Page Rank Calculation

everyone: Check out my new article Page Rank Calculation based on Bob Wakfer's PageRank Calculation Chart where I address the question "how many links do I need to get a page rank of 5 or 6?" and make 4 simple rules.

Rule 1: 18 PRx links to get PRx (i.e. 18 PR5 links to get a PR5).
Rule 2: 3.339 PRx links to get PR(x-1) (i.e. 3.339 PR6 links to get a PR5).
Rule 3: 101 PRx links to get PR(x+1) (i.e. 101 PR5 links to get a PR6).

As Walfer's calculations are based on links having 50 Total Outbound Links (TOL) there were some simple rules need to revalue links.

Rule 4a: Revaluing link to PRx: 6.25(PRx/TOL) PRx (i.e. PR8 130TOL: 0.385 PR8)
Rule 4b: Revaluing link to PR(x-1): 34.375(PRx/TOL) PR(x-1) (i.e. PR8 130TOL: 2.1 PR7)
Rule 4c: Revaluing link to PR(x-2): 187.5(PRx/TOL) PR(x-2) (i.e. PR8 130TOL: 11.5 PR6)
Rule 4d: Revaluing link to PR(x-3): 1037.5(PRx/TOL) PR(x-3) (i.e. PR8 130TOL: 63 PR5)

The article is a pretty easy read and explains how I came up with the rules.

 

Very interesting articles, thank you.

They have also explained something I have been worried for the last weeks since the last BL update. Also have gotten no answers in the two forums I visit.

That is that Google does not show all backlink information:

Quote from Bob Wakfer:

"The one trick to this link search is that Google does not display all backlinks. At one time it was thought that they only listed pages with a value of PR4 or greater. Today however, you will find backlinks reported from pages of lower PR values. So, at best, Google's backlink search seems to present some sample of pages linking to the site. Suffice it to say that this search is not a reliable measure of all IBLs to a page."

Thanks for clarifying this!

 

bobmutch, your logo is very cool

 

hurricane sh: hehehe thanks, made it myself if I may say so
etreus: Just in case I didn't explain myself in the post very well Bob Wakfer's PageRank Calculation Chart article is written by Bob Wakfer's and the Page Rank Calculation article is written by me bobmutch.
On the inbound links end of things go to Yahoo and do a "domainlink:www.example.com" and that will give you your inbound links. If you want to you can do a "domainlink:www.example.com -inurl:www.example.com" to exclude your internal links.

 

Bob who?

Bob you better get your posts corrected, you have Bob's last name mis-spelled in your posts, wouldn't want "The Donald" mad at you, would you Bob?

You could hear the words "YOUR FIRED" real soon man.

 

anthonycea: Well if I spell the name wrong he will have something to flame me on and not be so hard on the article

 

Don't count on it, the Compar Mafia will get you one way or another, once your in, there is only one way out

 

Anthony- youre out of control.. lol

 

Oi Bobmutch,

You really should get Bobs name correct - at your site anyway. It's just a common courtesy thing, when you're featuring their works.

You've only misspelt it once, but hey - I'm pedantic !!

Cheers for the article

 

john loch: anthonycea pointed that out about 6 hours ago and I quickly corrected the spelling. I just double checked this thread and my article page and as far as I can see the spelling is corrected. You may want to try doing a refresh if the spelling of Bob Wakfer is still incorrect.
"If I approached you to reproduce your work at one of my sites (which I may btw)." Really, well I would be honoured. Email me at and lets work something out.
I would also be open to suggestions for articles. The way I get some of my ideas for articles is by answering questions on the different forums I am subscribed to and then publishing an article on that subject.
Check out Scrub Warehouse for a bit of fun over at SEO Guy Forum.

 

I edited my post above..

Cheers, JL.

 

Bob we are only hard on you because we love you, remember youa famialy nowa, there is only one way out of the Compar Mafia.

Concrete boots, you sleepa wit du fisha, youa knowa wata meena bon?

Be cool G........

 

Anthony, I think you are on more than Jack Daniels tonight.

 

Sorry Boss, I won't give out no more family secrets, just breaking in the new kid, all business from here on in boss

 

Thanks for the article. It seems pretty straight forward. Based on your formula the PR for two of my sites should be different. I have listed their information below. Any ideas on why site #1 is so low and site #2 is so high?

Site #1 Current PR = 1 [resalebroker.com]
Actual BackLink Counts:
PR 0 26
PR 1 3
PR 2 4
PR 3 2
PR 4 3
PR 5 5

Site #2 Current PR = 5 [ezliving.com]
Actual BackLink Counts
PR 0 2
PR 1 1
PR 2 3
PR 3 2
PR 4 1
PR 5 3

Thanks!

 

ResaleBroker: Well thats kind of interesting but with 5 - PR5 links and your site only at PR1 I would have to ask when the you got the links? It has been 66 days since we have had a PR update but we have had a backlink update during that time if I remember correctly. So it would be my guess that the backlinks are new so Google shows them, but your new PR will not show on the toolbar until the next PR update.

 

How did I know you were going to say that?

What about the second sight though? By my calculations it shouldn't have a PR5.

 

If only Pagerank were this simple. Unfortunately, it isn't.

The link structure of the entire site of the page you are attempting to estimate the PR of has too much bearing on the PR to be able to guestimate the page's PR based purely on information about the incoming links.

Take a thorough read of this article and study the example and you will see understand why the strucuture of your site has such a huge impact on the PR of any page in the site:

http://www.iprcom.com/papers/pagerank/index.html

 

Thanks for posting that article!

 

I have learned a couple of things since I wrote the original PageRank Calculation article.

But before I tell you about those things I'd asked that you read the original article carefully. I clearly point out in the article that both internal and external links are used by Google in calculating PR. So Virginia Realtors' comment about it not being that simple is wrong. It is that simple.

It is a straight forward mathematical calculation based on all the PR values of all the pages that a page recieves links from. Nothing more and nothing less. If you think about this statement carefully you will realize it even covers the case of "PR leak".

Now what addition information I have learned.

To build a chart like this you are forced to make some assumptions. The major assumption built into my chart is that all the pages have an average of 50 outbound links. I was never very happy with that assumption, but it was based on the fact that Google recommends not having more than 100 links per page and so fifty seemed like a half way cut off point.

Since the chart was published Rustybrick has given us a new and very useful tool for analyzing the linking structure of pages. http://www.rustybrick.com/link_analysis.php This tool will tell you the number of backlinks a page has from pages of each PR value.

I took the results from this tools and plugged them into my chart. I found the correlation was good but not perfect. I started to adjust the "50 links per page assumption" and I found that when I changed this assumption to 40 links per page I got a 100% correlation with the actual results Rustybrick was reporting -- at least for the approximately 10 different pages that I test with.

So I am going to revise my article to reflect this. In the meantime if you would like a copy of the original Excel spreadsheet with the assumption changed to 40 you can download it here. And if you would like a copy modified so that you can enter Rustybricks results and then play "what ifs" by looking at the impact of adding addition links you can download that chart here.

Remember also, as has already been said, that it is very hard to do tests or forecasts on new pages or recent links because Google has not done a PR update in a long time.