I have certainly read a seo book saying this was the case for internal links - indeed, I seem to remember it saying that very large sites have a greater pr because of all the link flows between the pages.

However, surely if you were to  - in an extreme case - link out to 100 external sites from your home page, then wouldn't this reduce the pr of your home page?? That was why - logically - I asked if the same applied to internal links.

Recommended SEOers Resources

Many thanks for this explanation - it's clarified my thinking on the subject. I've now changed the internal link flow of my home site following replies to this thread and the other internal link building thread I started (I've put nofollow to a few unimporant pages and put extra links in to a key "website design suffolk" page) and - whether because of this or because of a few new external links or because of a mixture of the two - this page has now gone up in the serps again for "website design suffolk". Still a way to go as it hasn't got many good links yet, but it's now beating some less-optimised pages with a higher PR. Many thanks all.

Judging by this logic I guess we should provide LOTS of links (maybe even lots of links to the same page?) when creating content in sites like squidoo or weebly (where the system adds in lots of unwanted links - I was pretty miffed when using the Backlink check to see just how many links are on a Squidoo page!). That way - although the link juice might go to pages on your site you're not really targetting, at least your site as a whole gets a higher percentage of the pr?

Now lets go to PageRank Siloing.

All the pages and extensions for the user agent Googlebot are set to "noindex" in the robots.txt

Lets take one page calendar.php. Gogle should not index the page. And if it is was already indexed, next time Google crawls the page, should de-index it entirely.

Now lets say I am linking from my homepage to my calendar.php page editing the url like this http:// www.whateveryouwant. com/calendar.php/?bots=nocrawl

In the robots txt and added:

User-agent: Googlebot
Disallow: *bot=nocrawl
Noindex: *bot=nocrawl
Noindex: /calendar.php

We all know that robots.txt can accrue PR when an external site is linking to that page. Now what's the big deal?

First I ask the bot not to crawl or index the calendar.php page. Then maybe the bot comes arrives to that page via an external site (OBL), but still he will not index the page, but he will follow the links and will spread the PR.

How about that?

Please study the robots.txt of my forums: http://www.seoworkers.net/robots.txt or my http://www.seoworkers.com/robots.txt

Don't forget:

Guideline: "A page need to have an incoming link visible to Google to accumulate and pass on PageRank."

I am sure this post has some more watts light than other found on the Internet.

Are you sure? If I would agree, who can answer then? Rand? If not, who else?

Take a deep breath and read my current position on all this.

Matt if a page is not indexed by Google, or the page have a meta robots or robots.txt directive "noindex", the page can not build PR. If internal or external links point to a page being blocked from being indexed, it will pass the PR to the pages the page links to, and Google is allowed to index.

Adding the bots=nocrawl means that Googlebot cannot crawl, as long they still obey to robots.txt standards. The pages you disallow Googlebot to crawl still can receive PR, but only from other pages linking to it.

Forgive me for my arrogance, but I do not really care what Matt Cutts says, because I have the same opinion with Deepsand at WPW:

"As I'd earlier/elsewhere mentioned, Cutts plays fast & loose with the language; his statements are, as often as not, imprecise and/or self-contradictory. Whether this is by design or nature remains unknown."

As I said I am not a Matt Cutts fan or blind believer. I am contentiously running tests/experiments, and I share a part of those only, which I think it is clear why.

Nonsense. If you have page which was never and is still not indexed by Google, it does not exist for Google. Therefore no PageRank. If you have a link to such a page your disallowed Google to access, it could be if their paper is still valid, the PR assigned to that link will shared with all pages in the entire index.

Then you will say the nofollow is better, since the link will be invisible to google. Google reserves the right to pass PR if they think it will make sense. Another issue is that all 3 SE treat the "nofollow" attribute differently. Also you will probably have pages which must not be followed by Google, but that they should be followed by Yahoo, because they can help you boost your site relevancy, rankings, or whatever.

Being one step further in my research, came up with some new stuff. I avoid where possible using the bots=nocrawl thingy too. I prefer to use the meta robots directive "noindex" (not nofollow) for pages I link to and which I do not want it be indexed or have PR, I make sure that the page has at least one outgoing link to a page where I want to pass the PR. And this way I do create any dead end pages (dangling pages)

Example:

Page A links to Page B. Page B has the meta robots directive noindex. On Page B I have a link to Page C. If I want to Page A. Doesn't matter. You might would say, then you are creating a loop. That is not the case. If that was the case, think about a site navigation linking structure.

So what is the result of the above is that you do not leak any PR if you have A > B > A.
If you have A > B > C, the link PR on page A will pass it to page C. Again no loss.

If there are some other minor stuff in Googles PageRank algorithm we don't know, I think are irrelevant or need to care about.

That is all I can tell so far. I hope I explained it well enough.

As I explained above, you do not control the PageRank 100% in some cases, since we don't have the present paper.

Recommended SEOers Resources

I dont see how a page doesnt build PR but manages to pass it on? Isnt that a contradiction in the way pagerank works? If a page passes PR it has to have PR in the first place?

Matt Cutts says [http://www.askapache.com/seo/robotstxt-mattcutts-noindex.html] and i dont see any reason for him to lie about this -



So, if you want a page to take none of your site's Pagerank stock you have to block all it's incoming links with nofollow (or some equivalent link cloaking technique).


In this situation why would he be lying? Why would the quotes above not be true? Does matt want us all to read his comments, believe them and then f*ck up our websites? Also, a lot of what you are doing is based on what matt says so youre contradicting yourself. For example... you've mentioned various times about a site with X amount of pagerank can only ever have Y amount of pages in the main index. That came from matt c and forms the basis of the reasoning behind the implementation of pagerank sculpting (or siloing - they are the same thing).

- noindex does not mean that Google doesnt know it exists. It just means that it doesn't appear in the index. Pages are still returned in the SERPs that are noindexed when a specific url search is done but they dont have any snippets. So google reserve the right to display pages when you noindex them. Remember when ebay was noindexed? But it still showed up in a search for ebay? Or what about this in 2006 - http://www.webmasterworld.com/forum30/34757.htm

Do you have any proof of that? The evidence that i have from my own experiments and observations is that the nofollow link is completely removed from the links graph and all pr calculations.

Can you expand on this? Can you give me an example where a page should be seen by yahoo/msn but not by google?


But page B has also built some pagerank from page A, otherwise how does it pass pagerank to page C?

Matt, I am afraid we are running is circles.

You come up with an copy of an interview of Eric Enge with Matt Cutts, which I posted the original interview 2 or 3 times at that thread at WPW? http://www.stonetemple.com/articles/interview-matt-cutts.shtml Looks like you are not checking the links I am posting

I think you did not read this link I posted at WPW http://eduardblacquiere.com/non-indexed-page-pass-pagerank/.

Therefore, please try to answer me the following questions, before I go on.

- Does that mean that the page stores the PR for itself? If yes, why shouldn't the page be a dead end?
- If you have a page on your site that you never had a link to it and no one knows about it including Google, can that page have PR?
- If you have a page that for 6 months the PR bar is grey, does that still mean that the page has PR?
- If you have a page that was indexed in Google and had PR5, and all at a sudden you added to that page a "noindex" meta tag directive and you delete it manually in the GWT, does that page still has PR assigned.

Contradict? OK man. Lets try again. Can you please make a short list of the contradictions please?

I did not say that you said so. If a noindexed page has not a single link on it or it has a meta tags robot directive "nofollow", where can the PR be passed to? Nowhere or? So where will the PR go to? Will it be assigned to the noindexed page? When there are no links pointing to noindexed, how can PR be accumulated.

Googlebot picks up the link on page A and goes to the noindex page (page B). Then meta guard says: Googlebot this page is not for you. So this page does not want your PR. But Googlebot, give the PR to my friend's page C following the link you will find below. 

I have a page which was indexed and had PR5 and then I added an X-Robots directive  "noindex,follow,noarchive,nosnippet", went to GWT and requested a deletion. No disallow in robots.txt.
It have been deleted and the PR bar greyed out. I am linking to that page from very page of my site. And not hiding/blocking bots through nofollow attribute or the bots=nocrawl. It is a full authorized link. The PR bar remains grey. Never got white or a 0. Page B has outbound links to follow. Now what?

I think I already answered?


Thank you man!

If you can't trust the toolbar, what is the alternative that I can trust. If you say none, then you should accept the value you see there and based on that, figure out whats going on.


I can give you a concrete proof. Googlebot is a software. If that software did not come in touch with my application (web site or page) or my local drive, who cars  , it cannot assign or offer it something. PageRank is a Google patent and an Internet Standard.

Is Google estimating how many pages will be created on the Web and then they begin building PageRank for them? 

I am getting a "Buffer Overflow" warning man.