| 1:38 pm on Mar 4, 2004 (gmt 0)|
a links.htm page may hold little value for most sites and very rarely serves as a landing page. So if Google penalises pages named "links" then most sites won't feel any direct pinch. However, I can't see Google penalising links pages per se or, in fact, sites that have links pages - there are just too many.
What is possible is that Google will ignore links coming to you from pages that are clearly links pages. There is some logic in this. Links pages are usually created to help boost a site's popularity with SEs and they don't reflect "natural" linking patterns (where sites link to each other for relevant content).
If you already have a links page you may to rename it to be on the safe side, or better still create a directory. If you are working on a reciprocal linking program - that may not be a good idea at this time. If you are getting one way links from other sites try and ensure they are from some page other than their links page.
| 11:33 am on Mar 5, 2004 (gmt 0)|
Could Google really actually penalise links pages though?
A lot of sites will have links pages that are not intended mainly to help search rankings through reciprocal links.
By penalising links pages Google would be penalising loads of innocent people who just want to provide a good resource section for visitors?
| 11:44 am on Mar 5, 2004 (gmt 0)|
I would say the majority of pages "links" are full of crap. That is why I call mine anything but that. If something is mostly crap you get rid of it even though there may be some good in it.
| 1:55 pm on Mar 5, 2004 (gmt 0)|
Rubbish. There are millions of useful link pages covering topics from cancer treatments to using a PC for recording guitars.
| 2:13 pm on Mar 5, 2004 (gmt 0)|
|There are millions of useful link pages covering topics from cancer treatments to using a PC for recording guitars. |
Absolutely! If I cannot find the information I require on a site, I look for a link page to point me in the correct direction. And I mean a link page, easily identifiable as such, and hopefully annotated. The last thing I want is to scour through general pages in the hope their may be a useful link embedded within the text.
| 2:24 pm on Mar 5, 2004 (gmt 0)|
Setting all links pages on a PR0 might be a way to go for G.
Google to neglect or not to promote links pages still does not dilute their value built-into a web site where you might want to use that page for further information.
I just don't see links pages as SERP landing pages per se, so a PR0 would clean up the whole linking mess as of the SERPs, still you can find those pages at every web site if you want, need to use it.
| 5:20 pm on Mar 5, 2004 (gmt 0)|
Thought experiment: Let's say Google did decide to penalise pages called "links.html". What would happen?
Well, all the spammers who read these boards and know about these things would simply rename their pages to "fred.html". All the innocent people genuinely trying to provide useful links wouldn't know any better and would get hammered.
Hardly the result they'd be looking for.
I think these rumours arise because of a misunderstanding of PR. The way PR works, any page containing many links automatically gives relatively little value to the linked pages. And if Google tweak the algo more towards semantics and away from PR then the value of being listed on a links page might go down further. That's not the same as saying that links.html will be penalised.
If the results of Austin are anything to go by Google are more likely to increase the value of pages containing links but no content. :(
| 5:56 pm on Mar 5, 2004 (gmt 0)|
Just to be safe, I quit naming my links pages "links" over a year ago. In fact, I started weaving the links into the site content, since I only put relevant links on my site anyway. I ask the same of my exchange partners.
Relevant link exchanges are a valuable source of highly targeted traffic and don't depend on the whims of search engines. Might as well do it up right and get double returns on the effort.
| 5:59 pm on Mar 5, 2004 (gmt 0)|
1 My links page is called enllašos.htm, wich in catalan means links. Then a links keyword penalization wouldn't affect me unless they use all the synonimous of the word in all languages ;P
2 I do not believe G will ban link pages. They ARE natural link patterns! Reciprocal linking has always been a way to share traffic among sites much before link popularity existed. It's true that there are sites that, since G created the PR technology, abuse of reciprocal linking, but these are much more under control than it seems by some reasons:
a) Big files are not fully readed by the bots. Nobody knows where this limit is but very people place it between the 100 and 200 kbytes. Link pages too big will be partially ignored.
b) The PR score passed by a link is in inverse proportion with the total number of links in the page. A page will distribute the same PR either if it has one link or if it has some millions. Links pages with a lot of links do not pass too many PR to each page.
c) Links pages use to be deep in the sites structure (normally there is only a link to them in the homepage and/or the sitemap), so they have not too much PR to pass.
d) big sites use to automatize their links pages with databases and scripts, geting dynamic URLs. Google does not crawl too well this kind of URLs
If even so Google starts to ban links pages, are the PR algo will go to garbage: searching for widgets will give only results of personal pages from people who like widgets, are interested in widgets, etc.
In other words: PR algo is good enough to keep under control links pages without any additional penalizations.
All what I've said is only a point of view.
PS: sorry, my links file is called enllacos.htm. It's not safe to use characters like š, ˝, ´, etc in the web ;)[/edit]
| 6:46 pm on Mar 5, 2004 (gmt 0)|
If your original question is are they penalizing the page because it is called links.htm the answer is not in my experiences - if you are asking if they would do it in the future - who knows - but I doubt it for the same reasons given here already. Should it be done - not in my opinion - why tear down the structure of the web that has been based on these filenames since the beginning, just because a few use them for spam or other reasons? That would be akin to banning all pages that are named keyword-keyword.html because a small handful of people use them for spam.
| 7:21 pm on Mar 5, 2004 (gmt 0)|
My links.php page, having PR 5, does not show up as back link for the external pages it links to. With 24 outgoing links it should be able to pass a decent amount of PR, however, one of the pages it links to has only PR 1.
It does show up as back link to the internal pages it links to, though.
There are only 2 reciprocal links on it.
I guess Google is treating not only my Links page like that.
As that page is there for visitors looking for good info related to that on my site, I don't really mind that the page doesn't pass on PR.
| 9:51 pm on Mar 5, 2004 (gmt 0)|
It's not like this is in doubt.
Skip down the first hundred or so sites and then check links from the next several hundred pages to the sites they link to. Very few show up as backlinks, and where PR should be clearly passed, it does not (that's much harder to find since they are few medium PR links pages with a few links linking to very low PR pages that should be getting a boost).
People can either do the investigating or not, but it is an objective fact these pages seldom show as backlinks to other domains, and very clear that they do not pass PR.
Remember, these are the links.html pages that Google algorithmically rates the best, so whatever percentage of the time Google shows this group among backlinks, the percentage is very very likely to be even lower on the Internet as a whole.
| 10:04 pm on Mar 5, 2004 (gmt 0)|
I can't see Google depreciating the value of a page named links.htm just because of the way it's named.
Most pages named links.htm have tons of irrelevant links or links piled together that are not organized. I CAN see Google depreciating a page like that. It just so happens most of these amaturish pages happen to have the name of links.htm. See the correlation?
Wouldn't it be better to divide your links up into seperate sections naming each section page on topic?
Also I think a better term for "links" is "resources". Like it or not, an outgoing link should serve as a resource for your visitors.
[edited by: synergy at 10:51 pm (utc) on Mar. 5, 2004]
| 10:26 pm on Mar 5, 2004 (gmt 0)|
If Google were ever to turn its cannons on links pages, it wouldn't identify them by filename. Computers don't use words they use numbers.
Incidentally and quite off-topic, I heard a joke the other day:
There are 10 types of people in the world.
Those who understand binary and those who don't.
| 11:02 pm on Mar 5, 2004 (gmt 0)|
For what its worth, I moved my outbound links from "links.html" to "alinksynonym.html" and the page started showing up in backlink checks after the next update. I think Google would ignore links.html pages and would not care about any colateral damage.
| 11:53 pm on Mar 5, 2004 (gmt 0)|
I have a links.html page to various shareware sites. Somehow, it has a PR of 3 despite the fact that it has only one inbound link and that is from a page that has no inbound links - it's part of a frameset.
So, I guess that
a) Google does follow framesets (I think I read otherwise).
b) Google does not ban links pages.
I only noticed that this page was indexed a few weeks ago when it came up in SERPS for something I was looking for.
However, because the page only has a PR of 3, I can't tell if any of the links on it are counted, but the page itself is definitely not penalised.
| 12:21 pm on Mar 6, 2004 (gmt 0)|
|Google does follow framesets |
Maybe, but in any case, better if you have a <noframes> tag that every bot can crawl.
| 12:46 pm on Mar 6, 2004 (gmt 0)|
Why not rename your links page to:
The evidence is only anecdotal, but Google seems to admire these pages :)
| 3:51 pm on Mar 9, 2004 (gmt 0)|
|useless-directory-not-providing-what-you-searched-for-but providing-links-to-what-you-were-actually-searching-for.html |
Even when I'm almost sure that G won't ban links pages, it's also evident that to do that, G would search for the text links in URLs. Your filename wouldn't be convincing ;)
| 4:03 pm on Mar 9, 2004 (gmt 0)|
Amazing how many people in here are thinking black & white when with algo's, it's better to think in thousands of shades of gray.
Read steveb's post. Do the homework. Read back through the threads in WW (there have been about 700 on this topic).
There's hard evidence all over the place. All one needs to do is look. Or choose to assume that one's opinion is likely to be correct.
Hint: "ban" is far too strong a word. ;-)
| 5:04 pm on Mar 9, 2004 (gmt 0)|
|And if Google tweak the algo more towards semantics and away from PR then the value of being listed on a links page might go down further. That's not the same as saying that links.html will be penalised. |
I think this would be a major mistake... yes, links can be manipulated, but "semantics" (I assume you mean on-page elements) can be much more easily manipulated. Link exchanges take a tremendous amount of time... if done correctly.
|If Google were ever to turn its cannons on links pages, it wouldn't identify them by filename. Computers don't use words they use numbers. |
Actually, they do penalize themeindex.html (default zeus link index page) by name. At one time, I had a page with that name, renamed it to index.php and the PR came up very quickly. No other changes were made.
| 5:17 pm on Mar 9, 2004 (gmt 0)|
Some of us have done the homework. I have read through reams of stuff on this subject, and all I have been able to deduce is there are two camps, the yeas and the nays. There is little hard evidence that stands up to scrutiny, steveb's post being a case in point.
For it to be a valid test, we would have to compare the amount of PR passed from these pages now with the PR passed from these pages say about 6 months ago. So unless someone had the foresight to take a snapshot we are none the wiser.
And even if steveb's results were found to be valid, that would not prove any sort of penalty. One would have to filter out all the changes in the algorithm before coming to that judgment.
It is much more likely that the majority of link pages have always passed low or invisible PR. They are frequently low in the PR hierarchy, have many links with little text, and I suspect a considerable number exceed the suggested limit of 100 links per page. Changes in the algorithm may have heightened our perception of this.
We should also bear in mind that concern has been expressed that backlinks in general (not just from link pages) are not showing as we would have expected a few months ago.
| 5:39 pm on Mar 9, 2004 (gmt 0)|
steveb certainly has a point, but he pushes it too hard.
something is going on with some pages called /links.* , but that is only one part of the equation.
If you look at the pages that do not show up as backlinks, you will notice that they are are simply a big long list of links. There is very little additional content there. The ones that do show up all seem to have additional information about those links.
None of them seem to be list of *only* reciprocal links. There are links to other sites, and seem to be set up as a true resource for users.
Now, you can just go ahead and change the name, but if google decides that they keep wanting to go in this direction, don't you think that they will expand it beyond /links.* out to look at all pages that appear to display this behavior?
Instead of just renaming your page, why not fix your page so it will survive in the future too?
| 6:35 pm on Mar 9, 2004 (gmt 0)|
|Amazing how many people in here are thinking black & white when with algo's, it's better to think in thousands of shades of gray. |
Yes, this should be the way search engines work. HOWEVER, if Florida taught us anything it is that Google's algos are either
a) littered with bugs
b) littered with discontinuities (maths students might prefer the term singularities).
Personally, I believe both feature at the moment.
If discontinuities are present in the algo (e.g. sending pages that were top 10 to oblivion) then thinking entirely in shades of gray just isn't sufficient.
I've said it before, but I still think it's worth repeating. If Google engineers believe they can maintain a search engine using algos that have discontinuities (I call them "non-linear" algos for simplicity) then they are wrong. They'll end up creating a patchwork of sticking plasters. As software engineers they may think they know it all - they don't. No control engineer would tread the path that Google seems to have taken.
| 7:31 pm on Mar 9, 2004 (gmt 0)|
>>shades of gray...
What I meant was that the discussion seemed to be going in terms of 'banned' or 'not banned'. Clearly they are not all banned. It's not black and white like that...
But also clearly, they do not show up as often as they should in various kinds of investigations, already alluded to. As a group, they are somehow being discounted.
But of course, some do far better than others, and are thus far more valuable. Getting a link from an edu page is extremely valuable. Almost none of them will do recip links however.
I don't have the page bookmarked but as I recall, Brett looked into this a while back. If memory serves, what he found was that the number of "links.***" pages in the SERP's had declined markedly over time (or was it in the visible backlinks?).
Either way, G probably doesn't want artificially enhanced PR by way of heavy duty link exchanging. It's much better for G if people give links as votes, not as payment for other links. So it stands to reason that G will look for signals of artificiality and dampen related pages in some fashion, while attempting to retain assigned credit for pages that are more useful and designed for the user.
| 8:58 pm on Mar 9, 2004 (gmt 0)|
|So it stands to reason that G will look for signals of artificiality and dampen related pages in some fashion, while attempting to retain assigned credit for pages that are more useful and designed for the user. |
I quite agree, but Google may be doing this generally without having to decide if a page is a "links" page whatever it is called. For instance Google could simply downgrade the value of outgoing links from any page when the proportion of outgoing links to text reaches a certain level. This would explain why user-friendly text-heavy links pages may show up, but a page just full of links doesn't. But of course, this is just more conjecture.
Hopefully the situation may become clearer after the next backlinks and PR updates.
| 9:14 pm on Mar 9, 2004 (gmt 0)|
It's like a war - a balance of probabilities. There are thousands of millions of pages out there, and they are being ranked with only 10^2 , perhaps 10^3 parameters by an algo.
There will always be casualties.
But continuing in this military vein: Florida and Austin wiped out not only the enemy, but half of the own side as well.
It took Google (a shameful, IMO) 4 months to put this right.
Let's hope they can minimise this 'friendly fire' in future.
Certainly if the conflict breaks out again, I will direct my white-hatted troops to take cover at the nearest Amazon warehouse. Or failing that, one of those unsteady barricades put up by the directories.
Once the nonsense is over again, they can return to the security of a decent Internet site here in blighty - if it's still here :)
| 9:46 pm on Mar 9, 2004 (gmt 0)|
|Google may be doing this generally without having to decide if a page is a "links" page whatever it is called |
Understand. But my point, with respect to this thread, was that I believe Brett's check into this issue was specifically related to pages named " links.*** "
It seemed to be pretty compelling evidence that the actual filename " links.**** " was singled out for dampening...probably in conjunction with other factors.
Unfortunately, I can't find that particular thread right now.
| 12:24 am on Mar 10, 2004 (gmt 0)|
"It is much more likely that the majority of link pages have always passed low or invisible PR."
Just because you have not done any tests doesn't mean others haven't.
I personally was very skeptical about this in the fall because I found exceptions. It became obviously a reason for concern the next update as the percentages seemed reversed, very few links pages showed up. I saw I wasn't getting benefit I should, and that I wasn't delivering what I should. So I changed the pages, poof, next update everything is as it should be. My pages are showing as backlinks and passing PR, and two sites I contacted who also changed are giving me powerful links now instead of nothing.
BigDave's (to try to agree with him on this...) point is important to consider, but something we can't be sure about. Google's algorithm basically can do anything it wants. It can say if a links.htm page has less than twenty links, is on .gov domain, has x% words not in hyperlinks, etc., that this page will show as a backlink and pass PR. But that is a much more complex situation, which frankly isn't worth considering for most people. The quick and dirty fix of changing a page name, or contacting webmasters who give you a key link that doesn't show, these are easy and waste an appropriate amount of time on what should be a minor page on a domain.
The most basic thing is there is no reason to name a new page links.tld and plenty reason to name it anything else under the sun. The next basic thing to understand is that Google has simplistically made a move toward addressing a major issue for them: separating the influence of genuine "voting" linking from baloney-pure-seo linking. This is a very sticky issue, and we should expect to see more algorithmically complex "solutions" to the problem from Google in the future, despite the rather dippy first step they took. So... changing a page may do you no good in the long run, but it probably will help you three weeks from now.
| This 39 message thread spans 2 pages: 39 (  2 ) > > |