PageRank leakage

Guys,

who knows if some page have only 10 links, for example, and 5 of them are linked to the pages that are disallowed in robots.txt file what happens with 50% of transferred PageRank to this pages? Is it lost?

Best regards,
Paul

 

if I had to guess, I would think that google would probably ignore the links entirely if they're disallowed in robots.txt

I could be wrong, but that'd definitely be my guess

 

But the problem is that I have such a site with PR6 homepage and just PR4 inside pages with the link from home page!!! Not PR5, but PR4. I have a lot of disallowed pages!

 

A site can not leak PageRank. It is a myth started by pretend SEOs that either don't know what they are talking about or are trying to scam you into using their services. Compar posted a great article about these types of myths.

Linking Myths

 

Please explain, you all think that the real forumula for transferring PageRank value is:

Value transfered = Page's PR * 85 / (number of links minus links to the content that are not indexed by Google).

Am I right?

Best regards,
Paul

 

Does it really make sense to have a site's pagerank diminished because it has links? If this was true, everything would have a pagerank of zero.

 

No, of course not. This "leaking pagerank" baloney is amazing at how prolific it has become. Pages do not leak. To anyone that says they do, I challenge you to prove it beyond any reasonable doubt. Of course now I'll be challenged to prove that it doesn't, which I can't. Only Google can do that. But the burden of proof rests on proving that leaking occurs since on it's face, the argument is so full of holes.

(off my soapbox now)

 

Guys,

You don't hear me. Look at the formula below
---
PR value transfered = Page's PR * 85% / (total number of all links on the page minus links to the content that are not indexed by Google).
---

Do you really think Google subtracts links to non existent for him pages?

 

The noted physicist Lord Kelvin once famously remarked "I often say that when you can measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meagre and unsatisfactory kind; it may be the beginning of knowledge, but you have scarcely in your thoughts advanced to the state of Science, whatever the matter may be."

With that in mind, one might be well rewarded to look in on this mathematically analytic article about Page Rank.

One should read the entire article, but these remarks appear in it:

And later:

And a little later yet:

One may agree or disagree, but it is, I think, a sufficient answer to the question of where such "crazy" ideas come from. They come from mathematical analysis of the similarly mathematical papers Google's people have published.

If someone wants to read the entire paper cited above then critique its mathematical analysis here, fine: let's see it. Till then, I stand with Lord Kelvin: put it in numbers for me.

 

Exactly what I was asking for! I wasn't talking about the drain (leakage) from the individual page, I meant the whole site.

I have site that have something about 5,000,000 links within the site (20,000 pages indexed by Google with 250 links on each). And 1,000,000 links points to the pages of my site that are disallowed in robots.txt and not indexed by Google! So we can count them for outbound non-reciprocal links because PR value transfered to them in vane (They are not indexed at all) and never comes back from them because Google doesn't index them at all!

So it comes I leak 20% of my PageRank value!!!!

What can I do? May be the use of
---
<a href="http://www.my_site.com/page_allowed_for_Google.html" OnClick = "window.open('/page_disallowed_for_Google.html', '', 'scrollbars=1,menubar=1,resizable=1,width=800,height=600'); return false;">
---
will help?

And what happens to the links that look like <a href="#">? Do they reduce PR value transfered from page to page and cause drain (Here I always talking about the whole site!)

It's not mere interest. I really have site with PR6 Home page (it have very many inbound links pointing to it) and all other pages on the site have only PR4- value! So I guess it's because 20% leakage (drain)!

PLEASE HELP WITH A GOOD ADVICE!!!

PS If you ask me what kind of pages they are, here is the short extract from the article

So they are the pages with ScreenShots and all of them are identical and have difference with only one pic of the screenshot. So I disallowed this content for all search engines but it causes the PageRank leakage

 

A link to a page not known to Google--whether because they simply haven't gotten to it or because it is blocked off--is a "dangling link". Dangling links have, by Google's founders' own assertions, essentially no effect on PR calculation (see the paper I cited earlier), which only makes sense. So there should be no PR loss from links to hidden or blocked pages.

 

But why I dont' have PR5 pages on the site at all? None of them!

It's not mere interest. I really have site with PR6 Home page (it have very many inbound links pointing to it, Google shows 535 backlink, for example) and all other pages don't have inbound links from outside of my site and they all have only PR4 and below value! So I guess it's because 20% leakage (drain)!

 

Just so. But there is no connection between those pages' being hidden from the robots and the question of blocked/hidden links. If, by Martian mental powers, someone tomorrow hypnotized every single page owner who was linking to those blocked pages into hiding their links to those pages, those pages' PR would dry up to zero as fast as Google can re-reckon PR.

 

So If pages disallowed by robots.txt have PR value, then some internal pages passes PR to them and we have PR leakage!

How could it be that Google passes PR to disallowed pages but doesn't subtract this passed value from the whole 85% devided by the number of links value? So we have here the described in the article situation when there's a link to the page and page don't have backlink! And it's leakage from the site's whole value!

 

That is VERY interesting - thanx for sharing ...
alek

 

It's been a while since I read Brin & Page, but if I remember correctly the reason dangling links are removed from the calculation process is that for each iteration of the formula they would receive PR while not passing any on, thus skewing the results. The dangling links are added back near the end of the process and are assigned PR.

If that is the case then those pages will dilute the PR passed to your other links.

 

It says:

"Because dangling links do not affect the ranking of any other page directly, we simply remove them from the system until all the PageRanks are calculated. After all the PageRanks are calculated they can be added back in without affecting things significantly."

As one expert translated it, That's the effect of functionally useful, dangling links in the site. There's no overall PageRank loss. (http://www.webworkshop.net/pagerank.html#dangling_links)

So they will not dilute the PR passed to your other (non-dangling) links.

 

I agree that Phil Craven is one of the leading authorities on PR, and I'm not looking to debate him (he'd just squash me like a bug). I do think the dangling links question and leakage does deserve further examination. Your quote from Brin & Page is correct, but is it what they are really doing? If they truly remove dangling links from the equation until all PR is calculated then shouldn't the dangling links pages show a gray bar and no PR? In practice those pages are showing approximately the PR you would expect based on the pages linking to them. At what point in the process is google attributing PR to those pages, and how do we know that this doesn't dilute the PR distribution?

I don't claim to have the answers, I just don't think we have enough information to conclude that no PR is lost through dangling links. Phil, you out there? Care to set us straight?