| This 59 message thread spans 2 pages: < < 59 ( 1  ) || |
|Should the ODP/DMOZ Use NoFollow Tags|
What that change anything?
System: The following 5 messages were cut out of thread at: http://www.webmasterworld.com/directories/3460327.htm [webmasterworld.com] by webwork - 8:35 am on Sep. 29, 2007 (utc -5)
I think a nofollow tag would significatntly decrease the number of submissions to the site. The only reason that anyone I know submits there is to get g Google Directory listing.
Maybe the submissions I made over the last couple of months will come out of "limbo" now.
|Most informational content can be quite "old" and still useful. (Its use changes over time, of course.) And a corporate website may not change in structure, while its actual content is an up-to-date database (for instance, of items for sale, or scheduled performances). |
I'm thinking more along the lines of websites which say "So long and thanks for all the fish" in many different ways, rather than websites that are still up and active. Or websites for clubs which no longer meet, for companies which went bust, basically spaces on the web that have simply been abandoned. 404 checking doesn't pick up on this if there's an actual page there, and I've seen it a lot in some of the more obscure and local categories.
But back to the OP: would adding nofollow actually change anything? If it were implemented on all links, it might discourage a few SEOs from spamming Dmoz so heavily. Dmoz can operate whether or not it passes pagerank, because its usefulness to searchers does not depend on its effect on any search engine.
However, adding nofollow would discourage a lot of submissions, good and bad. After all, Dmoz sends very little direct traffic; its effect on traffic levels is almost entirely by proxy, through improved search engine indexing. If it nofollowed links, who would bother to submit?
|So long and thanks for all the fish |
Indeed, our robots aren't going to find those; we have to rely on human beings to notice them.
If you find such a link, you can help us to fix the problem by clicking on the update link at the top of the page and reporting it. Such update requests are prominent on editorial dashboards and most will give them priority.
|Indeed, our robots aren't going to find those; we have to rely on human beings to notice them. |
If you find such a link, you can help us to fix the problem by clicking on the update link at the top of the page and reporting it. Such update requests are prominent on editorial dashboards and most will give them priority.
I suspect I'm in the minority, but once in a while I do go through categories that interest me, and report all the dead links. That's why I noticed this issue. There are a lot of Tripod, Angelfire, Geocities, AOL and other homepages that turn up specific "not found" messages rather than 404s. It doesn't seem as though the 404 spider is checking for these specific messages, and it might be a good idea to spider for this type of text and flag such pages automatically.
On a further note.. there's alot of other people and websites who pull data from dmoz, whole categories, subtopics, you know - and why would dmoz as the original source add a nofollow which complicates things on so many levels for both themselves and for everybody else using their data? It's open - not closed - the traffic from dmoz might seem minimal or non existant .. but for alot of websites it's a reality from the search engines too. Why would Dmoz want to do as Google want people to do - just because? - google is one of the websites who are pulling data from them, because they know that they would not have been able to sort the internet without those reviewed and trusted sources to begin with?
Well, the point is, other people DON'T follow the ODP "nofollow" decision (whatever it is.) Google can put "nofollow" on their directory links, or not, regardless of what DMOZ does. Ditto for everyone else.
Changes like this would naturally be discussed in the internal forums, and of course this particular change has never gotten any support from the community, because there's no reason to suppose it would accomplish anything we're interested in doing.
The issue isn't whether dmoz links out to trusted sites or not...by the way they link to sites that explain how to make fake id's, sites that promote drug use, sites that promote drinking human blood, that's just off the top of my head, I'm sure anyone could find even better "trusted resources" if they tried.
But that's not the point, the issue is the growing number of editors who abuse the fact that dmoz supposedly passes an extra boost to it's links in the search results. And they're not listing tons of porn and spam, or making 1000's of dollars by selling links, they're just very quietly listing only their own sites while keeping out their competitors or listing them badly...nofollowing the links would make it more pointless for webmasters to become editors and favor their own sites in the categories that they control.
the issue is the growing number of editors who abuse the fact that dmoz supposedly passes an extra boost to it's links in the search results.
... they're just very quietly listing only their own sites while keeping out their competitors or listing them badly.
(1) Do you have a list of those editors? If not, how do you know what their number is, let alone whether it's increasing or decreasing?
(2) Assuming such editors exist, it's irrational to suppose they could suddenly develop scruples about using GOOGLE directory to, um, "pass an extra bust to ITS links.
(3) One editor _can't_ keep a website out of the directory. But he can get himself kept out of the directory by trying.
(4) As for "listing competitors badly" -- that's a form of abuse which would be visible to anyone who looked at the listings in the directory. Have you ever seen such a thing happen? Did you report it via report_abuse or quality_feedback? And what happened to it?
If we don't see things like this happen, and you don't report the actual events, then we shouldn't be expected to believe that you know they ARE happening. And if we can't believe something, then it would be irrational to act on it.
So if you have evidence of a real problem, then report it. And the people who care about the directory can take whatever steps are appropriate, for the specific problem and for general improvement of the directory.
But at this point, apparently nobody knows about any problem, whether or not "nofollow" would address it. And nobody knows about any effect of ANY kind that "nofollow" on dmoz.org (but not AOL or Google directories) would have.
How are you going to list a "competitor" badly in dmoz? Seriously. Do you know how the internal systems work of dmoz with the editors and the systems behind the public view? I haven't even figured out how to list a competitor badly yet? - please explain with an example, I would really like to know on an as-is basis, rather than abstract allegations - furthermore - if you have any evidence of any of the claims you make against any and all editors .. please, use the ABUSE button. It's not there to look nice - it's there for a reason.
NoFollow wouldn't do anything - the links and everything is still there, reviewed, accepted, rejected, placed on hold, sent to other categories, all those things which happens behind the scene - nothing would change - just annoy the people who pulled data from dmoz and used it as a quality resource and sent robots out which were programmed to like ... respect the nofollow tag.. (that's google we're talking about) and they would have to change the robot script to "follow it anyways and calculate anyways" and a link in dmoz would perhaps not matter as much anymore.
[edited by: Webwork at 9:13 pm (utc) on Oct. 4, 2007]
[edit reason] Please tone down the posts, including avoiding religious references. Thank you. [/edit]
It's easy to list a competitor badly, just move them to a really remote low pr category and write a bad title and description, or just leave them in unreviewed forever...
Anyone will tell you, editors included, that dmoz does not exist in order to pass pr and/or give sites a boost in Google.
Nofollowing it only reinforces that statement.
[edited by: Webwork at 9:14 pm (utc) on Oct. 4, 2007]
[edit reason] Please don't "take the bait" and perpetuate problematic posts. Thank you. [/edit]
If you see spam in DMOZ, report it.
(hope the URL is OK to post here, it seems relevant and authoritative enough to me, if not, mods are free to zap it)
Much more effective than trying to put nofollow tags all over the ODP, which I consider ridiculous.
>It's easy to list a competitor badly, just move them to a really remote low pr category and write a bad title and description,
And if this really happened, it would be easy to _see_ it.
Where have you seen it?
If it's really a problem, it'll be a visible problem.
>... or just leave them in unreviewed forever...
You can leave a site in unreviewed forever even if it isn't a competitor.
In fact, _every_ editor leaves nearly _every_ site in unreviewed forever. And it's not just easy to do that, it's impossible not to. No matter how much you edit, your total is an insignificant part--a fraction of a percent--of the whole. The OTHER 99+% of the listed sites, are sites that you personally left unreviewed.
In fact, when you think about it, you don't even have to be an editor to leave a site in unreviewed. People who've never seen a computer can do it without assistance of any kind. And everyone does it. Non-editors do it for ALL sites. Editors just do it for MOST sites -- and an editor who has reviewed a million sites, still hasn't visited a tenth of a percent of the web.
So this isn't a problem. It's a fact of life for everyone, and there's no reason to single out competitors.
The ODP is quite different from wikipedia.
Wikipedia is actually intended to be the repository of the knowledge, not a collection of reviewed links.
The ODP goal is to build a collection of reviewed sites, sorted by topic. Kind of a human powered quality score.
The collection and reviewing of sites it the core of the ODP.
In wikipedia there is a problem with people spamming links directly into the site. "nofollow" is for wikipedia far from a good solution: if GOOG doesn't count links from wikipedia it means that those of us who get referenced by wikipedia for borrowing the ideas and the research we performed don't get any reward for our work anymore.
The ODP's review system is far from a single editor doing things. There are rather strict rules in dealing with one's affiliated sites, there are reviews of categories, there spiders and editors are supposed to review categories every so often. And for those thinking there's little oversight of new editors: just try becoming an editor and see how long it takes before you'll be spotted.
I believe to level the playing field one of two courses should be followed.
One, totally revamp the system so that all submitted sites get a review in a reasonable time frame, say 3 months. Submitted sites should not be allowed to simply hang in limbo forever. If a site is reviewed and rejected then the person that submitted should get a detailed explanation why the site was rejected. Those reasons should have some type of standardization on them so that everyone has a fair chance to getting in.
Two, if they won't change the model then no follow tags are probably a more fair way to treat the sites. At least in this situation a poor competitor's site doesn't get the benefit simply because of luck or some other factor that you can't control.
[edited by: Webwork at 4:34 pm (utc) on Oct. 5, 2007]
[edit reason] Good points. Just tidying up a bit. [/edit]
We don't think of it as a 'playing field' to be 'leveled'. We think of it as a 'structure' to be 'built.'
Which suggests an entirely different set of goals, and an entirely different kind of activities to achieve them.
|I believe to level the playing field one of two courses should be followed. |
The ony problem with your suggestions are that they are only relevant if DMOZ was there to provide a listing service for webmasters and an editors job is to process submitted sites. Editors are under no obligation what-so-ever to even use the pool of suggested sites when looking for sites to add to a category. They can get sites from wherever they like. Most do at least scan the pool of suggested sites.
Forgive my rather noobish understanding...
But I've always been under the impression that having a link from ODP was quite significant, particularly since the search engines incorporate ODP's data into their index. And from what I understand, people started exploiting this, much like Wikipedia was exploited for link juice, not to mention the stories I've heard about editor's editing their own niche in order to keep competition out, or "selling" listings or holding listings for "randsom"
I've been waiting for a link from ODP for 2 years. I did not know about pagerank or nofollow or trusted at the time. I just wanted my site being exposed.
Now as I've grown more jaded through out web developement, I came to the understanding that most people don't look at ODP as a source of reference like others have pointed out, but as a way to increase their positions in the SERPs, and that game is preventing sites with exceptional content to get ignored or passed over.
I think they should use the nofollow tag, just like Wikipedia, so people will stop gaming the site. Let it remain human edited, thats fine, but ya gotta keep in mind that some (not most or all) of the humans have alterior motives for editing at ODP.
And now that I have some more coffee in me, and I'm thinking clearly, I would like to clarify my position...
I think nofollow should be used on ALL directories or sites that are generated out of user-contribution such as ODP or Wikipedia.
Having a site referenced by a *.gov or *.edu, to me, seems to be alot more authoritative than being referenced by a directory.
"Editors are under no obligation what-so-ever to even use the pool of suggested sites when looking for sites to add to a category."
That's true. That's why this whole thing is not an issue of spam, or whether the links are trusted, it's an issue of control.
I'm an anonymous editor, no one knows which sites are mine and which aren't or even how many different editor accounts I might have.
Nofollow may not be ideal but at least it's something.
[edited by: Webwork at 2:38 am (utc) on Oct. 6, 2007]
[edit reason] No fuel needed to advance the dialogue at this point. [/edit]
>I think they should use the nofollow tag, just like Wikipedia, so people will stop gaming the site.
If you'll read previous discussions of this issue, you'll see it repeatedly stated that "The Google directory doesn't use 'nofollow', so the games would continue whatever the ODP does."
This seems to me so obvious that I can't figure out why it is so often overlooked.
But there's a more subtle psychological point that I would make. And it's this. Playas aren't smart. They're sly but dumb. And their strategies are based on ignorance and chance, not on logic.
A playa thinks like this: "Now, a link from dmoz might help. It can't hurt. Some people say it helps a lot, some people think it helps a little. Lots of people talk about it, that must mean it's important. I don't know, and I'm too dumb to figure out. BUT THAT DOESN'T MATTER. I have a winning strategy. It's: try to sneak a dmoz link by fraud and deception -- in other words, share my ignorance as much as possible, and try to leave as much to chance as possible. That will keep other people from using intelligence and knowledge to get an advantage over me."
(Now, from a short-sighted self-centered perspective, ignoring what other people think about him, that looks like a good strategy. And we aren't talking about honest people here. So this strategy works--so long as he can outrun his reputation.
You see, it doesn't matter if he doesn't try to learn the rules BECAUSE HE WASN'T PLANNING ON FOLLOWING THEM ANYWAY. He was planning to cheat. And therefore, for all all practical purpose, there's no point in doing things like "nofollow", because that's just changing rules to give the cheaters something else to ignore.
Again, nofollow isn't ideal. It isn't "something". It's nothing, absolutely nothing. It doesn't address the fact that the ODP has influence beyond its website, and it doesn't address the fact that cheaters don't stop cheating just because someone makes an arbitrary change in the rules.
|If you'll read previous discussions of this issue, you'll see it repeatedly stated that "The Google directory doesn't use 'nofollow', so the games would continue whatever the ODP does." |
I'm not referring to Google's Directory, I'm referring to ODP.
However outdated and antiquated the toolbar is, looking at one of the categories in ODP where my link would appear, gives it a page rank of 7.
That would be a very nice link for my site to have.
That would be a nice link for ALOT of sites to have.
And to be honest, in all my years of surfing the Internet, I have not once seen www.dmoz.org appear in the top 10 for a term I was searching for, like BSOD or mickey mouse, or glass of water. So I seriously wonder how one can hold the position that in today's world of the Web, how DMOZ can be used as a directory reference. Maybe back in the day when everyone was exchanging ip address's in BBS's to update their HOSTS files, could I see a use for ODP, but today, its a non-paid link directory that has become obsolete.
Take the weight it carries away, and who will use it?
I sure wish there was a study done on how much contributions to Wikipedia raised or dropped after placing nofollow on the links.
>I'm not referring to Google's Directory, I'm referring to ODP.
The ODP (more precisely, the OD) is published online at many websites. One of the most important is Google. Google's directory _is_ the OD. There is no other Google directory.
Every website that publishes the directory can potentially pass page rank. But Google is one of the more important page rank passers. You simply can't ignore Google when you're talking about page rank effects of the OD!
And so long as Google is passing page rank, on your psychological theory of abusers, then "nofollow" on some OTHER site will have no effect. (On my theory, even if Google were NOT passing page rank, "nofollow" would have no effect.)
>And to be honest, in all my years of surfing the Internet, I have not once seen www.dmoz.org appear in the top 10 for a term I was searching for, like BSOD or mickey mouse, or glass of water.
I have. But that's OK.
>So I seriously wonder how one can hold the position that in today's world of the Web, how DMOZ can be used as a directory reference.
No doubt the dumbing-down of the universe is accelerating, but there are still a few people who are clever enough to figure out how to use sites even without their appearing in your searches.
>Maybe back in the day when everyone was exchanging ip address's in BBS's to update their HOSTS files, could I see a use for ODP, but today, its a non-paid link directory that has become obsolete.
Nothing good ever really dies. Admittedly, my last programming project for the DDP-316 machines (the original IMP, if that means anything to you) was in assembly language on punched paper tape. And I kept backups on cuneiform tablets. The algorithm to convert dates from Sumerian eponyms to Greek Olympiads never quite worked. But I hear parts of the source code are still floating around in .NET DLLs.
Folks, this thread feels like it is winding down, which also - and often - is the point where ODP/DMOZ threads catch a second foul wind.
It's been an interesting discussion up to this point. If there's no new insights relating to the use of NoFollow tags let's call it a night, instead of revisiting other well worn issues.
Just one more thing, dmoz has requirements for using their data, nofollowing the links could be added as a requirement. That would mean the Google directory and all the others would have to nofollow their links as well...
I think that as long as it has 10000 links to one site, and that data is exported out to other various directories pushing the IBL list to 30,000 high quality links, nofollow should be placed on those links.
I'd love to have 30,000 PR7-PR9 links pointing to my site.
lostjake, I think on more careful review you'll find that the 1000+-link sites are not personal sites or even business sites, but are major reference sites with hundreds or thousands of individual contributors.
You might also want to check the PR of the pages containing those links. They are deep enough in the ODP structure not to be "PR 7-9".
Your attitude is common enough, that it's worth emphasizing the facts that should make a rational person feel better eventually. (Hey, we're all humans, we know about not being able to control visceral reactions.)
As I've posted before in response to the OP, I think they should use Nofollow, if not to at least counteract an abused and manipulated system.
[edited by: engine at 5:20 pm (utc) on Oct. 8, 2007]
[edit reason] See TOS [webmasterworld.com] [/edit]
I am not sure why DMOZ is being asked to change its behavior to fix a problem that seems to exist with Google.
Let's get Google to fix the prob. How about Google only honoring links that have an explicit follow attribute in the rel tag?
Sure, it would mean some work for webmasters to add follow atts to links they want to pass PR (etc).
But, individually, that is surely worthwhile to correct the current situation.
I think a typical editor attitude is that ODP links tend to counteract the abuses of SERP activity: and therefore when an SERP perp calls for "nofollow" our reaction is "aha, we're successfully doing our job."
This is also an emotional reaction, because we're also human. (It's conceivable that we could be failing at our primary mission EVEN THOUGH we were frustrating webmasters.)
I've never considered "being human" a personality attack before. A personality attack would be suggesting you should have taken it as a compliment, which of course I wouldn't do.
There is a point at which ODP threads show signs of issue exhaustion. It's a judgment call when that point has been reached and, at that point, it is the policy of the Directory Forum to lock threads so that they retain their value and their focus isn't lost or diminished.
It appears that most of the major points have been aired. Therefore, this thread is now being closed.
If any latecomer has a particularly insightful post to that adds value to the consideration of the issue you may start a new thread.
Thank you all for your contributions.
| This 59 message thread spans 2 pages: < < 59 ( 1  ) |