Your diagrams seem to have been created just to bolster your point.

the pages that page A links out to are not part of the PR equation only inbound links are used.

Originally Posted by nixies78@yahoo.

As you can see each page starts with a value of 1...

We can see that each page starts with a value of 1 only because you have given them a value of 1 LOL! Why have you given each page an initial value of 1?

PageRank is not and has never been a major part of the ranking algo. The anchor text was and still is the major part.

Actually your example of a tracking link could certainly pass linkpop value on depending on the type of redirect used. A 302 will not but a 301 would.

So gaining useful traffic is not part of SEO
Quote:
For SEO purposes these links are useless. They will not increase your ranking at all. The links are useless for anything but traffic.
This is probably the most stupid thing I've heard. SEO is ALL about gaining useful traffic. Far too many "SEOs" go about gaining top rankings for totally useless or "willy waving" words and phrases.
Sales and conversions are the only thing worth mentioning. What is the point of being No1 in the SERPs if nobody actually clicks through to your site?

Well I’m going to ignore all this part it's off topic. The question here is if adding outgoing links will reduce your page rank, nothing to do with the ranking algorithm, just how page rank is calculated.

Quote:
outgoing links do not need to reduce the value all they need to do is send their bonus to different pages, in this case by adding an extra link out of your site you would be sending the value to other website rather than keeping it on your own.
NO NO NO NO NO! The calculation is for incoming links to page A. the outgoing links on page A are not factored into the calculations so they cannot have even the slightest effect on the value.

If this were the case every page on every directory that exists on the Internet would have a PR of 0.
The real value of the PR of any given page is continually changing up and down as new pages are found and added to the index and other pages drop out off the index. But the PR value of any page is NOT reduced by outgoing links, all that happens is the amount bestowed to each of the linked pages is reduced.

I don’t think you have understood the equation, by having a lot of outbound link doesn’t make it impossible to gain PR. The majority of pages in most directory do have a very low PR, the main exception is DMOZ, because it has so many inbound links.

Your diagrams seem to have been created just to bolster your point.

I created the diagrams to illustrate my point, what wrong with that?

the pages that page A links out to are not part of the PR equation only inbound links are used.

Some evidence would be nice, so far all the evidence seem to back up my point and you are just repeating yourself.

We can see that each page starts with a value of 1 only because you have given them a value of 1 LOL! Why have you given each page an initial value of 1?

That is the permise of the equation, you have to start with a value other wise everythign woudl come out as 0, If you time and divide 0 you always get 0. It doesn't need to be 1 it can be 2 or 10 it doesn't matter. 1 is just the logical number.

I don't think you are usuing the equation correctly. The equation you seem to be using is this :

PR(A) = PR(A) + (1-d) + d (PR(T1)/C(T1) + ... + PR(Tn)/C(Tn))

.. because you have added 1 to the incoming pagerank for page A. The equation presented by Brin and Page is PR(A) = (1-d) + d (PR(T1)/C(T1) + ... + PR(Tn)/C(Tn)).

EDIT: You are misinterpreting this equation.

Originally Posted by nixies78@yahoo.

EXAMPLE 2
Now look at the same equation but I will add one outgoing link from Page A. Note how now the value that page A adds to it linking pages is split evenly between the two pages it links to, so rather than 0.85 it gives each page half that 0.425.

The value that A contributes is not half of 0.85, it is half of its pagerank.

Also, example 1 is an artificial situation in that the only page linked to by A links back.

Yep I think I have got part of this equation wrong:

To calculate the PageRank for a page, all of its inbound links are taken into account. These are links from within the site and links from outside the site.

PR(A) = (1-d) + d(PR(t1)/C(t1) + ... + PR(tn)/C(tn))

That's the equation that calculates a page's PageRank. It's the original one that was published when PageRank was being developed, and it is probable that Google uses a variation of it but they aren't telling us what it is. It doesn't matter though, as this equation is good enough.

In the equation 't1 - tn' are pages linking to page A, 'C' is the number of outbound links that a page has and 'd' is a damping factor, usually set to 0.85.

We can think of it in a simpler way:-

a page's PageRank = 0.15 + 0.85 * (a "share" of the PageRank of every page that links to it)

"share" = the linking page's PageRank divided by the number of outbound links on the page.

This is from here,
so the figure are off on my examples.

However this doesn't change the point, structure of your site will determine hte pagerank and the adition of a outgoing link would still reduce the page rank on a certain page.

There is no website distinction - all pages from all websites are just individual pages for the PageRank calculations. We make website distinctions because we are only interested in our own sites, but they don't exist for the PageRank calculations.

But let's take a 3-page website as an example, and let's say that each of the pages links to the other 2 pages - all to all. While that link structure exists, each page will have identical the PageRanks. What happens is that a page's conferable PageRank is divided equally between the pages that are linked to, so each page in the site will receive half of the conferable PageRank from each of the other 2 pages. I don't think that anyone would disagree so far.

Now place an additional link on one of the pages to another site. Suddenly, that page is dividing its conferable PageRank to 3 other pages instead of 2. So it is only passing one third to each of the other 2 internal pages, which means that those other 2 internal pages have less PageRank to confer to the pages within the site.

The net result is that there is less PageRank within the site because of the added link to another site's page. That's what is meant by "leaking" PageRank. It isn't leaked directly from the linking page; it is that there is less PageRank being confered around the cluster of "our" pages (our site).

That's the working of the orginal PageRank equation. Mathematics like that aren't variable.

So the idea that linking out of a site "leaks" PageRank from the site is correct.

I can show it another way without the need to refer to the equation...

B&P said two apparently conflicting things in their original papers:- (1) the sum of all PageRanks is 1, and (2) the average of all PageRanks is 1. Which is true doesn't matter. What matters is that the total amount of PageRank in all the pages in their index is finite. It is either 1, or it is equal to the number of pages in their index. Which it is, is not important. The important thing to note is that it is an exact number, it is a finite number, and the number is not variable.

What it means is that linking from one page to another does not alter the total amount of PageRank in the index. I don't think that anyone disputes the fact the IBLs to a page increase the PageRank, according to the amount of PageRank that the linking pages have. Since the page's PageRank has increased, then for the exact total amount to remain unchanged, one or more pages somewhere must have lost some PageRank. If other page(s) didn't lose PageRank, and the receving page increased its PageRank, then both of B&P's statements must have been wrong.

They weren't wrong, so increasing a page's PageRank necessarily means decreasing another page's PageRank.

All of this concerns the original PageRank equation and statements, of course, since it's all we have.

The idea that this site or that site wouldn't have such a PageRank value if linking out of the site leaks PageRank is incorrect. People usually refer to DMOZ with that particular argument, but a quick look at DMOZ's IBLs, and it becomes obvious where it's PageRank comes from. Forums often have a decent front page PageRank, but it doesn't tend to steadily drop as you move down the levels. They are much more likely to show a 0 toolbar value in their threads - just like this thread does.

The calculations need a lot of iterations, and the idea of arbitrarily allocating a page with 1 comes from B&P's statement that starting each page with 1 results in fewer iterations being needed for the results to converge to a point where more iterations are virtually meaningless. And that's true. Remember that all pages start from scratch in the calculations - they don't start from their existing PageRank values. So starting all pages with 1 makes the calculations shorter and quicker.

Having said all that, I'm not going to get involved in this thread. I thought I'd shed some light on PageRank, in case it helps, but I won't be discussing it.

Sorry - I should have added something concerning the exact total value of PageRank to my last post.

Take an index with an exact amount of PageRank in it that is spread between all the pages, and a specific link structure. Do the PageRank calculations so that each page has a value. Then add a link from one page to another page, and do the calculations again. The page that received the link, and all the pages in the link chains beyond it, will have their PageRanks changed. They will all gain. To maintain the exact amount of PageRank in the index equalling 1 (or equalling the number of pages in the index), the linking page, and those in the other link chains beyond it, will lose PageRank. It is a consequence of the equation, but it is easy to think of it this way, when it's realised what B&P said about the sum and average of all PageRanks.

If anyone doesn't understand it, the best thing to do is sit down with a pencil and paper and actually apply the equation to a 3-page site, starting each page with any number you like, but 1 is best. Easier still, use the calculator that was linked to in this thread, make a few links, select "raw data", and "calculate". You'll see the values change through the iterations. All PageRank calculators apply the equation, so you can use of them that show the raw data.

It's no good making points like, there is no mention of a minus sign, or such a site wouldn't have such a PR if..., etc. Do the calculations with the equation, and see what happens. It's all there in the equation

The idea that outbound links don't leak PageRank because they aren't factored into the equation, and because there is no subtracting part in the equation, is true if the calculation for each page is done only once. But the PageRank for each page is calculated at least 40 times - the calculation is done for the entire index at least 40 times (at least 40 iterations), and that's how the decreases come in.

Thanks for that PhilC, or should I say.... Sergey.

I shall scrutinise your words tomorrow. I hope you stick around.

No if a directory add a link to your website they are giving you a little of their page rank. Yes it does drain the Page rank of the directory and slightly increase yours. This is why directories and website use javascript and redirects to pervent the pagerank getting drained from their site.

What are you basing that on? Why would google take PR away from someone just because they link to another site? What purpose would that serve?

By the way, Dmoz.org is PR9 and they have 5,190,540 outbound links.

Since there seems to be so much disagreement I thought I'd post a link to an article explaining page rank. I can't vouch for the aticle since I just found it with a quick search for 'page rank explained'

It looks like most of the other articles I've seen on page rank and I do think it's one I had read a long time ago.

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

We should all remember though that page rank is not as important as it once was. It's relevant links that are important now. Having links to your site from other relevant sites is going to be more important to being found in search engines that the PR of the site linking to you.

I also believe PR has something to do with Time too. As one of my sites has 0 SEO, as when it was made, no consideration was given to web design, just a personal home page, that isn't very popular, but it's old, so it has a PR 1, even though no real backlinks were attempted. Obviously, Time plays a factor. Not just backlinks... or outgoing links.

What are you basing that on? Why would google take PR away from someone just because they link to another site? What purpose would that serve?

By the way, Dmoz.org is PR9 and they have 5,190,540 outbound links.

st0x we have been over this a couple of times already, it is based on the equation that provided year ago on the pagerank calulation: http://www-db.stanford.edu/~backrub/google.html

Dmoz has already been given as a example:

People usually refer to DMOZ with that particular argument, but a quick look at DMOZ's IBLs, and it becomes obvious where it's PageRank comes from.

The majority of pages in most directory do have a very low PR, the main exception is DMOZ, because it has so many inbound links.

I also believe PR has something to do with Time too. As one of my sites has 0 SEO, as when it was made, no consideration was given to web design, just a personal home page, that isn't very popular, but it's old, so it has a PR 1, even though no real backlinks were attempted. Obviously, Time plays a factor. Not just backlinks... or outgoing links.

Google may have made some changes to include time but I am very specitical. Personal I think that the PR1 you see may either be a residue for a old PR update, or your site may have enought internal pages indexed so that it can get a PR or 1 without any links from extrenal sites.

I can't get to that page at the moment, but "backrub" was the name given to the algorithm while it was still in a lab, im pretty sure huge changes have been made sinse 95.

I still don't see what purpose taking PR away from a site would serve, why would they do that?

Their could well be changes, I'm fairly sure their are a whole bunch of filters applied to which links/pages pass page rank.

But I haven't seen anything to suggest that it isn't still based on the orginal equation.

I still don't see what purpose taking PR away from a site would serve, why would they do that?

It doesn't actually take away PR it just give the bonus elsewhere. Bear in mind that one pages incoming link is another pages outgoing link. Also that the calulatation is done from scratch each time.

I've been thinking

Although the equation quoted may be used as a basis for pagerank calculation, I think that there may be cases when it is not used. Consider the situation where I have a brand new website consisting of only one page (A). I then proceed to accumulate inbound links from directories to get my page spidered. In doing so, I attain some pagerank - all conferred by the directories that are now linking to my page.

Call pagerank for page A : PR(A)

Assuming nothing else changes (I know...) I then add a new page B to my site, and add a reciprocal link from A to B. The pagerank originally conferred by the directories remains unchanged. However, page A now experiences an increase in pagerank - conferred by page B. Let's look at B's pagerank first before considering how much is conferred back to A.

PR(B) = (1-d) + d(PR(A))

which is always >= (1-d) regardless of pagerank conferred by AThis is a key point.

So PR(B) >= (1-d)

On to page A's increase in pagerank, caused by the new link from B:

PR(A) = PR(A) + d(PR(B)) which is > PR(A) + d(1-d)

Woo hoo! I've increased the pagerank of my homepage simply by adding a new page to my site and linking to it.

If I add a page C and link to it (no other links from C), I increase page A's pagerank still more.

PR(A) = PR(A) + d(PR(B) + PR(C)) which is > PR(