Is Paying for Links in Profile Picks Cheating?

I am not moving the goalposts.

This is a conversation its not about "winning"

--------------------------

You said that in the SL search there were too many subjective factors for it to be compared to a race. I don't agree, but thats because I think you are getting bogged down by detail. When the question is really a basic one.

But .. since you think it wouldn't apply .. I offered Figure Skating which includes large numbers of subjective factors much like the SL search .. and they still don't allow steroids.

---------------------------------------------


The basic question by the way is

Should places have the Search ranking they Earned?

or

Should places have the Search Ranking they Earned+Cheated to get?

They also don't allow you to pay the judges for a favorable rating

Aye but Linden Labs does allow paid picks.

However I would like to point out that sports companies do pay competitors to wear and endorse their products, which is perfectly legal and a much closer analogy to the concept of paid picks than cheating is.

They do have the search ranking they earned, they earned it by using the currently available tools, if they get caught cheating they get kicked out.

This rightful place malarkey is the kind of rubbish I hear football fans talking on radio phone ins on Saturdays.

"We're a much bigger club than them, how dare they beat us, we've got better players, we deserve to be higher up the table".

"but they have more points than you?"

"Yes but they use boring long ball tactics, we play elegant pretty football"

Ciaran ..

If you wont consider the possibility that paid for picks is cheating .. you will never see my argument.

You do not have to agree, but until you actually consider the possibility you will keep coming back with posts like this.

You have to think a bit more abstractly.

You are still get caught up on the "If Linden Labs doesn't do anything, then its all fine" thing.

That's generally how rules and laws work Colette. There are national speed limit signs on my route to work, they used to mean end of speed limit but one day someone decided that letting people hare around at silly speeds wasn't sensible, so they changed the rules and the signs now mean something else. The people who previously drove on these roads at speeds that are in excess of the current limit weren't speeding back when the sign meant something else.

There's a difference between something like paid picks and something like, say, finding you can manipulate auctions.

Nothing about this amuses me, Kitty. The question of this thread is, "Is paying for links in profile picks cheating?" It's a very simple question and I'm trying to keep it simple, without shooting off at tangents. In order for something to be cheating, somone must be cheated, and I asked 2 questions about it. If you're not interested in straight, simple answers, there's no point in replying to me at all, so we'll leave it that, shall we?

A bunch of numbers go in, a single number comes out and everything sorts by that after filtering. You may not know exactly how all the numbers are weighted to yield a single number but you know that if nudge the numbers enough the end result will be more to your liking.

So break it down and look at some of the techniques that I personally find rather questionable:

* number of search entries

Most stores will list once (different outlets might list several times), someone here lists his same store 5 times by dividing it into 5 different parcels.

The benefit is quite obvious (three of them rank high for the primary keywords): the more listings that do well and are spread out, the more someone is likely to notice you.

The downside should quite obvious as well: if everyone decides to list themselves 5 times you're not going to find anything anymore and why would they stop at 5? 20 times is even better, 100 times is even better than that.

* keyword repetitions

First of all make the distinction between a keyword that simply occurs naturally and one that is out of place.

There's nothing wrong with "Tabby Cat (Large)", "Tabby Cat (Medium)", "Tabby Cat (Small)", "Black Cat (Large)", etc if that happens to be what it is. On the other hand appending "neko, feline, animal, kitty" to every object is clearly primary keyword padding/spamming to me. The sole reason to add those keywords is to influence relevance for that keyword and it happens 50-70 times so it's really not negliable.

If repeating a keyword more often has an influence on making a parcel more relevant for that keyword then it's simply a matter of who can keyword spam with the most prims.

You can shorten it to just a keyword and a number everyone is free to choose, the higher the number the more relevant you'll become for that keyword.

* paid for picks

"5000L$ each week by Jaxvan" (http://world.secondlife.com/group/f0c4f441-a7ae-7329-92c9-746163807b03) is an picks incentive group with a weekly lottery: 2076 members

Save everyone the trouble of having to buy picks and just make it a number everyone is free to choose as well.

* interlinking parcels

Part of what determines any page's weight is the weight of the pages linking to it (yes, that's a simplified way of putting it)

If page A links to page B and A has a high ranking factor then B will rank higher as well (popularity rubs off).

The more interlinks you create, the more everyone benefits so once again save everyone the trouble and let them pick a number that they want to have.

* "Googlebombing"

"Miserable failure" currently returns 0 results in places so it would just take 1 person to create a pick called "Miserable failure" made on a certain Linden's land to emulate one of the more known Googlebombs. Relevance is nill here, someone said "A is linked to B" and the search engine makes the connection without question.

If enough people all say "A is linked to B" together (by having a picks named that way, or by injecting links onto the parcel page) then you B climbs the search ranking for A.

(The only occurances I've come across so far have been accidental but it clearly works on the SL search and nothing keeps someone from "optimizing" this way even though it should be obvious why it's not exactly a legitimate technique)

---

I probably missed a few but what exactly is it about search that isn't quantifiable? You don't know if repeating your keyword 5 million times is better than having 5 million people add you in their picks but that's ok: you can do both at the same time!

You can argue that it all works out well enough for the web but for one you're talking about human readable continuous text that's being indexed there. If 50% of all text consists of 3 keywords then you can discard it but in SL that could actually be a perfectly valid page because search pages are really anything but human readable continuous text. Web pages don't compare to SL search pages.

I really fail to see anything about the new search that is not based on playing with numbers or where merit or "relevancy" come into play. Whoever plays dirtiest comes out on top and for someone who actually has the use search that's hardly a benefit and it's no different than what came before.

(And no, I don't think the old search was all that great either but in my opinion they're really both equally awful as far as ranking is concerned. The new search does have better keyword matching but that's not all that important when the new ranking is as big mess as the old one, in some cases it even makes the problem worse because you're dealing with more results)

I think the only answers that are really wanted are:

a) Yes, it is cheating, so we (replier) will stop doing it.
or
b) No, it is not cheating, and here is the titanium argument that proves this, thus enabling you (poster) to do it too without moral qualms.

I know that was what I wanted when I made my earlier post about booby traps.

It's nothing like it.

Sports companies pay big names to wear their gear. The company is buying reflected glory. They want people to associate the success of the athlete with the company logo. People can see it and know it for what it is.

For a parallel with paid picks.
An ordinary person of no special fame or sporting prowess walks into a store. There's a sign up. "If you take this sports gear and wear it, we'll pay you."
Plus (if there is a Picks effect on search ranking) then for a complete parallel, the John Doe has to tell people how great the sportwear is.

Paying for picks is Pick Stuffing the search engine. It's the same family of backdoor abuse as is keyword stuffing - which practice is generally acknowledged in the industry as being unethical.

http://forums.secondlife.com/showthread.php?p=2137327#post2137327Want to try again? There's plenty more where that one came from.

Some people think it is, other think it's not so: "it depends".

Is that simple enough for you?

Simple answer.

If someone cheats to get higher up in search ranking, the parties cheated are
1. the people who would have been ranked higher than the cheater had the cheater not cheated.
2. The people who tired of TPing to places and missed out on alternatives because the cheating parcel took up their time and patience.

The cheater is jumping the queue.

Someone pays someone for endorsing their product, come on, how can you say it's nothing like it. It's very much like it.



However another analogy and one that you might be able to associate with Joe Public is referrals, refer product x and we'll pay you y. That's a very common system. One Linden Lab employ even, funny how yet again a Linden Lab tactic is ignored, but I guess that's somehow different? The words "Absolute" and "Balderdash" spring to mind, and I'm not referring to the avatar of that name.



No it's in the family of referrals, which is a practice that is not generally acknowledged in the industry as being unethical.

Someone here divides his store into departments - yes it does have real departments - and lists those departments seperately, targetting the products in each department (fireplaces, tables, art, etc.). And if it were a simple matter to divide off other departments he would do it. Do you find fault with that? What if each department was in a different sim? Would that be acceptable? If it would, what is wrong with the departments being adjacent to each other?

I already said I'd be perfectly happy with a normal "yes, I see your point, but I disagree" post from Phil .

If you'd read the questions, you wouldn't have written that post.

That isn't even a poor attempt to answer the questions, Kitty.

And quoting things that were said at other times *about* other times, doesn't apply to the current time, or to *any* time other than when they were written. That was just a wasted effort on your part.

--------------------------------------------------------------

Oh well. I'm done with this thread. I tried to get it back on track, and just discuss the original question logically, but some people want to take it off every which way.

It looks to the engine like a referral. It's meant to look that way. Technically from a system design point of view is is a referral.
The reality is different.
The reality is that it is an artificially generated referral under the control of the entity being referred to.

Such a referral is in reality no more significant TO THE SEARCHER than a website page being referred to by another page in the same website.
Read that again.
If you engineer separate websites purely to create referrals to your own website, those referrals should have no more weight than internal website references, but the search engine can't detect the abuse.

It's definitely cheating. It's definitely dishonest. It's definitely unethical.

However such a practice can be circumvented to a degree if traffic is still a factor because the traffic on 5 parcels won't reach the heights that one parcel would.

There are however other factors to consider with multiple parcels for the world as a whole, why should someone who owns half a sim be restricted to the same number of search listings as someone who owns a 512? Landing points and such like mean the half a sim is harder to navigate so I don't see that multiple parcels for the same store even, should be restricted.



I'm not sure how Linden Lab have configured it but keyword spamming shouldn't work that way, indeed if configured sensibly it will have negative effects.



I haven't got a Scooby Doo what you're talking about here.



There are too many variables that's why. Not everyone has the same tastes, what one person sees as quality another person sees as a pile of pants. The issue of "rightful place" has to be easily quantifiable or it's bunkum to make such claims. The search here makes "Rightful place" bunkum on many levels.



I'm not a fan of the new search, they have a system designed for places in a world of commerce. They were supposedly doing something about that but no comment has been forthcoming.

If it were an army of alts I'd agree with you, but it's not. I've been through this before but picks are user made decisions.



Paid picks are not cheating, they're out in the open. No referal is important to the searcher, the searcher doesn't have the option to rank by picks. They aren't the same website, I can't emphasise this enough, paid picks operate on the premise of different individuals choosing to participate. They are very much a referral based system.

Again, an army of pick alts would be what you're suggesting.

So where is the flag on a profile pick that shows it as being a paid endorsement?

Come now, in magazines the ads made to look like legit articles have to have the rider text that can be read stating they are paid adverts.

It's hard enough to find anything as it is without needing duplicate results cluttering things even more.

Searching for the general keyword three of the five "department" parcels shows up in the first few pages; if everyone did that then results would become utterly meaningless.

And yes, I do think I get your point: just like a site has an entry page and then branches out into pages which will address specific topics within the scope of the site it's better if someone lands on the page that's relevant rather than the entry page and then navigate the site from there.

SL doesn't work like that though, it's top-level (entry page) only and sub-navigation needs to be done by the parcel owner through things like tp prims. I'd actually love it if search would just override the landing point so you can tp straight over to a specific item listing rather than be forced to find your way to it.

If they're outlet stores on multiple sims then it's equally cluttering things up (tends to really annoy me when I'm looking for a specific name and a mall-like sim with an outlet ranks higher on the store name than the store itself) unless each store sells entirely different items so only one of them will be in the results for any given keyword.

Okies, imagine I sell cats and I parcel my store up per colour since it's an openspace there's plenty of room.

You want to buy a kitty to sit on your couch and you like sculpty animals better so you search for "sculpted cat".

My store appears somewhere on the first page so you work your way down, take a look and decide that my cats aren't really what you're looking for.

You move on to page 2 and there I am again, but the title is different so you click and "hey, I've been to this one, meh".

You land on page 4 and there I am yet again with a different title.

---

If everyone did that you might as well give up trying to search altogether, it would be an immense mess. There's a certain expectation that if you've already seen one store it's not suddenly appear again and again as you jump from page to page.

I'm not sure if anyone ever did find out whether GSA will penalize anything to begin with.

If it's targetted solely at indexing documents or intranets it doesn't need any of that. Most people don't keyword spam their own documents, if a word appears 10 times in one document and 100 times in another you can make an easy claim the latter is likely more relevant.

If Google created GSA with the expectation that it will be fed trusted input then it'll perform horribly when that assertion isn't true and in the general case you can't really claim that all business owners are equally honest when it comes to their search ranking so.

Those 3 appear on pages 1, 4 and 8, and each is very clearly stated in its title what it is, so nobody who does that search is going to be misled by them. If they were in different sims, it would be the same. If you find fault with it, your beef is with the LL search team, and not with me. Or perhaps you'd like to criticise places like B&B for having 2 seperate stores. That would just as silly too.

You're inventing things. I never said anything like that, or even hinted at it.

--------------------------------------------------

Look, Kitty, I'm sorry that you are miffed that I don't want to be friendly with you any more, but I think the reason is good. You never used to get involved in these threads to this extent, but now you are very involved in replying to my posts, and I can only believe that it's personal. You don't even write straight replies to them. Instead you invent things like that website bit, and cloud things by writing a lot without even attempting to answer the questions you quote, and so on.

I'm not going to be friendly with you again, and you know why. I won't allow you to screw up threads just because you want to hit back. You can either join Colette and Pie, or you can go back to posting sensibly. It's your choice.