Do Links from Links Page Have Vaule?

[I, Brian]

Originally Posted by GoldDust

I have set up a number of seo test sites - and one of the sites has 1 pr5 linking to it and the test site has tlots of of content but no pr has shown up since creation of the test site - is there a minimum number of links needed to gain some pr?

Every page has PR (even if less than 1!) but so long as the PR5 link can be followed by Google then the expectation would be that such a page would have a PR that will show on the toolbar - but all in good time.

Note that the toolbar only displays PR publicly, and it is noly the public record that is updated every 4 weeks or so - Google itself always has a record of the real value and that is applied to the SERPs in the rolling index.

[compar]

Google does certainly use PageRank as a relevancy measure - that is precisely why so many people are PR possessed!

Where is your authority for that statement? PR is a mathematical calculation based on the PR value that has been passed to your page by backlinks, both internal and external. As such it is an indicator of the number pages you are linked to and the PR value of those pages. Now I would agree that pages with a high PR may have good content and that in part can explain why people want to link to them and why they have a high PR.

But I don't think PR has a thing to do with Google's judgement of the relevance of the link or with the placement of your site in the SERP. You yourself touched upon the real measure of relevancy when you said:

but lower PR pages can outrank higher PR pages on various other non-PR factors - anchor text for one.

That statement is true because Google doesn't use PR to judge relevance. They use anchor text and since Florida I suspect they are also using other types of semantical analysis. But I don't think they are now or have ever used PR as a relevance measure.

[I, Brian]

Heh, I think we are getting lost in semantics here.

Search Engines attempt to list results they consider relevant. for Google, one aspect of this is the implementation of PageRank.

I'm certainly not claiming that PR is involved in sorting themes and topics - merely that it is part of the overall ranking process.

For themes and topics we have implicit algos such as Hilltop (or even Topic Sensitive PageRank, if we follow Dan Thies reasoning on this) - and also, of course, the possibility of Applied Semantics technology being involved.

[compar]

(or even Topic Sensitive PageRank, if we follow Dan Thies reasoning on this)

I wish John Scott would jump in on this. I think you are dead wrong. PageRank is a mathematical calculation only and has absolutely nothing to do with relevance.

You have some Bible related sites. If someone with a PR8 Satanic Cult site put up a link to your web site it would add to your PR. But it sure wouldn't be relative.

There is topic sensitive relevancy value given to a link, but that is totally unrelated to PR.

John Scott, opinion please.

[compar]

On the basic question about the value of links from links pages. If a Google backlink search for the linking page get this result

<Your search - link:http://www.somedomain.com/public_htm...links16_9.html - did not match any documents.>

is it probable that a link to that site will be considered a link to to a "bad" neighborhood? And would you immediately drop that link?

[compar]

Brian,

Here is a direct quote from the Hilltop paper that you think verifies that PR is a measure of relevance.

PageRank [Page et al 98] is an algorithm to rank pages based on assumption b. ((b) authoritative pages tend to point to other authoritative pages.) It computes a query-independent authority score for every page on the Web and uses this score to rank the result set. Since PageRank is query-independent it cannot by itself distinguish between pages that are authoritative in general and pages that are authoritative on the query topic. In particular a web-site that is authoritative in general may contain a page that matches a certain query but is not an authority on the topic of the query. In particular, such a page may not be considered valuable within the community of users who author pages on the topic of the query.

I have added the bold to illustrate the main point, which is that PR assume that linking pages are authoritative, but cannot possibly know that they are, because the calculation is query-independent.

[I, Brian]

Originally Posted by compar

I wish John Scott would jump in on this. I think you are dead wrong. PageRank is a mathematical calculation only and has absolutely nothing to do with relevance.

You have some Bible related sites. If someone with a PR8 Satanic Cult site put up a link to your web site it would add to your PR. But it sure wouldn't be relative.

There is topic sensitive relevancy value given to a link, but that is totally unrelated to PR.

John Scott, opinion please.

Compar, when you are asking about "relevance" what are you actually asking about? Are you talking about "themes" or just general ranking?

I think you are quite misunderstanding what I am saying - I am not at all suggesting that PR works by a Hilltop system, or similar, and that PR is theme related - in case that is what you are reading my posts are saying.

I am simply saying that PR contributes to the ranking process. That's why you generally will not see very low PR pages rank no.1 for highly competitive terms. Anchor text - and other factors - comes into it - but PR normally is seen to act as a modifier to teh SERPs.

Does that help better?

[I, Brian]

Originally Posted by compar

Brian,

Here is a direct quote from the Hilltop paper that you think verifies that PR is a measure of relevance.

I have added the bold to illustrate the main point, which is that PR assume that linking pages are authoritative, but cannot possibly know that they are, because the calculation is query-independent.

Yes, I quite agree - I tried to make this statement quite explicitly in the post before your three:

Originally Posted by I, Brian

I'm certainly not claiming that PR is involved in sorting themes and topics - merely that it is part of the overall ranking process.

By that I mean that the exact same PR10 page "Bluewidgets" may rank higher than the exact same PR1 page "Bluewidgets" - because Google sees the PR 10 page as more relev...(hang on, I'd better not use that word!)...more suitable.

[John Scott]

Originally Posted by I, Brian

Google does certainly use PageRank as a relevancy measure - that is precisely why so many people are PR possessed!.

Brian,

I'm not understanding what you are saying. Are you simply saying that PR affecting the ranking of pages - or are you saying that PR transfer is affected by the relevancy of the link?

[I, Brian]

Hm, odd that I don't seem to be communicating very well. Is this a cross-Atlantic thing, or am I simply being sloppy with my technical terms again?

Are you simply saying that PR affecting the ranking of pages

Yes - my understanding of PR is that it contributes as a modifier to other present factors, when calculating the overall ranking of a particular page.

- or are you saying that PR transfer is affected by the relevancy of the link?

no - this would require a different type of concept algo - such as this.

[compar]

Ok One last kick at this can. In another thread http://www.internet-marketing-resear...529.html#43529 Johan007 wrote:

PR + Page Copy = Rank for non competative words

Anchor Link Text = Ranking of competative words

If you except this formula, and I tend to think it is accurate, then PR makes no contribution towards ranking for highly competitive search terms. This is contrary to what you Brian just said above.

The only contribution PR makes toward SERP placement is in the case of non competitive terms and there it is used in conjunction with the on page factors.

So back to the orginal question. I contend that if you want to rank for a highly competitive term you need 1000s of links as John suggests. These links need to have good anchor text and satisfy the other criteria that Google may be using today to determine relevance, such as local rank, hilltop, and applied semantics, but their PR contribution is unimportant.

Therefore even if they are from links pages with low PR value they may still make a major contribution towards relevancy ranking for highly competitive terms. And frankly if you want 1000s of backlinks I don't see how you are going to accomplish this without doing reciprocal linking which results in your backlinks coming from links pages.

[I, Brian]

Ah, now, compar - after exhorting for a simple reference to PR being relevant for SERPs, or no, you then claim I'm wrong because you were only ever talking about "competitive" keywords anyway! And then repeat precisely what you've been told on this thread!!

Google certainly does seem to be delivering a different basic set of results for different sets of words - but these are not necessarily "competitive" - and being able to clearly state that PR is not at all involved in one of these sets of words really does require some support of that presumption - especially if it's going to be so strongly postulated.

The simple fact is, the post-Floridan system is a mystery. Some people see hints and suggestions of specific algos at work - we've mentioned Hilltop, LocalRank, and Applied Semantics technology at work here - even a new form of PageRank called Topic Sensitive PageRank.

However, we should be very careful, here, not to imply there is any known "truth" of how post-Florida Google operates in its exact details. We may infer - and some people may strike out with brave hypotheses - but these are otherwise only based on partial observations.

Do you have any specific perspective on how PR is being held accountably absent post-Florida for specific keyword ranges?

[compar]

Well I have to apologize. I thought for sure I read you say PR contributed to competitive term ranking. But on rereading I cannot find it. So I'm sorry if I quoted you falsely.

The postulation is not mine it is John Scotts. But I happen to agree. I think most people agree that you can get ranked for infrequently used search terms virtually on content alone. But if you want to rank well for frequently used, highly commercial, terms you need 1000s of backlinks.

The evidence of the latter is proven by the fact that people keep asking on forums like this "how come site X ranks well for search term Y when Y does not appear anywhere on the page?"

The answer always is that the page has a lot of links that use Y in their anchor text. So this really is not a very daring postualtion. For many SEO gurus it is virtually a given.

There is no such thing as "Topic Sensitive PageRank". I think you got that from a missreading of the Hilltop paper. They talked about Topic Sensitive Ranking, i.e. relevance ranking. They never suggested that there methods should be applied or could be applied to PageRank.

Here is exactly what they said on this subject:

An alternative to PageRank is Topic Distillation ...........................Unlike PageRank, Topic Distillation is only applicable to broad queries, since it requires the presence of a community of pages on the topic.

So instead of suggesting "Topic Sensitive PageRank" they say exactly the opposite they say there is no similarity between Topic Distillation and PageRank.

[I, Brian]

You are certainly right that on-page elements alone won't make a site rank first for highly competitive search terms - and John's method of brute force links and anchor text is one solution.

However, that doesn't preclude other possibilities - *if* Hilltop and themes are to be accepted to play some role (and they may not) then there's room for a far more selective strategy.

In this way, an "on-theme" link exchange program could attempt to take advantage of issues such as theme implicitly, and "expert pages" from Hilltop (If they can get away with it). But link exchange alone is not likely to be of overwhelming benefit - it really needs to be part of a wider SEO strategy.

Interesting, though - there is actually a "Topic Sensitive PageRank" - linked to it discretely - here's the full paper at Stanford:

http://www.stanford.edu/~taherh/pape...erank-tkde.pdf

Dan Thies has suggested that TSPR is playing a specific role in the post-Florida results.

[Fruit & Veg]

Screw PageRank. Forget about it. Google use many criteria to rank a page - you could have PR8 and still be No.67 in the SERP - what are you going to do about that? Delete links with less than PR5? That's nuts.

BBC.co.uk site has PR9 - that proves that PR is horse muck cause the BBC website is second to none.

Getting back to the original question, get links from sites that your users will appreciate. If they have no PR - they might do in a few months. As someone else pointed out earlier - links can drive traffic. Get links from directories and info sites as they don't usually require a reciprocal link.

[John Scott]

The topical pagerank would be relevancy based, but it's not a PageRank you can see in the toolbar. It's two entirely different calculations, and we don't have access to the topical pagerank.

[compar]

Brian,

I didn't recognize you in your new disguise Are you hiding or turning over a new leaf.

Ok Let explore this a bit more. The problem with the term PageRank is that it can mean a couple of things. PageRank as reported by Google is a mathematical transfer or sharing of PageRank between linking sites.

However relevancy factor can also be said to contribute to a pages ranking. I like to refer to this as index or SERP placement. But it is where a site ranks in the SERP. It is a very easy semantical step from there to talk about a page's rank or even page rank when what is meant is what I call index placement.

It is also very easy to forget what PageRank really is. It is the only published measure Google offers and a great many people believe there is a direct one to one relationship between PageRank and ranking in the index -- my index placement. So everytime we see the word 'rank' in regards to links we have to look very hard at the context to see if they are talking about PageRank or index rank.

That is why I said people are PR possessed. They think that PageRank is the only thing needed to get high index placement and it is not. It probably is one of the least important factors.

Applied semantics is something that could only be used to judge relevance. My reading of the local rank patent application makes me think that it is really Google's attempt to patent the technology or methods described in the Hilltop paper. In any case neither of these technologies or method are being used in PageRank calculations. They may be being used to determine relevancy ranking and index placement. Whereas pre Florida it appeared that anchor text was the major relevancy agent.

Now as too John's brute force method. To get a 1000 links does not mean that you have to get non relevant links. For popular, commercial sites it would be easy to get a 1000 links from site with exactly the same basic content or information as your site. It means linking directly to your competitors, but if your site isn't interesting enough to hold your visitor you are going to lose them anyway. Whether they go to your competitor or just click away from your site, you've lost them in either case.

And if I have several hundred links on my links page to my competitors doesn't that qualify my site as an 'expert' site? So I am accomplishing or fulfilling all these thematic, local rank, Hilltop criteria for a high relevancy ranking. Not high PageRank but high ranking of my page in the SERPs or index.

[Rob]

Originally Posted by compar

The question is, is there anyway to get 1000s of links without building large link directory sections in your web site and exchanging links with a lot of pages with little or no PR.

I'd say, only if people voluntarily link to your website, not requiring a link back. These (often personal) pages are usually not of much PR value though (except for directories such as Yahoo!). You will also need quality (unique) content and a lot of time, or good press coverage.

[I, Brian]

Originally Posted by compar

Brian,

I didn't recognize you in your new disguise Are you hiding or turning over a new leaf.

Oh, I'm changing my profile across the internet.

As for the actual discussion about PR: I don't think there's any real disagreement betwen us - the stumbling block seems to have been a miscommunication, nothing more.

[altyfc]

The normal way, if not the only way, to get thousands of links is by participating in a reciprocal linking scheme.

We have over 60,000 links to Aardvark Travel which I am pretty proud of (or just around 9,000 if you do link: on Google). Yet we don't participate in a reciprocal link scheme, per se... we welcome people to link to our site, and actively encourage it, but it is not a requirement should they want us to link to them.

Aaron