Forum Moderators: open
As far as I can see, if you exchange a link with somebody, there are three possible scenarios:
1.) Both links, mine and the one I get from the other web site, have exactly the same PR value. More theory, but even if this would be the case, both web sites would loose PR due to the damping factor, right? Not very attractive.
2.) You give more PR than you get. Well..
3.) You get more PR than you give. Nice - but you probably wont find that many webmasters willing to exchange a link if they LOOSE some PR, do you?
I of course see the other benefits of exchanging links - but in order to raise your PR, how might exchanging links help? That is, unless you find some webmasters not interested in SEO.. ;-)
Can someone tell me what I am missing?
outgoing links decrease your PR as long as your linking page is not a dead end (this is another, different case) and connected to the rest of the pages of your site. Of course, this is a kind of second order effect since the transferred PR from your link page to the rest of your pages is lowered because of the additional link (i.e. transferred PR changes: d*PR/N -> d*PR/(N+1) for each of your internal pages) and this decrease the PR of your link page. Try some simply analytical or numerical calculation.
By the way, the main reason for introducing the damping factor are leaf nodes. These leaf nodes lead to degenerate eigenvalues, thus there is no more a unique solution. By changing d from 1 to 0<d<1 you can easily show that there is a unique solution for the resulting system of linear equations. Also, maple leafs would led to PR=0 (for d=1) for all pages outsites such structures. Or simply: setting d=1 would just work if you could reach each page from each other page (of course, not directly).
Made In Sheffield
PR is deceased:
for page (A) it changes from 1 to 14/23 (which is lower)
for page (B) it changes from 1 to 11/23 (which is also lower)
PR is deceased:
for page (A) it changes from 1 to 14/23 (which is lower)
for page (B) it changes from 1 to 11/23 (which is also lower) <
Yes the PR of the site is reduced as I said, but an individual page's PR is not reduced by having a link on it.
the PR of page A - which is an individual page with a link on it - as well as the other page (B) on that site as well as the PR of the whole site PR(A)+PR(B) is reduced (by adding a link to page C on the second site).
Even if PR leakage is real it won't be much, and you'll get it back as soon as one person "from the wild" links to your site because they find it useful.
Don't try to manipulate google using link exchanges and such artificial link popularity stuff.
You want long term results and long term results are accomplished by creating great content and THAT content will get people to link to you on their own will!
From my own PR 8 site I often write a front-page main attraction that includes a link to content I thought was great and that I'd like visitors to see! You see? I didn't link because I wanted to give them PR, I linked because I thought they had great content to offer and as a consequence they got a bit of PR from me.
That's how it's supposed to work. I don't believe in artificially inflating yourself. I've seen people get banned for this. If a Google hand-review determines your links are solely for self-inflation of PR you are in trouble!
> PR(A) = 0.25 + 0.75 PR(B)
PR(B) = 0.25 + 0.375 PR(A)
PR(C) = 0.25 + 0.75 PR(D) + 0.375 PR(A)
PR(D) = 0.25 + 0.75 PR(C) <
The PR for the entire test group is 1, hence the starting values of 0.25. Page B is the only page in site A/B to lose PR after the link from A to C because Page A's PR is now distributed between 2 pages instead of 1.
So no, Page A does not lose any PR through having an extra link. So no, a page does not lose PR by having links, but it doesn't gain as much PR the more links there are on a page linking to it.
Cheers,
Nigel
you have to compare the situation for the 4 pages before adding a link from page (A) to page (C)
PR(A)=PR(B)=PR(C)=PR(D)=1
with the situation after adding a link
PR(A),PR(B)<1
Obviously, page A and page B loose PR.
Thus the PR for each page on that site decrease. This is independent from the value of d (0<d<1), but the amount of decrease depent on d.
The situation is the same for more complicated structures. Assuming there are no dead end, the average PR in such a system is one and the total amount of PR is N (the number of pages). Adding a link leave the total amount of PR unchanged since the number of pages doesn't change. Thus an increase of PR of some pages always lead to a decrease of other pages.
Also I have to say that I'm not worried about PR leakage. Of course, I have external links on my site and I'm not only focussing on PR. I'm linking not for PR, but for content to other sites. And I know that external links can improve the ranking (not PR). However, I just want to show that outgoing links decrease PR (assuming that the original PR algorithm is still valid) - nothing more, nothing less.
I have successful sites with thousands of outgoing links, but yet, under 50 incoming links. If my PR was diminished with each outgoing link, then by this theory, why don't I have a minus 8,000 PageRank? To the contrary, I have pages with hundreds of outbound links with PR4 - PR6 which out-rank every single competing site. PR is a portion of Google technology, whereas content is is another important consideration.
Search engines like informative sites with quality outgoing links. If you have a quality site, just "sticky" me and I will add your link. You wouldn't believe how many times I gain a new client looking for something outside of my offerings, and even better yet, looking for my competitors.
I see your point. A loses PR because it doesn't get as much from it's link from B because B get's less from A. (Algebra revision was needed :)). The point I was trying to make is that adding a link to a page does not reduce it's PR directly. However because the amount of PR kept within the site decreases A does lose a little PR (which I had missed).
Cheers,
Nigel
Because PR isn't so simple - you have to calculate PR according to the corresponding formula. First of all it is impossible to get a PR < (1-d). Secondly, when you simply subtract the number of outgoing links from the number of incoming links you are oversimplifying. You have to look at the transferred PR (from other pages to your page and vice versa). PR also depends on a number of other parameters as the structure of your site, number of pages, where you place your outgoing links and ...
> To the contrary, I have pages with hundreds of outbound links with PR4 - PR6 which out-rank every single competing site. PR is a portion of Google technology, whereas content is is another important consideration.
I have never said that PR is the only point to look at. Of course, there are other important factors as anchor text or title. We agree in that point, but that wasn't the topic of this discussion. I have mentioned twice that outgoing links - even if they decease PR - can improve ranking (which is the important thing).
I think somewhat of a language barrier was broken here. I was stressing a point. When I said:
> I have successful sites with thousands of outgoing links, but yet, under 50 incoming links. If my PR was diminished with each outgoing link, then by this theory, why don't I have a minus 8,000 PageRank?
doc_z, this is what we call a rhetorical (yet sarcastic) question. It was not begging for an answer, as it provides its own.
I don't understand whether you felt that my reply was pointed toward something you posted, but the point was missed by a mile.
Mine is the perspective of the "Master", and not the "Student". I was stressing multiple points in my post, and not poking at anybody.
I think that both points may generally be taken well. PR has a strong place, as does content. However, to confuse or commingle them was not a part of either point.
I have simply never seen any instance of loss of PR due to outgoing links, which was a point of confusion as it seems. Diminution of PR gain for the recipient of a link is true. Whereas, a loss for the granting page is not an instance I have ever experienced, nor is it a theory for which I have seen any credential.
indeed, there was a language barrier on my site. It seems to me that I have to improve my English ...
However, my mathematics works well.
The decrease of PR follows directly from the original papers. A simple intuitive explanation (a proof in a mathematical way can be found elsewhere) is given in msg #37 (The situation ...).
The reason that you normally don't see a drop of PR is simply a result of the ToolbarPR. This just shows an interger of a log scale and not the exact real PR. Thus a decrease in PR by adding a link leads in most of the cases not to a decease in ToolbarPR.
Hmmm.. ofcourse I'll keep my effort going, in getting oneway inbound links anyhow ;-)
This makes sense, given the enormity of a three billion URL index.
<added>
Good point, McMohan. The most effective SERP inflation efforts I've seen in the last few months involve a lot of links from outside the domain, but not necessarily high PageRank.
