Ratings site linked to by Lindens in blog is shady to say the least.

*scratches head*

well i see this here linky at the very first post that leads to a blog which lists several sites (which I guess are supposed to be shady) one was banlink which I have no idea if its shady or not but the subsequent posts talked about banlink which was listed in that there blog post

so having not experienced the other stuff and whatnot its as good as anything to discuss on a thread about sites which keep track of people who did "stuff" hehe

if not then i'm puzzled too

*scracthes head some more*

*wanders off to look at shiny things*

course then there is this other link here

http://meratalk.com/blog.php?p=304

named mera talk which says banlink goes linden and well mera pixel helped create banlink with travis lambert so i put mera and meratalk together and got something or other that says

that "banlink goes linden" from a site named Meratalk (or something like that) lol

*oo shiny*

Yeah,Wilhelm. Agree. Before I start to look for shiny things too, should we all then just set Banlink aside and resolve to forget all the others listed in the link from the first post even exist because one of them (at least) is allegedly "shady?"

"Shady" I'd be delighted to help nail, skin, and cure in the sun on a south-facing wall. Unshady, I'd rather not pick on. But the others, besides Banlink, all seem to be about "reputation," and while a couple of them do strike my as hype-ific, it's not immediately clear to me who might be "shadier" than whom. I mean extortion is a pretty strong word, ya know...

After looking over the other sites mentioned ...

Ratepoint - looks like a popularity contest

Real Reputations and Trust Net - Look like organized Gossip

Sloog and Slicr - are just tallies of people saying certain places are good. The numbers that make a place popular are laughable. Sloogs most popular is 81 people, While Slicr's is 6.

Ban Link also looks like the only system that is developed to a really functional point.

I still dont see the need for any of them.

Lets not loose sight of how its actually used.

We (luskwood) are part of this system... we trust a very very select few other places (just 3-4 of them) and there are a select few places who trust us.

If someone has attacked one of those places sufficiently to get 'enrolled' into the system, innately nothing happens, other than they are banned from the place that put them in. There is no 'automatic' ban from all of sl etc from this, its just the place that put them in, and they'll be booted from there.

Bans are entered with reasons, for lookups so people don't 'fall through the cracks' at all, thats just silly. Now, if the place that put them in is trusted by us, and as i said, only a couple of places have earned that level of trust with us, they will not be allowed on our land either, and if they are bounced, they will be told why.

If someone habitually bans lots of random people, they will not end up on trusted lists, and those bans will never 'spread' anywhere. Its an opt-IN style system, we must elect which bans will carry over onto our land, a random ban from a random member will *NEVER* affect anyone trying to enter luskwood.

Likewise if we have a problem with someone, and we put them in, and we do this *very* infrequently... that person will not also be 'banned from sl' or even a large majority of it... they will only be banned from a very few select areas which know us and have elected to trust our judgement.


If someone is banned in this way, they *do* get a chance to appeal to us, we *will* listen to them, and if they are sincere, they will be removed from the list. Its really that simple.

If someone on our trusted list does NOT remove bans when asked like that, chances are they will not be on our trusted list for very long.

If there is a specific dispute, we can elect to 'white list' someone, even if they are banned from a trusted place, so that they can come into our land, and we have done this several times, though it is generally not our preferred method... i.e. we much prefer it if the dispute is worked out with the group who initially put them in.


Most of the complaints and arguments against the system have been incoherent, wrong, misinformed, or just flat out lies. The purpose of this system is to make managing a social area in SL slightly less onerous for all of us involved. 'vendettas' etc will not work because people will become untrusted.

Also... one thing to keep in mind... what was just a little laugh or prank for you, could have deeply hurt someone else, and as i mentioned in a previous post, that is something all of us are accountable for in the real world, and its been long since overdue since that kind of accountability was present in SL as well.

If banlink is abused by people, it will not serve its central purpose, hence, it is honestly *not* abused by people, since those involved have a very deep seated interest in making it work well for everyone

I don't see the need for any of them either.

coco

The clue is the link in the very first post that leads directly to a comment about the service in question. Not too hard to figure that one out really.

okay so its not banlink and has something to do with a link one of several in a blog? are we speaking of the one by shaun altman? err real reputations? i dunno stabbing in the dark here but?

If its that one I sent a notecard telling him I would not wanting to be participating and to make sure I didn't end up on his rate system and kept a copy of the notecard etc. That one is not clear and he never answered. Some are illuding to the fact that you can be rated in there but if you want to deal with your rating or see it you have to pay to get in or something. (which of course is not quite well legal in many places if you create records that are for public usage you can't charge peopel for a service they didn't want to be a part of in order to deal with it if its not opt in. This wont last if this is true. If he chooses to rate people without consent and wont let them see the record without paying there is an issue here, but again I actualy didn't see him confirm that)

In any event he doesn't have my permission to charge me money so I can see myself on a system or put me on one. (if that makes any sense) Anything that is not opt in will not get much support from me. If its not opt in okay but you can't charge for random ratings. Only way you can charge is if it IS opt in.

Now I"m talking in circles because this entire thing well half of them make very little sense and aren't fully thought out.

How do you charge for a service and keep informationon people without their consent and if somehow you collect their info and they want to see it charge them to see it? Can't be done hehe the words extortion and fraud are the first things that come to my head and they get worse from there

err and why is it that no one will use the name and say "figure it out from the blog" or whatever is everyone afraid of a rating system?

If he charges fine I'm actualy not to worried about a rating system (I have no use for them) but I am worried about people saying lies and making libelous statements etc (slander etc). Can't really start a rumour mill of negatives about people tell them they are oon there and having hanous things said about them and then say "no you can't deal with it till you pay me money" in any event such is life. If it is indeed shady we are not sure yet. If its not then well its just a rating system I"m not interested in. Right now no opinions can be formed because I have not seen anyone experience it nor have I experienced it myself cause i'm not interested in rating systems..

*passes out from run on thoughts*

Don't think I'm going to spend much more time on these other rep-rating schemes. On a cursory look they all seem not worth bothering with.

As they say in the Middle East: The dogs bark; the caravan moves on...

I think my system is well thought out (although I'm biased), although it's not very easy to explain. The problem is that if I made it simpler it wouldn't be nearly as good.

If you think there's something wrong with it, however, I'd like to hear about it. I spent a good deal of time thinking of how to design a system that's reliable and abuse resistant, and will appreciate good feedback on it.



Because it's very hard to present a coherent argument against all reputation systems. To argue that nothing of the sort should ever exist you basically need to say that nobody can ever declare their opinion about anybody or anything in public.

The moment you express an opinion about somebody, or make a list of bad places, you're doing the very basics of what a reputation system is, only with less automation.

People at least seem to realize the irony of protesting against BanLink by effectively creating something quite similar to BanLink, and try to wiggle out of it by hiding their opinion. I don't think that changes things much, as the fundamental nature of the thing remains the same.

Allana's original post has a link which points to another post of hers. It describes a firm that requires you to pay a premium membership fee to view the reviews and comments made by others about YOU. Sooo, may or not be construed as a shady practise. You dont have to pay the premium membership if you dont give a damn about what others are saying about you. But if a dozen people IM you and say, " MAN, YOU SUCK! IVE HEARD FROM A REVIEW SITE THAT YOU ARE NOT TO BE TRUSTED AND THAT YOU ABUSE SMALL ANIMALS AND YOU GRIEF AGEPLAY SIMS!" Then your gonna be reeeeal interested about whats being said on that site and thus your curiosity will force you to pay the fee. If your being reviewed on some third party site, I think you have the right to see what is being said about you, without having to pay to see it. Charging a person for the privilage to defend themselves isnt right. If a person is being reviewed on a site, they should at least have access to see their own portfolio, and a way to reply to commentary. Its your reputation on the line, and you should have the abilility to address whatever comments are being made about it, even if you are locked out of the rest of the site untill you join up and pay the admission cost.

I agree there. Which is why I don't do that. TrustNet costs money (with a free trial) because it's rather more resource heavier than simpler systems. But paying for it only gives you the ability to rate and see other people's ratings. Being rated is no different with or without a subscription.

I was originally confused because the original post merely links to the blog post announcing the end of the ratings program. But sure enough, Allana posts the message you describe. It's a comment to the blog post; look way down in the 300's. Or do what I finally did: ctrl-F and search for Allana.

I suppose the thing that makes me most concerned about BanLink, Travis, is the fact that I don't think your flavour of it will be the one that dominates the grid some years from now.

Trust networks ultimately fail, if in fact there are two levels. In this case the 'admins' and 'everyone else'. Which is what we have here. Not in the club? Not a friend of anyone in the club? Then you better watch your step.

So what keeps BanLink on the up-and-up? One factor: Travis himself. It's the human intervention that will somewhat prevent this from being gamed (as long as Travis runs the dominant version of it). Maybe.

Travis governs the process, and he's a good governor. But make no mistake, he's key - without him, system abuse would be answerable to... nobody. Luskwood is also a bit of a special case. Here we have community members that deeply care about this sort of thing, have been a constant target for years, and are deeply invested in 'doing it right'. The Shelter, NCI, Luskwood - all strong 'pro' cases for BanLink.

But these are not the general cases - most of the grid is populated by casually interested parties, not so thorough, simply seeking fast remedy.

Whatever the outcome, we shall all see what happens.

Anyone use the Real Reputation system by Shaun Altman, url http://www.slrealreps.com, or the RatePoint one?

There's a video of instructions for RatePoint at http://tools.ratepoint.com/sldemo.

I just tried to click on the SL screen shown in the instruction video to change where I was looking.

the girl in the video got -5 from me because she didnt' clean up her box...

Wouldn't surprise me... though we'll never know since I'm sure part of employment for LL involves an NDR (non disclosure agreement) as you said.

Sheesh.... Even after seeing your post I kept trying to turn the volume down by using the Music slider in the video. I feel stoopid now

I have put a note on the front page of my profile (anyone who wants is free to copy/paste it) saying in no short terms that I have no part on these reputation sites and essentially that anything they put there has no credibility in relationship to me. I'd urge everyone to do the same.

I have no plans on using them. I keep wondering where the law stands on these things. We have in the UK things like the Freedom of Information Act and Data Protection act, which can be used to check what information is held on you and to enforce the correction of it if needed.

"Real" names may not be involved, but then Reg Dwight and Katie Price still likely have a vested interest in potential libel against "Elton John" and "Jordan". I imagine some of the leading lights in SL have avatar names that are quite valuable to them, as brands.

It will be interesting to see which, if any, of these opinion harvesting systems falls foul of the first lawsuit.

My understanding is that those laws allow correcting incorrect information. What I store about a given avatar is name and key, plus a table of key1/key2/rating type/score.

The data is automatically correct as it's obtained from the grid. In the cases where the system doesn't have complete confidence in the data, it's marked as dubious until retrieved from a trusted source.

This for instance applies to avatar names. If an avatar isn't known to the system yet, and the source of the name can't be assured to be 100% reliable, it's marked, and the name appears with a "?)" following it in the output. This mark disappears when information is submitted from a trusted source.

At the time, this data is nearly inexistent, and mostly consists of remains of incorrect data produced by a bug present in a very early version. Your're very unlikely to ever notice this, as the moment somebody with a scanner comes near one of the people whose data is incorrect, their information will be corrected, as the scanner is considered to be a trusted source.

Soon, the remains of it will be completely eliminated by using a script that will verify all the remaining ones.

Regarding the other data stored, such as the ratings you gave, you can see it all in the TrustNet website. This only applies if you use it of course.


I would doubt very much that any of them.

On the data protection side, they should be in the clear due to gathering public data. Your name and key are public in SL, and that's all really all that's needed. No attempt is made to link this to RL information, nor is it needed. That data about is correct as it's obtained from the grid. The ratings stored are an opinion, and are exactly what was submitted. The submitter can review, edit, and delete them at any time.

On the libel side, that would be the submitter's problem if they submit something that would qualify as libel, but I think that opinions are generally exempt.

Hey Ill turn this around -

What are the reasons we need any of these systems?

The only one I see is banlinks stated purpose - to combat greifing.

So what is the purpose of the ratings systems?

The DPA also indicates that you must provide all personal data held about someone to that person on request.

http://www.opsi.gov.uk/ACTS/acts1998/19980029.htm

I suspect it is arguable whether data held on an avatar would constitute personal information. If the RL identity of that avatar is well known, it probably would - otherwise, I don't actually know, and it might be an interesting test case.

You're allowed to charge a (capped) processing fee for requests anyway though.

I run TrustNet.

In my case, the main aim is providing additional information. The scanner tells you the overall opinion of the people you chose to trust of a person. This isn't intended to replace the user's brain, rather to augment it.

My own usage of it is that as somebody who can kick people out of Luskwood if needed (I'm a low level moderator and can't ban people), I like having some extra information about people. If somebody with a bad score shows up I hear it (it plays a sound depending on rating). This doesn't mean however that I do anything to them. I simply check out who is it, and who rated them, and based on that take a decision that ranges from not doing anything to paying more attention. If negatively rated people start showing up all around, I start paying a lot more attention. The same goes the other way, if I hear there's a plan to attack Lusk soon and people start showing around, I'll pay a lot less attention to people with good ratings from people I know well. Used in this sense, it's simply an aid when dealing with 20 avatars, and doesn't enforce anything. If people attack, I submit negative ratings for them, so that I know if they appear nearby again, and other people are warned.

The scanner has also the start of a search function. The idea of it is that you can ask the server to find you a well regarded scripter, builder, etc, inside your group of trusted people. So if you search for a scripter, it returns the list of the people with the highest ratings in the scripting category. Searching for people negatively rated isn't possible.

I've got a security orb in progress (will be released soon). This is rather more banlink-like, but still not the same thing. It acts based on score thresholds. You can tell it to only kick people you personally rated (making it work like a normal orb, only with an unlimited length list that can be changed from anywhere), or set a threshold that requires multiple people to submit negative ratings before they're considered unwelcome. This can be used to organize democratic moderation, by trusting a few moderators, and requiring several of them to agree before somebody can be banned. It can also be set to only allow access to people you rated positively, or have a high enough score. If somebody is kicked out by it, they will get an IM explaining why they were kicked out, and why their score is the way it is (they get a list of people who rated them).

I've also got an API that allows making direct queries to the server from the owner's point of view. I use this on my dome's door in Clara. There it's mostly a gimmick, but works well for testing it. This is intended to be used by people who run some sort of custom security system and want to integrate TrustNet into it.

The aim of the system itself is accurately representing people's reputation within a group, providing useful information, while being very resistant to attempts to disrupt it. Unlike simper systems, TrustNet doesn't have global ratings, and it's perfectly possible to have a good score in a place, and a bad one in another. I think this is good because it's a lot more precise, and unlike the old system, doesn't discourage exploration. Global systems may encourage sitting all day in a place where you know you're welcome, to avoid being rated negatively by somebody elsewhere. If some griefer rates you negatively it's not a big deal, as they probably won't be trusted by anybody of importance anyway. If they are, their ratings will be seen by those who rely on them, so they can decide to remove their trust in the griefer.

I don't intend such things to be taken as gospel, however. You can always ask the server where it got that score from and get a full list of people who contributed to it. It's the user's responsibility to decide who to trust, and what to do about the results.

It's not supposed to be a list of evil people either. The vast majority of ratings are positive, and I intend to put more emphasis on the search feature in the future.

So the purpose of this essay is your system lets you decide if someone has a bad Reputation in a certain locality? I suppose with the added benefit of a personality contest to see who is "trusted"

How does this differ from word of mouth in a locality?


Im not complaining about any of these ratings systems. I just dont see the use. I, like everyone else SL rezed after 2004 have not really ever used rattings except as a pat on the back.

In 2 years I havent needed them, I just dont see whats changed now.

Actually the link goes to the exact comment, though evidently your browser is not getting you there correctly.