First of all there really isn't such a thing as "leakage". You don't give any PR away when you link to a page. The target page gets a value that is some proportion of your PR. But you don't give PR away. The more OBLs on a page the less PR value shared with each of them. See my PageRank & How to Get It article for a complete explanation. And can you explain how linking your "link page" to only your home page is going to help anything?
Professional SEOs who has interpreted Page and Sergei's PR Documentation, they wrote articles about it, like you did. It's us to believe, choose and interpret which one is logical or not. So I am not going to discuss what is right or wrong. Whether you believe or not this is my opinion. And please don't be mean
I didn't mean to be "mean". I just believe that you have misunderstood what you apparently have read. You just have it all wrong, and before you lead someone else to believe this false information I felt I should offer a more conventional view and understanding of the subject. So please answer my question and explain the value of linking a link page only to your home page????
Sure Compar, will be pleasure. If you don't mind I want to use your own words first. "Normally our initial concern is to transfer as much PR as possible to our home page. The way that this is done is to make sure that there is a Google readable link on every page pointing back to the home page." I took these sentences from your article. OK our aim is the same. I want to maximize Homepage PR. So what we need is all internal pages needs a link pointed at homepage. is that ok? right. So let's say we have A (homepage) B(links Page) C(products page) to maximize Pr for homepage, linking structure should be like A<->B A<->C. Am I correct? ok let's go on. As your theory there is no leakage... So if we build our internal link structure as I explained above what would be the PRs of these pages? I am not sure why this is happens but; If I took as a reference of O'Reilly's Google Hacks Book and Professional SEOs who claims that this theory is logical, A should have (approx. values) 1.5, B should have 0.7 and C should have 0.7 . So in this situation what happened to B's and C's 0.3 PRs. Honestly, can you tell me what's this called, cause I can't find out why this is happens? Right, let me answer your question now... Let's say you have 20 obls at Links Page. What would be the value of transfered PR to these obls if you add more link to this page? Would it be diluted? Answer is right, yep? So if you place a link pointing to only Home page as internal link, more PR value would be transfered to Home page unless you add link pointing at for all internal pages. If you add link for internal pages at Links Page then Pr value would be shared by the other internal pages as well. So you would not maximize your Home Page Pr value...
But if you link to all internal pages you increase the value of all pages and then they pass their increased value to the home page. So I really don't think there is any difference.
So, is this it: Only the home page links to the links page and the links page has a few forward links to the rest of the website? Question: What other areas of the website would be suitable for a forward link from the links page? Thanks
I see a few diagrams, a lot of suppositions, and conclusions that might be of interest if it weren't for the false premise. Page leakage doesn't exist. If you want to dispute that, don't show me an article claiming it does. Show me page leakage in action, unequivocally. Anywhere.
Well Minstrel I'm glad to see that I will have company in taking the heat on this subject. I don't know why this obsession about PR leakage is so prevalent or where it originated, but the absolute obsession with it and the great ends that people go to apparently avoid it are simply ridiculous. Thanks for your support, but I fear we are fighting a loosing battle. Some people will never believe that PR Leakage doesn't exist. They have no evidence. They have no proof. But they will fight to the death anyone who says it doesn't exist.
What was the false premise that you saw? I think it is for two reasons. Firstly because it is the wrong name for the concept and much of the discussion that I see boils down to semantic differences. If it was called "loss of potential PR" which is what it is there would be less obsession and more understanding perhaps? Secondly because it is a difficult concept to grasp. The idea that the average RAW PR for all pages in Google's universe is exactly one and the fact that for every PR increase there is a corresponding PR decrease somewhere else in Google's universe is not easy to conceptualize for most people. The fact that PR can increase on one page and decrease on another within the same domain is just a logical consequence of the calculation of PageRank. Linking off a page deducts nothing at all from the page and those that think it does are misunderstanding the PR calculation. Equally those that believe replacing an outbound link with an internal link will have no effect on the PageRank of the linked to page are also misunderstanding the PR calculation. Fortunately from the practical SEO point of view it makes no difference whatsoever and simply selecting outbound links on the basis that they will help users is by far the best strategy. - Michael
Well, first, that PR leakage exists. The article assumes that it does and then goes to great lengths to show us how it would work if it did, none of which is of the slightest interest to me because the premise is incorrect. No, I don't think so. It would still be just as incorrect. Call it "loss of potential sanity" or "Fred" -- it's still wrong. Correct. "Voting" for a web page (by linking to it) does not weaken the page doing the voting any more than voting for a mayoral candidate weakens the power of my vote for a school board candidate or a national leader. Or any more than saying "good morning" to you weakens my ability to say "good morning" to someone else.
It is like superstitious behavior or religion: People search desperately for a glimpse into the Google inner sanctum in the hope of gaining a slight edge over their competitors. If you conceptulize PR as a bank account, it would make sense: Pass some of it over to someone else and you will have less in your account. Great theory. Except the premise is false. Now we're on to the newest "Google insight" which claims that "reciprocal linking is bad" -- just as groundless and just as hard to dislodge.
I don't disagree with his conclusions about the mechanism of PR calculation. But the point he doesn't make is that if the outbound link had been reciprocal there would have no net difference to either site. The increase would have balanced out the dilution as I prefer to call it. Now in real life many of our links are reciprocal, or if not reciprocal then OBLs are frequently balanced by IBLs. So the entire thing is in general balance. I personally think that anyone who uses any method to mask his/her OBLs is dishonest and is cheating everyone s/he links to or who links back to him/her. Then of course there is the question of the value of PR in today's algorithms. So the entire subject tends to be obsession about nothing.
Actually Hagstrøm does not argue from a premise at all he simple states a coherent set of propositions which explain a class of phenomena. As he himself says "I hope this will give a more visual idea of PageRank - and at the same time answer time-old questions - such as does PageRank leak exist? and which linking structure is better?" He answers the first question at the end with "That depends on how you define PageRank leak" which does not indicate he is arguing from a premise but rather the opposite. If you read through his article from the beginning here. I would be interested to hear from you which particular elements of his explanation you disagree with and why. Voting is often used as an analogy for the PageRank calculation but I am afraid that it is slightly more complex than you presuppose. For a very good explanation of PageRank using the voting analogy I suggest you read Ian Rogers fine article here. If you follow it through to end you will see a paragraph that reads "If you give outbound links to other sites then your site's average PR will decrease (you're not keeping your vote "in house" as it were)". This is not a premise or even a hypothesis it is a logical consequence of the PageRank calculation as expounded by Brin and Page in their original paper. Anyone is of course at liberty to disagree with Rogers and Hagstrøm (and many others) who explain the logical consequences of the PageRank calculation. However I have yet to come across anyone is able to point to a sentence in their respective articles and demonstrate an inaccuracy. Maybe you will be the first Minstrel or perhaps you compar? - Michael
No, it's an assumption or interpretation. There are numerous articles and posts "explaining" how this so-called leakage works. As I said above, if you're trying to explain something that doesn't exist, it makes no difference how eloquent your explanation or how many times you say it or how many people say it: You can't create something out of nothing. As I also said, show me an example of this and I might begin to be persuaded. I've already read all the opinions on the myth that I need. Don't show me another article -- show me an example.
If you are saying, as you seem to be, that Rogers and Hagstrøm have got it wrong simply because "they are wrong" then that rather precludes a debate. If you say that they have got it wrong because you can demonstrate where in their logic they have made an error then a rational discourse may follow. It's up to you and the ball is in your court. Both Rogers and Hagstrøm give clear examples and Hagstrøm even has a downloadable spreadsheet which demonstrates PR flow quite nicely. - Michael