group of people collecting additional support from others because they fear themselves being targeted by this group, and do not want to lose their reputation for arbitrary reasons (or for the underlying reason of opposing/pushing back against this group of people).

This particular proposal removes the ability to label someone a scammer for an arbitrary reason. If there is pushback to someone being labeled a scammer, it will take many people labeling the person a scammer, and if there is enough pushback, the person will effectively not be labeled a scammer.

I don't think someone needs to have actually scammed to be labeled a scammer, and it is sometimes appropriate to label someone a scammer if they are showing "red flags", however it is my belief it should be pretty clear cut if someone is being labeled a scammer. In any case, everyone will be free to review the text (and references) of trust ratings and come to their own conclusion if it is safe to trade with someone or not.

Not many people know how to determine who has included who on their trust list.

I am not referring to BAC, but under the old implementation of DT, you had at least one person who not only left many controversial ratings, but also refused to even defend his ratings, and chose to instead troll anyone who tried to defend themselves (which, BTW sent a clear message to others to not even try to defend themselves), and you were in no way held accountable. You were not the only one, Blazed had many people on his trust list who left many controversial ratings, and I am not aware of him being held accountable in any way. Both you and Blazed either ignored the problems, or pretended they didn't exist, and never engaged in any substantial discussion on the ratings of those in question. Going back further, CITM had a trust list that many people found very controversial for many months, if not over a year, and to my knowledge he did not encounter any problems in conducting business, and was only removed from DT1 when he left an indefensible negative rating for Dogie, who was able to concisely defend himself -- I have my doubts that the rating in question would have even mattered had it not been for dogies contributions to the forum.

So no, I don't think having someone on your trust list who leaves controversial ratings will reflect poorly enough on you to resemble anything close to you being held accountable.

In this example, everyone who has BAC on their trust list would need to exclude the volatile person, which is not ideal because not everyone will notice the problem right away.

A better example of this is probably Blazed, who as mentioned above, had many people on his trust list who had no business being anywhere near DT. I consider him to be generally trustworthy, have successfully trusted him with $x,000 in the past, and would probably be willing to trust him with $xx,000 if we were engaging in a trade that I believed to be mutually beneficial for the both of us. His ratings were similarly good. His trust list was not, nearly his entire list was compromised of people who I don't want in my trust network, and as such I eventually excluded him, knowing how the trust system worked and how to use it.

Like others have pointed out, the easiest solution is to advertise custom trust lists.

Those that come here and decide to stay can choose who they want to trust.
Don't scare away curious newbies by making them sign waivers, agreements and lawyer up before their first post.

Just like the merit system, those who have screwed over others will not like this system. 

That was my dilemma, there are folks who's reviews I would like to see, but not necessarily give them my vote for DT1.



I thought of this solution, and even created an alt for that purpose, but I have not implemented it yet.

You can add users to an alt account, then add that alt account to your Trust list. This way, you trust them one level less deep.
I haven't used this option yet though.

I agree the current trust system is complicated enough already, if not too complicated. Promoting custom Trust lists is more important now. I've suggested a big banner like the April Fool joke before, that had a lot of response in just one day.

Your questions seem perfectly reasonable and intellectually honest. I would again however refer you to the linked posts from my last statement above for review, but in brief....

Q: "How do you see that implemented ? Who would apply this tag to scammers ?"
A: The same people who do it now, only they would need to provide substantiation to their claims rather than it being a subjective free for all as it is currently.

Q: "How do you verify the violation of applicable laws ?"
A: The same way the forum currently addresses scams and fraudulent activity (other than getting out the red paint bucket), via threads in "Scam Accusations" and "Reputation" to submit the evidence for public discussion and review.

IMO the problem is Theymos is simply hesitant to put his ego aside and admit his little social experiment he has been subjecting us all to has been a total failure, and some of the people he has been most critical of are providing legitimate solutions. That is a difficult thing to admit when you are wrong some times, especially when you went out of your way to punish those that tried to warn you for the act of pointing those issues out. I am sure he feels it would degrade his authority, but of course if this place is a giant shit show he is not going to have any authority or respect from anyone anyway. I think he is underestimating the positive impact of meaningful reform, and the resulting support that will come of it and just hoping he will never have to make that hard choice and go through that process.

Set trust depth to 0 and explicitly add all users you want on your trust network.

When the changes happened in January I also thought we might want separate lists for users whose trust scores we want to see, users whom we want in our trust network, users to "vote" for in DT1 election... but that's not the intent of the trust system, is it? We should add users whose judgement we trust and the other stuff is just sugar, mostly to make it easier for newbies to find their way around. More complexity creates more room for manipulation and we definitely need less of that.

Yes, I agree with you, the trust system is not perfect and subject to improvement. And yes, we don't need more staff intervention.
I hear your alternatives, but these are my questions about your alternatives, and would like to know how you would address them.

When you say :


How do you see that implemented ? Who would apply this tag to scammers ?
Is it the staff ? Or the tag appears when the trust score would have been -16 ?

How do you verify the violation of applicable laws ?

I am not trying to annoy you, but rather trying to picture how it would work.

---

To be honest, I think that the trust system has flaws, but I also think it is something very difficult to manage and Theymos will always try to do his best to make things better.

I'm not seeing your example there, isn't that the accountability we are talking about? If BAC leaves controversial ratings, it would reflect poorly on me as me lending my trust in his ratings. As a result, I should remove him from my list, or others will doubt my ability to be a judge of ratings and exclude me. That is a feature we want correct?

As for the ratings only part, if BAC has someone on his trust list that I think is too volatile to be on DT, the process starts again, the individual gets pressured, BAC would get pressured, and then I'd get pressured in that order if nothing is done about it. In that case, I could exclude that person and it'd be settled.


There are a few features that I'd like to see changed around a bit, but I still think the biggest problems with the trust system are interpersonal problems. Its like inventing the chainsaw, and then deciding, yeah this technology is no good, the people are juggling with them instead of cutting trees.

What I am arguing for is inherently SIMPLIFYING the trust system, not complicating it. We don't need more staff intervention. I am telling you the alternatives right now as you do your best to ignore and dismiss them.



You are right. It is super easy to read red and green numbers. The problem is they are often not accurate and based on personal vendettas or retaliation for criticism and not on actual fraudulent behavior. This is what new users will not understand, and this is the issue. Saying that it will be fine after they are around a while is kind of like saying "oh its ok if Timmy plays in traffic, after a while he will learn that it is not a good idea!" Your statement inherently excludes new users, the very users it was intended to protect. This is not logic but the superficial appearance of logic.

I have explained this again and again and again and again and you all feign ignorance over and over pretending as if I have not already responded to your criticisms. This will require LESS staff intervention, will provide a more equitable trust system, prevent large amounts of abuse and fraud, as well as most importantly defuse conflicts BEFORE they happen, not after they are already a clusterfuck.

I've thought of this option, and I do think it can be helpful.  It would be nice if we could add a custom depth to each of inclusions, but like o_e_l_e_o mentioned, it would add complexity to an already complex system.  Maybe a custom depth setting can be implemented as an option, and not a requirement.




This is one of the philosophical debates about the current system; is it a "trader" rating or a "trust" rating?  Should it be limited to those who've completed trades with one another, or is more than that?

It's my opinion that the trust system is designed to be more general and broad, and therefore more useful than merely a "trade" rating.  There are many attempted scams that don't necessarily involve a trade of goods or an exchange of currency.  And anyone could simply start trading in rinky-dink transactions that amount to only a few dollars and still earn a high trade rating.  If it was limited to only trades, then every one who dealt with CanaryInTheMine would like a trusted individual.  We know that's not necessarily the case.

I don’t think the current implication is working. This is for many reasons.

My proposal should completely eliminate any trust disputes for controversial reasons. If someone starts leaving inaccurate ratings with inaccurate comments, they can be blacklisted.

It is only possible to exclude someone from your own trust network. This will do nothing to address how others will view trust left by this person. As mentioned, the majority of businesses use the default settings so any rating left by anyone in DT, and if considering doing something such as giving an unsecured loan, looking at these ratings is logical.

Further, it is not possible to exclude an individual rating because all ratings are calculated in trust scores.

The calculation for if a warning is displayed doesn’t need to be done by users and I don’t see any major drawbacks if individuals don’t understand this.

This is exactly what we don't need, the 2 things the world already has too much of is lawyers and politicians.

Ironically tecshare's stupid idea would require an entire "justice system" be installed, were going to need lawyers (prosecutors and defense), judges, appeal judges, supreme court justice's, jury's, grand jury's, Attonery's General, an FBI cybercrime division plus administrative staff to handle all this justice!  We need lawmakers now to define the laws.  Fuck ya all these rules sound like awesome sauce and so much like what the forum is all about.  I mean the forum motto is MOAR RULES PLZ after all!

inb4 "poll manipulated by gangs!!!"

I would love to see examples of the suggested trust score formula applied to the OP and to some other individuals it appears to be designed for.

Would people not be able to put user into distrust list Lauda and his group would quickly lose power.

You will never see them agree on that.