6 replies to this topic

[rmjvol]

It had been a long held rule that the Google link command only showed backlinks from pages that were PR4 & higher.

In this thread from August 2003, Debra brings up the point that she's seen the Google link command returning links that were below PR4. Jill & others question & then confirm her observations. Qwerty resists the analysis but finally hops on board.

While I don't have the benefit of looking at exactly what you guys saw then, this issue brought up a debate for me at another board earlier today. I found the most direct info here at your forum so I thought I'd share some research. In the big scheme of things, this issue is not that important but it's stuck in my head now

Disclaimer: We all know that the TB PR is an approximation of real PR & that the TB PR can be flaky at times. Anytime I say PR, I mean TB PR unless noted otherwise.

I checked the backlinks on highrankings.com.
http://www.google.co...c...nkings.com/

That's a lot of backlinks so I just poked around the first 100. Checked every non-indented page & a few indented ones. About half down I came across a link from a gray TB page (shows a cache):
www.whatsworking.biz/full_story.asp?ArtID=13

Couple down I see this with gray TB & no cache:
www.webpronews.com/ wpn-3-20031203ShouldYouPayForInclusion.html

Next a nice auto generated looking page with PR1, shows cache:
www.aallix.com/finding.php?page=marketing+consulting

Next was a gray TB, shows cache.
www.cheverus.org/LibraryLinks.asp?cat=20

Finally was a gray TB, shows cache page.
www.voelspriet2.nl/links2/cgi-bin/r.../rate.cgi?ID=69


That's all that I originally found in the 1st 100 backlinks. My initial reaction after finding a couple was that the old PR4 & higher rule has indeed changed. Then I thought I saw a pattern. All dynamic urls... except 1. So is there a pattern?

Let's look at them 1 at a time.
www.whatsworking.biz/full_story.asp?ArtID=1
The home page has TB PR5. site: shows 182 pages indexed but all the articles show gray TB. Until I refresh a few times. Then I see a nice green PR4 on the TB for that page. Next...


www.webpronews.com/ wpn-3-20031203ShouldYouPayForInclusion.html
No cache, huh? If you watch closely, you'll see the url indexed by G gets redirected to www.webpronews.com/ebusiness/sitepr...rInclusion.html. Hmmm. Follow the site's navigation from index & you'll see you gotta go thru the archives to the 2003 archives to the summary of the article then thru a redirect to get to the real article. Try link:www.webpronews.com/wpn-3-20031203ShouldYouPayForInclusion.html & you'll see that G'bot actually sees a redirect page as a link to the page in question. With all the redirection going on, I don't find it hard to believe that the link: command & the TB could be out of whack to our eyes. I'm gonna chalk this one up to too much redirection to get a good read on the real PR of that link. Next...


www.aallix.com/finding.php?page=marketing+consulting
Looks like a dynamic, auto generated page on a site that's basically trying to catch a bunch of keyword traffic. The home page www.aallix.com/ has a PR6. With only 30 links off that page, I'd expect all the next level pages to have PR6, PR5 or PR4. But that's not what the TB shows. Only 1 internal link goes to a static looking url. www.aallix.com/top_keywords.php That page shows a PR5. All the other internal links have a dynamic variable in the url & they all show PR1. I don't buy that. I think the TB is wiggin out on those urls. Next...


www.cheverus.org/LibraryLinks.asp?cat=20
The only link to this page seems to be from www.cheverus.org/library.asp which shows PR4. None of the "Internet Resources" pages under the Library section show PR. The page in question doesn't link to highrankings.com. It links to rankwrite.com which redirects to highrankings.com. We can debate whether LibraryLinks.asp?cat=20 should really be a PR4 page but we don't know for sure. If it's not a PR4, it's probably close. I think the TB is getting dizzy between the dynamic url & the redirect. Next...

www.voelspriet2.nl/links2/cgi-bin/r.../rate.cgi?ID=69
I can't read the site but I found the listing pointing at rankwrite.com on this PR6 page
www.voelspriet2.nl/links2/pages/Men..._en_Discussies/
Google's displaying the cgi redirect page on the site that gets shot to rankwrite & then to highrankings.com. Again, I think this is a case of the TB not being sync'd up with what the index shows in the link: command. I think that the link is giving credit from that PR6 page or at least passing enough PR thru the cgi to count as PR4 or better.


I also looked briefly at the joeant link mentioned. While there's no backlink there anymore, there is a redirect going on so I'm betting that's what was messing up the TB PR when you guys saw it originally.

There's a total of 9 urls in those 1st 100 links that contain a ? character. I see that there's a couple forum posts in that link list that I should look at but it's too late & I'm beat.

I know this isn't complete & full data but I think it's more than was originally raised in this thread that started the "rules have changed" comments. Just consider it food for thought.

My current working theory is that the TB PR gets messed up on either dynamic urls, redirects or some combination of them both. I think that the old rule of the link command only showing backlinks with PR4 or higher is still valid. You just need to know that it's *Real* PR4, not necessarily TB PR4.

Anybody want to chime in on what I've pointed out so far? Anybody want to check the other 500 backlinks?

Ross

Edited by Scottie, 06 February 2004 - 09:24 AM.

[powerofeyes]

Hello Ross, Welcome to HR forum
Regarding PR4 links as backlinks yes that theory still exists for static URLs, 99% of the static URLs I have seen showing backlinks will be PR4,
Now coming to dynamic URLs, especially PHP and ASP urls have a tendency to display a toolbar PR of 0 of 1 or 2 but still show as backlinks, According to me, google has a different version of the PageRank when it comes to dynamic URLs not the one which shows on the toolbar,
As you said the toolbar PR might not be assigned to display the exact pagerank of dynamic urls or they might get messed while assigning PR to a dynamic version of the Pages, So it is no a problem,
The real issue is between the dynamic page and the static page, So conclusion it is safe to have static type URLs whe it comes to google and for its backlinks,

[cbp]

In the past, I always thought that the PR threshold was a very high 3 and not a 4.

Now, Google only returns what a Google employee called a "sample" of backlinks - the criteria for being in the sample is now more than just a PR threshold and involves other factors which I do not think that anyone has worked out yet.

[OldWelshGuy]

Or maybe they have worked them out, but are just not telling lol

I agree that things are a changing with regard this, but then again, if google are giving more weight to a link from a page local to your search, all this 'carved in stone' stuff goes out the window.

It makes SEO even more fun

[Scottie]

Welcome to the forum, Ross!

Are you really in Neyland Stadium West? I marched in Neyland in the flag corp back in my college days... I was in town back in the fall and was shocked to see how much the Strip had changed... time flies!

I've edited your live links per our guidelines.

[Grumpus]

First things first. If the page is in Google's index and it's also showing as being "linked to" some site and it's grey, it can mean only one thing - your toolbar is burping. If you go back to these "grey" pages another day, it'll show some PR there. I promise. It may be "guessed" PR, but it'll still show something. (The only other reason for a grey toolbar is if the page was picked up by the fresh crawl, but the PR database didn't get updated yet. In this case, that page wouldn't be shown as "linking to" another site, so...)

Now, there's the more interesting point that you've stumbled upon. Dynamic pages are, in fact, calculated a bit differently than static pages. It can take as long as 4-6 months for "real" PR to be calculated for a set of dynamic pages. In the interim, a "guessed" PR value is used and is calculated in the same way that "guessed" PR is calculated for static pages that Google has found via a link from its own site, but it hasn't found any links from other sites. Guessed PR is almost always calculated by taking the PR of the root site page, subtracting 1 per "/" or "&" in the URL and adding 1 for the "?" if it exists.

There's also the fact that it can take Google a while to figure out what the variables in the URL actually do for a dynamic page. In essence, it looks at page.asp?var=1 as the same page as page.asp?var=2 until it can figure out just what the "var" variable controls. I don't really have the time to write a novel and explain all this, but suffice it to say that if you can take advantage of this knowledge, there are some serious SEO benefits. Take a look through my various posts here dealing with pages with common pointers showing different data being pointed to on each page for some more clues. (My third post in this thread will be a good starting point, maybe).

In essence, if there are enough common elements on a dynamic page without it tripping over into the duplicate content range, then I suspect there is eventually becomes a "real PR" value for each page (taking each variable into consideration) but also a blanket PR value that gives "page.asp" an overall value, regardless of the variables attached to it. Link to any one of them, and the hidden "overall power" PR goes up. It's probably not weighted the same way, and may really only come into play with TSPR type calculations, but there is some evidence that that value exists - we can just never see nor confirm that value. How it all works will always be a bit of a mystery, I guess.

G.

[mcanerin]

Welcome, Ross!

Ian