28 replies to this topic

[jdevalk]

Wikipedia has added nofollow on their outbound links just a few hours ago, after a surge of new spam due to another SEO contest...

[JohnMu]

SEO contests suck.

They make things worse for all of us, as if there weren't enough real sites with search engine issues on them to solve.... No, it's better to be #93 in some contest than to help some small-time webmaster get his family-run shop back into Google. Sigh.

John

PS Welcome to the forums, jdevalk! I hope you weren't waiting for the nofollow to delurk . What else drives you?

[jdevalk]

agreed, i hate these contests too... They cause all sorts of havoc.

Hehe and thx for the welcome I'm an SEO, read a lot of seo blogs all day, and a bit active on DP forums, but these seemed nicer

[yannis]

The no-follow should not really worry anyone with links in Wikipedia. The real value of these links is that they provide traffic to your pages.

Yannis

[JohnMu]

I tend to agree with Yannis, it's about traffic ... (but SEOs like good links anyway ) ... and most of the Wikipedia links that I've been following seem to be extremely lacking with regards to value (for Google) when compared to links from other sites. I don't know if Google is treating them differently or if they are on some "user generated content - spam warning" list (or most likely algorithm), but that's what it feels like to me.

That said, I did have one site that was doing very well in Google, mostly thanks to a few good links from the Wikipedia (as far as I could tell) -- someone cleaned up the categories and almost all value attached to the site seems to have dried up (the site is non-profit, doesn't sell anything, not even ads, but I feel kind of stupid putting the links back in myself ... even though they were removed as "spam").

The Wikipedia is a strange beast when it comes to SEO

John

[Nadir]

Wikipedia is part of all these large websites that are killing the web.

With so many links and no outbound links, they'll just be seen a giant hub in Google's eyes and rank first for any query, which causes smaller websites or other relevant resources to remain unknown by the average surfer.

[Wit]

I think Wikipedia is a special case, and has been for some time. I'm not convinced that a link from them was counting for as much as you would expect anyway -- link-juice-wise that is. For some reason, I tend to believe that Google likes WikiP pages them a lot for the content, not for the (outbound) links.
For that reason, I don't expect the nofollow to influence things gravely.... For the same reason, I haven't even TRIED to get links from them since a year ago. Not even for traffic (which is useless in the niches I serve).





----edit: added word that got lost

Edited by Wit, 21 January 2007 - 05:47 AM.

[bwelford]

I think this is an excellent move on Wikipedia's part. If you're motivated to edit an entry and you get some traffic by the excellence of what you wrote, well so be it. It certainly works for me.

In a sense, Wikipedia is correcting the fallacy in the whole Google PageRank approach. It's like Heisenberg's Uncertainty Principle. There are some things you can't measure. If you try to measure them then they're not the same. Once Google says inlinks will boost a web page's relevancy, then of course everyone, often supercharged with dumb computer programs, generates as many inlinks as they can.

If Google had been smart they would have kept PageRank as a corporate secret like the KFC or Coca Cola secret recipes. It's a great principle provided you don't tell anyone about it. So they should have gone public with the BrinBoost, naming it after the other founder. The BrinBoost would be a measure of the value of the outbound links or outlinks you put on a webpage. If everyone thought that the BrinBoost was the key factor they would be motivated to improve the quality of what they write.

You could even have a BrinBoost 'thermometer' with BBs from 0 to 9. They could then have devoted say 5% of their staff to supporting the BrinBoost approach.

They wouldn't abandon the PageRank thinking of course but it would be buried within those 100 factors they use for assessing relevance. This secret would be known to only a few high priests within the Googleplex. So they'd have the best of both worlds. They in fact would have improved relevancy since people would not be trying to create all those irrelevant inlinks. The PageRank approach would work even better. Who knows Yahoo! and MSN/Live might even have been unaware of the Google secret weapon.

Ah well, it's too late now and we've all got to suffer.

[EGOL]

I really like that thinking, Barry.

Imagine how much effort has been devoted to generating links.... it would probably pay off the US National Debt.

[yannis]

I think this is an excellent move on Wikipedia's part. If you're motivated to edit an entry and you get some traffic by the excellence of what you wrote, well so be it. It certainly works for me.

It also works for me. The Wikipedia is perhaps one of the few resources that belongs to everyone irrespective who you are or where you live. Wikipedia has grown to the point where a normal web user should begin his search queries from there, rather than the search engines. Try it! you will be amazed. At present I consider the quality of the information on Wikipedia better than any printed encyclopaedia.

Abuse takes various forms. Some try to promote their own websites (either for pagerank or traffic) and others are zealots promoting their own political views and then there are the graffiti artists. The first two are the easier to handle and it does not need the 'no-follow' ban to achieve it. If you visited a link from the Wikipedia and found an adsense site you will be tempted to go back and remove it. The second is more difficult to control - like most religious or political ideas. However there is always a counter-camp that keeps checking these pages. In most cases a balance is achieved and the ideas of both camps are well represented. The graffiti artists are more difficult to control as they will just deface a page and run.


Yannis

[joedolson]

I can't honestly say that I mind the fact that Wikipedia has gone nofollow - I just wonder whether it'll make a difference. After all, WordPress went nofollow but I still get thousands of comments and trackback spam. Fact is, as long as only Google claims to actually not follow "nofollow," these kinds of spam drivebys will continue to be considered valuable. Maybe not as good as they used to be, but people will still do it.

It's kind of like putting your thumb in the dike...but with a hole the size of a basketball.

[DianeV]

That's true. Same here, and I've always had to completely moderate comments on my blogs.

BUT using Akismet has meant that nearly *all* the spam comments (out of 2600+ in the last three weeks) have been quarantined.

There's also the Bad Behavior plugin, which apparently determines whether incoming comments fit a pre-determined set of spam comment parameters. If I need to, I'll use that one, too, but so far, it hasn't been needed.

Edited by DianeV, 21 January 2007 - 02:06 PM.

[bragadocchio]

If Google had been smart they would have kept PageRank as a corporate secret like the KFC or Coca Cola secret recipes. It's a great principle provided you don't tell anyone about it.

But they couldn't, since PageRank doesn't belong to them, but rather to Stanford University. It's also possible that the PageRank technology was part of what made them popular. When we think of some trade secrets, like the KFC and Coca Cola recipes, there are things about those that are known - afterall, they are recipes. The exact ingrediants of PageRank can only be guessed at, and there are elements that we don't know anything about.

For instance, how does the algorithm treat dangling nodes? That is, pages that don't have any links pointing out from them, such as a PDF file, or a txt page?

How do the ranking algorithms used presently by Google treat sites that don't have any links pointing outside of their own pages? Or that use some prophylactic measure like nofollow on all of their links? If a majority of sites started using nofollow on all of their external links, following the leadership of the Wikipedia, would it have an affect on how sites are ranked?

There are many other ways that a search engine could measure the value of links from a site like wikipedia than just a pagerank value. For instance, entries in wikipedia all have explicit timestamps associated with them - easy to find and measure (see: Temporal Analysis of the Wikigraph for an interesting analysis of the temporal aspects of the site).

What we don't know, though I've seen some discussion of it, is whether a newly discovered link has as much value as one that has been around for months, or for years? The historical data patent application from Google a couple of years back seemed to consider the age of links as an aspect that could affect scoring of pages - though whether or not that is done is part of the secret recipe. We just don't know.

Should Wikipedia use nofollow on all external links? The reason cited is that they want to reduce vandalism. Perhaps they should note that all links are nofollowed clearly on all pages that can be edited if they want to meet that intent. I suspect that even if they did that, the amounts of vandalism to wikipedia wouldn't be reduced in any meaningful way. It appears that the cost of vandalizing wikipedia is very small, and even the rewards of a few visitors to links may be considered valuable.

[projectphp]

Exellent! I don't like nofollow for multiple reasons, but this does seem like an appropriate usage. "We can't trust these links" is true, because they can be added or changed at any time, so it is a valid(ish) usage (not that I think it is a good idea... but that is another story )

[Mano70]

Just for your info, the Cre8asite forum is mentioned here:
en.wikipedia.org//wiki/Wikipedia:Administrators'_noticeboard

Off Topic

I wonder what kind of people that are wasting time on a lame SEO contest. From what I can see there can't be much to learn from this - non competitive keywords starting from scratch on brand new domains.

Edit: Not a live link anymore since the redirect messes with the url.

Edited by Mano70, 21 January 2007 - 07:38 PM.

[AbleReach]

Just for your info, the Cre8asite forum is mentioned here:
http://en.wikipedia...._spam_onslaught.

Not any more!

Wikipedia does not have a project page with this exact name.

We are listed under references here, for the SEMkit review: http://en.wikipedia.org/wiki/Dan_Thies

And under external links here, for the comprehensive robots.txt tutorial:
http://en.wikipedia....lusion_Standard

Edited by AbleReach, 21 January 2007 - 07:31 PM.

[Mano70]

Cre8asites redirect messed up the link, copy paste the link below and scroll down to For yoru entertainment:
en.wikipedia.org//wiki/Wikipedia:Administrators'_noticeboard

[Halfdeck]

In a post Matt Cutts published an hour ago, he says:

I don’t expect this change to affect Google’s rankings very much, but it’s good to see the Wikipedia folks paying close attention to link spam (and open to refining their trust for external links).

To me, his statement implies that, like many already suspected, those Wikipedia links were not passing much link juice to begin with.

[JohnMu]

This is going to be the year of the nofollow -- link blocks all around

Google added it to its own Google Groups (which are close to uncrawlable) and has even added a redirector script for profile links in the Google Groups.

When all links are nofollowed, we will be back where we started, :rofl:

Spinning your (and Matts) comment a bit further, Halfdeck... Google will not treat the wikipedia special, that's too much work (and requires manually editing value tables, bah! Manual labor is for the other engines). They will use algorithms to recognize sites like that and perhaps treat the links in them differently. How can you recognize a wikipedia-like site? If the links are being treated differently, perhaps they still have the same value even when nofollowed?

John

[rustybrick]

Danny Sullivan suggested in yesterdays daily search cast that if we ever link to wikipedia, we should use the nofollow.