Sunday, January 6, 2008

Number of Pages - Will an Increase of Pages Give Your Site Higher PR?

Number of Pages - Will an Increase of Pages Give Your Site Higher PR?

As you add pages to your site, the RPR of the home page and the subpages will in most cases drop. While this sould not be an issue with most sites, the myth that adding pages to your site will increase the PR of the site pages needs to be disspelled.

For an example if you have 5 page site no matter if it is Fully Meshed, Hierarchically, or a combination of the two you will get the same average real PR. When you add a new 6th page that page will start with a real PR of 1. (This page would link to one of the pages in the site and have a link to it from one of the pages.)

If the average of the other pages is above 1, in all cases the average real PR will drop. When the average real PR drops for a group of pages that means that some if not all pages will end up with less real PR.

The following are a few diagrams and calculations to verify that this claim is true. Most small sites have a linking structure that is fully meshed. I have in this article put forth that adding more pages to your menu will dilute the real PR of the current pages. While some in the industry are willing to accept that, they seem to be saying that the real PR is dropping in this case because of poor linking structure.

The following examples use the two main linking structures, Fully Meshed, Hierarchically and combinations of the two. The following examples and calculations show that adding pages to your home page with 2 way links will dilute all the other pages real PR that have two way links to the home page.

In the following 4 samples you will see a small increase in the home page real PR but in the subpages you will see a drop of up to a factor of 10.

3 Page Site & The Real PageRank After 40 iterations

The first example is a 3 page site with page "x" having an inbound link to the home page "a" and the home page "a" with 2 way links to 2 subpages "b" and "c". Page "x" has a real PR of 1000 at the start of the first iteration and pages "a", "b" and "c" start out with 0. After 40 iterations the home page "a" has a real PR value of 3064 and the subpages have a real PR value of 1302.

See links that shows iterations and the source.

4 Page Site & The Real PageRank After 40 iterations

The next example is a 4 page site with page "x" having an inbound link to the home page "a" and the home page "a" with 2 way links to 3 subpages "b", "c" and "d". Page "x" has a real PR of 1000 at the start of the first iteration and pages "a", "b", "c" and "d" start out with 0. After 40 iterations the home page "a" has a real PR value of 3064 and the subpages have a real PR value of 869.

See links that shows iterations and the source.

5 Page Site & The Real PageRank After 40 iterations

The next example is a 5 page site with page "x" having an inbound link to the home page "a" and the home page "a" with 2 way links to 4 subpages "b", "c", "d" and "e". Page "x" has a real PR of 1000 at the start of the first iteration and pages "a", "b", "c", "d" and "e" start out with 0. After 40 iterations the home page "a" has a real PR value of 3065 and the subpages have a real PR value of 651.

See links that shows iterations and the source.

21 Page Site & The Real PageRank After 40 iterations

The next example is a 21 page site with page "x" having an inbound link to the home page "a" and the home page "a" with 2 way links to 20 subpages "b" to "u". Page "x" has a real PR of 1000 at the start of the first iteration and pages "a" to "u" start out with 0. After 40 iterations the home page "a" has a real PR value of 3072 and the subpages have a real PR value of 130.

See links that shows iterations and the source.

4 Page Site Fully Meshed & The Real PageRank After 40 iterations

This example is a 4 page fully meshed site with page "x" having an inbound link to the home page "a". Page "x" has a real PR of 1000 at the start of the first iteration and pages "a", "b", "c" and "d" starting out with 0. After 40 iterations the home page "a" has a real PR value of 1914 and the subpages have a real PR value of 1252.

See links that shows iternations and the source.

5 Page Site Fully Meshed & The Real PageRank After 40 iterations

This example is a 5 page fully meshed site with page "x" having an inbound link
to the home page "a". Page "x" has a real PR of 1000 at the start of the first iteration and pages "a", "b", "c", "d" and "e" starting out with 0. After 40 iterations the home page "a" has a real PR value of 1695 and the subpages have a real PR value of 994.

See links that shows iternations and the source.

6 Page Site Fully Meshed & The Real PageRank After 40 iterations

This example is a 6 page fully meshed site with page "x" having an inbound link
to the home page "a". Page "x" has a real PR of 1000 at the start of the first iteration and pages "a", "b", "c", "d", "e" and "f" starting out with 0. After 40 iterations the home page "a" has a real PR value of 1550 and the subpages have a real PR value of 824.

See links that shows iternations and the source.

Adding Pages To SubPage By Two Way Links

Here is 4 more diagrams and with a bit of code to show the iterations and a link to the code that produces those iterations so you make I am doing my calculations right. At this point I am now looking to see if I can find any configuration that will let you add 2 Way or Meshed Linked pages with out a loss of real PR to the home page and/or subpages. (I changed the code to work out the iterations for this group to PHP.)

4 Page Site Before Adding 2 Way Links to Subpage

See links that shows iternations and the source.

4 Page Site With 1 - 2 Way Link On Subpage

See links that shows iternations and the source.

4 Page Site With 2 - 2 Way Links On Subpage

See links that shows iternations and the source.

4 Page Site With 3 - 2 Way Links On Subpage

See links that shows iternations and the source.

What Do These Diagrams Show Us?

Well I think it is pretty clear when you add pages via a two way link to the home page of a website that the real PR of the other subpage will drop.

lso when you fully mesh a page into a site that has a fully meshed link structure the real PR of all pages goes now.

Check My Numbers On Phil Cravens PR Calculator<

If you would like to check out my calculations you can drop over to Phil Cravens site and take his Google RageRank Calculator for a spin.

At first when I ran two ways links to the home page on Cravens RP Calculator I didn't believe the numbers when the home page when up slowly with each 2 way link that was added (the other subs dropped for course). So I set out and did up the formulas my self and get the very same number.

Try Finding A Linking Structure That Doesn't Bleed PR

When you try out differnet linking structures make sure you have the home page an inbound PR of around 1000. The calculations of sites where they are only dealing with there own real PR of 1 don't produce the same numbers. You should also turn the Mode Selector to Real Mode that that the calculation will not pick up one way links or orphans. The Simple Mode will include them into the calculations but as Google drops them you don't want to be looking at calculations that won't happen.

Again when you add pages with at least a 2 way link (one way links are orphans) to a site that has a high PR home page it really doesn't matter where you place them you are aways going to decrease the real PR of some of the pages on the site. Even if you select one page of the sub pages and hang 2 way links of it. It will still decrease the real PR on the home page.

The only combination that I have been able to find that doesn't decrease the real PR on your home page is when you add 2 way links directly to the home page. But when you do this it will decrease the PR of the other subpages. Even when you add a 2 way link to a subpage that has a two way link to the home page the home page real PR drops.

Have fun with the calculator and hopefully this clears up this issue seeing how there is nowmath to support what I have been saying.

By the way my finding are not new. While I orginially found out how adding pages can drop TRPR and hence Real PR by experience, PR experts and people involved in the text link ad industry have been aware of this for a long time. This is no new discovery and I consider the things I have been presenting in this thread be ABC's in PR.

I am not saying that to take a wack at those that have disagree with me but it's just a fact. Any one that has even taken the time to read the main articles that are out there and has done there home work should see this. Even the site you just asked me to look at shows this same thing and comments on it.

What Other PR Experts Have Said

Mark Sobek's article, which I consider the most advanced articles I have named in this post (see links at the bottom of this post) , even has a section called "The Reduction of PageRank by Additional Pages". While I don't agree with all the positions he takes in his article, he does come out and clearly state "Since adding pages to a site often reduces PageRank for already existing pages, it becomes obvious that the PageRank algorithm tends to privilege smaller web sites." http://www.miswebdesign.com/resourc...pagerank-6.html

Phil Craven's article is quite short but still its very clear that he has a good grasp on PageRank. In his "Google PageRank Explained", he also brings out that adding pages to a site one or more of the existing pages will suffer a PageRank loss.

"The new page will, of course, aquire PageRank from the site's existing pages. The effect is that, whilst the total PageRank in the site is increased, one or more of the existing pages will suffer a PageRank loss due to the new page making gains."

Ian Rogers in his article "The Google Pagerank Algorithm and How It Works" shows how you can return link 1000 pages from a one page and then link that page to the home page and product a real PR of 331 on the home page in Example 13 as seen at this link http://www.iprcom.com/papers/pagerank/. The problems with this sample are numerous. First its spam (which is notes), 2nd the Google bot is going to cough at a page with 1000 links on it, and 3rd, in the scheme of things real PR of 331 is nothing. One inbound link from a PR4 will give you that. Baring that issue Ian's article is one of the better article out there.

Chris Riding's makes the statement in his article that "Fundamentally, however, adding extra internal links to your outbound pages is one of the most (if not the most) important internal factors to improve you PageRank." Page 27 http://www.voelspriet2.nl/PageRank.pdf . This is a good point and gets me to add another item to the "Things that can cause yoru TBPR to drop" list.

Things that can cause your TB PR to drop
9. Removing of internal links from pages that have external links. (Internal links on pages that have your most external links will change the about of PR voted off site.)

0 comments:

 
Copyright 2009 Make Money Blogging Online. Powered by Blogger Blogger Templates designed by Deluxe Templates