Page Rank Bleed Check out my article on Page Rank Bleed. This article examines if page rank bleed is a factor you need to calculate when you are considering what page rank will be tranfered by your inbound links. This article used the number in Wakfer's Google PageRank & How to Get It Chart. Enjoy!
Actually, I think you meant "Page Rank Bleed", which is a myth. At the risk of seeming to be blunt, the spelling error repeated 4 times (including the thread title) in your post as well as in your page title doesn't fill me with confidence or inspire me to believe that you have a "must-read' article.
Minstrel: Thank you for noting this, it has been changed. I would be interesting in hearing your view on the subject of Page Rank Bleed and article support from authoritative sites that you base your opinions on.
That this subject repeatedly surfaces is astounding. Yes, PageRank bleed exists; no, a page with outbound links does not "bleed" PR. The "bleed" is of the site as a whole (or, more correctly, by some or all of the individual pages of the site), on the assumption that every page in a site has at least one link to some other page of that site. (A page of a site that has no links to any other page of that site--or to any site also owned by or of interest to its maker--is strikingly rare, but such a page would not generate any "bleed".) The "bleed" is simple enough to understand: a page's "strength"--its ability to contribute to another page's PR by linking to that other page--is shared equally among all pages it links out to. The more pages it links out to that are not part of its own site (or a related or co-owned site), the less that goes out to the links that are to the same or a co-owned site. This is kinderspiel. To simplify: if a page of my site has only one outbound link, and that link is to the front page of its site, it is sending 100% of its available "strength" (I am trying to avoid saying "PR" here, because that's not exactly how it goes) to that site's front page. If you now add a second link, which is to some external independent site, the "strength" that that page is sending to its site's front page is halved. That is the "bleed" that external linking causes. Naturally, on a real site, with the various cross- and inter-links among the pages, the calculation is by no means simple, but the principle remains the same simple, obvious one: what you spend outside doesn't stay at home.
Yes, whatever "strength" or "PR" or whatever you want to call it a page has is "passed on" to other pages in proportion to how many pages it is linking to. However, if you spend any time browsing forums/fora you will see that the term "page bleed" is not usually interpreted this way, but rather to mean that as a page links out to other pages it loses some of it's PR, thereby "weakening" the original page: This is the part that's a myth.
Yes, it is. One that, like the stubborn old donkey I am, I keep posting and posting, here and there, trying to dispel. It's like trying to scoop up the ocean in a sieve.
Owlcroft/minstrel: So the position you take is that when a site home page from Site1 links to the home page of Site2 that there is PR transfered from site1 to site2 but no PR lost or bleed from site1. I think that most people see transfer and bleed as having 2 different means. Transfer meaning that PR that is give from site1 to site2. Bleed to most seems to mean the amount of PR that is "bleed" or lost from site1 to site2. Am I understanding what you are saying correctly? Now if this is true, what about phil craven's and ian roger's articles where they show in their examples fully meshed set of pages with all have PR1. And then when they do an external link out of that fully meshed set of pages there is a "loss" of PR. Are you saying that there is no lose, just a transfer of PR and the orginal PR retained?
Yes. I am aware of the theoretical arguments about "PR bleed". I've just never seen any evidence whatsoever that it exists in reality. Period. On the other hand, I'm not going to debate the issue with you. If you choose to believe it exists, then by all means hoard your links -- I don't argue with Jehovah's Witnesses when they come to my door either...
minstrel: Ok so you would see the term bleed to refer to the idea that some PR is lost when an link is given? And you would use the term transfered where a link transfers PR to another link but that PR is not bleed or lost? Can you think of any good articles that support that view? I am sure you are aware of both Phil Craven's and Ian Roger's articles that appear to take the position that PR is bleed.
If you have read those articles, I don't know how I can make it any plainer. It will help all of us if we stop using terms like PR and transfer. PR is the result of a many-stage iterative calculation that takes into account all links on the internet (at least all that are on pages Google has indexed). A link from Page A to Page B does not "transfer PR", which is why I tried to avoid such terminology. It does register a vote for Page B that will be used in calculating Page B's eventual PR, and which vote has a value proportionate to Page A's eventual PR and to the total number of outbound links Page A has. The value, or weighting factor, is not "PR" as such--it is a vote whose value is, as I say, proportionate to Page A's PR. If Page A "votes for" X pages, the value of its "voting strength" is doled out evenly among those X many. In a real election, as held in the U.S., a citizen can only vote for one candidate for a given office. But we can imagine a scenario in which a citizen could vote for any number of the candidates--voting, say, for three candidates, with each then receiving a third of a vote. The citizen has not "transferred" anything to the candidates, and is thus not out anything because she voted. But if she votes for one candidate, that candidate gets all the value of her vote. (In some elections, a voter can vote for X many out of Y total candidates for multiple positions--for example, there might be 6 open seats on the county board and 13 candidates for those 6. A citizen can vote for up to 6--or can vote for only one. If the top six vote-getters are to be elected, such a "bullet vote" strengthens the odds that the selected candidate will outpoll the others.) Obviously, then, every time a page "votes for" a page outside its own site, the strength of its "vote" for pages within its site is proportionately diminished. It's all math. There are no maybes or guesses.
Actually, to expand on your analogy, years ago Nevil Shute wrote (and I read) a novel about a future society in Australia (not sure of the title any moe - possibly In the Wet? or A Town Called Alice?) -- everyone had a basic 1 vote unless they were in prison or something, but some people had moe than 1 vote, I think up to 6 (it was many years ago that I read the novel), with people being awarded extra votes for things like education and contributions to society. In that novel, voting for a candidate would be similar to your analogy, since votes from some people/pages in effect would carry more "weight" than votes from other people/pages... I know it's not a perfect analogy but heck -- I haven't even thought about that story in probably 20 years...
Owelcroft: Thank you for you time and the points you made. I read Craven's and Rogers' articles again, and they talk about PR being increased, decreased, channeled, being lost, injected, drained, aquired, distributed and leaked but not transfered. I don't think any of the things you mentioned were thinging I have not read before but I would agree that voted is a better way to say it. I looked at my orginal questions to you again and if I exchange vote for transfered that conveys what I wanted to ask and I think you have answered that question. Craven's and Rogers' articles seem to support that Page Rank can be "lost" from a site by outbound links (meaning that when the calculation is done at PR updates you will have less) and "injected" into a site by outbound links. It is all math I agree, and while we have the orginal formula they are many things that we don't know. minstrel: I think your analogy is a go one cause there are things like this that Google may be doing that we don't know.
minstrel: I would agree with that. There are a number of different opinions and conjectures on PR and just you say, there are just that. But Craven's, Rogers' and Sobek's articles that are held highly in the SEO community all hold that page rank can be lost (all three articles use the work "lost"). I am wondering what authorive articles you would cite that have lead you to take the position that the deduction that PR can be "lost" is as you say a myth.
1. "held highly" by whom? 2. I said before you should not confuse opinion or conjecture with fact; perhaps I should add, "do not confuse reputation with anything but reputation" See my earlier post -- I have no interest in debating this with you. Believe what you like. As I have already stated, if not here then elsewhere, I've seen no evidence of this hypothetical loss in page rank. And I have no inention of modifying any of my linking practices because of a dubious theory. If you believe it, then don't link to anyone. If you don't link to anyone, good luck with your web site.
My article, which you quote in the opening thread, supports this view. PR bleed and link hoarding are ridiculous. They are at best a theoretical construct that has virtually no effect in reality. It is only anal retentive and very selfish web site owners who worry about this. Linking should be freely participated in and a shared exchange among sites. Anyone who masks links, or refuses to share legitimates links, on the basis of PR bleed is a joke.
ministrel: I am not wanting to debate it either. But I have no problem with a free exchange of ideas. I just asked if you know of articles from authorive sites that support your view that page rank bleed is a myth. I you know of any paste them up, I would like to read them. Held high by whom. Well in my travels around SEO sites I see they articles referenced often. I don't have a list of name, but it seem that most people seem to know who they are. Personally I don't take a position on the bleed issue. I have lots of links on my site. I don't hoard links. compar: I would tend to agree, I think link hoarding is a selfish pratice and I chuckle when I see it. Your article may support that view but I didn't notice that it addresses it directly. I would be interested in any articles you know of that deal directly with this issue.