Google PR Funny behaviour

have never seen this before!

[moltar]

So I was just playing around with URLs and noticed this strange behavious.

URL: [example.com...]
PR: 8

URL: [example.com...]
PR: 7

URL: [example.com...]
PR: 6

URL: [example.com...]
PR: 5

URL: [example.com...]
PR: 4

URL: [example.com...]
PR: 3

URL: [example.com...]
PR: 2

URL: [example.com...]
PR: 1

URL: [example.com...]
PR: 0

[Brett_Tabke]

Ever hear the older seo's talk about just 'throwing it all in root'? Now you know why. Directory depth certainly is a criteria in the algos of the majors. It is also in the "pr guessing" game algo that the foolbar uses.

[rookiecrd1]

Very interesting. Does this mean if you have a pr 8 site its better to make a page at www.site.com/widgets then it is to link to a site made at www.newsite.com

What do you guys think?

I have tended to notice that subpages on a high pr site seem to get ranked much higher in serps then a separate page that you link to from that high pr site.

[moltar]

Now you know why.

Yes, now I actually beleive that theory :)

[JaySmith]

Very interesting. Does this mean if you have a pr 8 site its better to make a page at www.site.com/widgets then it is to link to a site made at www.newsite.com
What do you guys think?

I have tended to notice that subpages on a high pr site seem to get ranked much higher in serps then a separate page that you link to from that high pr site

It's definately better to create the page on the old site for ranking purposes. There is no Sandbox for pages so it will get your indexed and in the serps quicker.

This is one of Google's biggest flaws. I am seeing more and more of www.authoritysite.com/choose-high-paying-keyword-that-has-nothing-to-do-with-site/...

I have been able to get indexed and listed in top 2 pages in days using isp web pages that give you a subdirectory from their home page while my 7 month old site is sitting in the sandbox.

What I did is use these pages to "brand" my real site so that I get the branding I need. So to answer your question, its a good idea to do both.

[edited by: JaySmith at 3:55 pm (utc) on April 19, 2005]

[moltar]

using isp web pages that give you a subdirectory from their home page

That is amazing idea. My ISP gives me 6 home pages with 15MB space each!

[JaySmith]

I am on page one for high money compitive keywords with only a SINGLE link from my sandboxed site. This shows me that google RESPECTS my sandboxed site (PR7) to pass the rank but due to its kneejerk sandbox filter, pulls it from the results (not in top 1000).

For people who think the sandbox does not exist, try this.. You will see there is definately some type of aging filter put in place.

[webdude]

I tried this on various sites of mine and the PR seems to stay the same. Any ideas why?

[Atomic]

That is amazing idea. My ISP gives me 6 home pages with 15MB space each!

I have retained an AOL account for over a decade because I have 7 logons and each has its own web page "space" where I can add content. So it's like having 7 websites for about $14.98 with the bring your own access plan.

[diamondgrl]

I see the same as webdude. I don't get it. Are other people seeing this?

[steveb]

"I tried this on various sites of mine and the PR seems to stay the same."

It's totally untrue. Ridiculous thread.

[moltar]

I tried this on a public, very well know site. You can duplicated it yourself.

The original tests were performed on W3C Link Checker using Firefox v1.0.2 (with Google Pagerank Status v0.9.3) on Ubuntu Linux. I just tested this on Windows using both IE and FF. It produced the same results.

PR 8: http:// validator.w3.org/checklink?uri=www.example.com 
PR 1: http:// validator.w3.org////////checklink?uri=www.example.com

[Marc_P]

the foolbar

I wasn't hip to that new terminology, I like it, or is it just a typo? ;-)

[trimmer80]

i have finally found an example of this behaviour.

Click on your inbox to go to your stickymail, ^^^above^^^, then look at the url and then add slashes between the .com and the stickymail.cgi eg.
.com////stickymail.cgi

PR decreases with every slash.

I have always wondered how google can allocate pr to these pages. My guess that it doesn't, but the toolbar gives an approxomation of the pr by looking at its directory depth and root domains pr.

[steveb]

You might want to use the Google toolbar instead, especially before posting something so obviously wrong.

[slate.msn.com...]

PR7, must be PRonemillion with out the slashes

[larryhatch]

Each slash drops (toolbar) PR by one. Don't we want to go the other way?

Is there such a thing as a reverse slash, a negative slash or ..
how about the BACK-slash! Sort of an anti-slash?

OK, never mind. -Larry

[Freedom]

I remember the old days when they used to update PageRank. It may not mean anything, but it would make it easier for my link campaign on a rebuilt site.

[ncreegan]

You might want to use the Google toolbar instead, especially before posting something so obviously wrong.
[slate.msn.com...]

PR7, must be PRonemillion with out the slashes

You're comparing a site with mod rewrite and dynamic content to sites without.

[diamondgrl]

I still don't quite get it. I've tried this on sites with dynamic content and static .html, sites with mod rewrite and sites without. I still get no PR dampening effect as described here.

On the other hand, I did try that validator site that moltar suggested and was able to repeat what he showed.

So unless my brain hasn't kicked into gear, it seems to affect only certain sites. Can anybody explain this?

[webdude]

I'm with diamondgrl,

Only can see this on the one example, but no others.

[trimmer80]

i think it happens for pages that either cannot be crawled or maybe have been crawled with a different query string. Thus my stickymail example.
I would say that it is not PR at all but just the toolbar trying to estimate pr.

oh and Steveb, I guess as usual, everyone else is wrong and you are right.

Try this

[webmasterworld.com...]

[webmasterworld.com...]

[webmasterworld.com...]

[webmasterworld.com...]

[moltar]

You might want to use the Google toolbar instead, especially before posting something so obviously wrong.

I said later on:

I just tested this on Windows using both IE and FF.

And IE obviosly had a real toolbar, not an extension ;)

i think it happens for pages that either cannot be crawled or maybe have been crawled with a different query string. Thus my stickymail example.
I would say that it is not PR at all but just the toolbar trying to estimate pr.

In the case with W3C link checker, yes, it cannot be crawled, but many people link to it. PR was calculated based on incoming links, so it shouldn't affect anything.

[jomaxx]

Why is everyone so excited about this? This isn't "real" PR as used in Google's ranking algorithms, it's just an artifact of the way the toolbar works. I've seen this same basic demonstration of PR degradation years ago.

[composer]

This page has PR 10 (Real PR)
www.google.com///////////about.html

Can someone try this with Toolbar pls?

[steveb]

"oh and Steveb, I guess as usual, everyone else is wrong and you are right."

What are you talking about? What was posted originally is total nonsense, as anyone can check with any URL. It's not news either that the Google toolbar has never read complex query strings right, ever, but what does that have to do with the original post? And what's then point of post those stickymail URLs where none have pagerank?

Is it still April Fool's day in some places?

(Or are people seeing this mysterious behavior with static urls all using Firefox)

[edited by: steveb at 5:52 am (utc) on April 21, 2005]

[steveb]

"www.google.com///////////about.html"

It is a ten because those slashes don't matter.