I dunno, but I've been using Owlcrafts via.php script to block this leakage for over six months. There was a good few months *before* I came accross this wherein we linked directly to close to 200 sites in our directory. I was really excited in those days, and used Zeus (if anybody knows of it), to crawl the net and create this themed directory for link building purposes. The idea was to e-mail them and tell them we've added their link and can you please do the same.. and all that newby stuff.. lol.. Later the vast majority wouldn't reply because our link pages didn't have PR. Our main site internals linked similarly had PR5 in those days. So I said what the heck and used the php script and set the robot.txt to dissalow. I did this with these sites only. All others who I would trade links with from time to time were linked to directly without script. Sure enough, in the next update two months later, our link pages went to PR5s and PR6s. I know we were doing some link building in those days, but I have no way of explaining the PR0s while the links were outbound and clean, whereas many internals of our domain linked similarly were showing a reasonable PR4, and some PR5. This leads me to believe that linking out to sites who you have no reciprocle arrangement with in a manner that their links are one third or more compared to your own internal links will harm your PR. Of course the occasional link or two to quality resources is never a bad idea, and should be done directly. Also a site giving five site-wides to others will have a harder time reaching a good PR level compared to one which does not link out this agressively. Isn't this much clear?
PR leak doesnt exist. How does any links/resources page have any PR? They almost never have any links pointing at them apart from maybe a sitemap link and can have hundreds of links pointing away from them. And yet my links pages have PR5 exactly the same as my index page. Bah. Humbug.
Yes. And Rodgers and Hammerstein and anyone else who wants to can create as many hypotherical examples and demonstrations as they wish. For the third time, Michael, I am saying I really don't care about any of that: Show me a site that has supposedly suffered PR leakage and we'll talk.
Show me a way to record/measure a small changes in RAW PR on demand and you could see for yourself. The absence of such a mechanism (at least outside of Google labs) precludes the evidence that you ask for. Fortunately because we are discussing a formal procedure (the calculation of PR) it is possible to build a model and observe the effects of PR flow in that way. If you believe that the Rogers the Hagstrøm models are flawed in some way just say so. Your observation is based on the false assumption that Toolbar PR can be substituted as measure for RAW PR. No one knows (outside of the Google labs) how the eleven (0 -10) toolbar icons are related to RAW PR or even if they are related at all. For example sometimes you will see a Toolbar PR for a page that has not even been indexed. - Michael
So, if I understand you correctly, any PR leakage that might occur would be too small to measure, i.e., for all intents and purposes invisible? Or alternatively that no one but Google would know how to measure it, however small or large it might be? Then what exactly are we talking about? Some hypothetical effect that cannot be seen or measured? What's the point of that?
No I am not saying it too small to measure, I am saying there is nothing to measure it with. No not a hypothetical effect but a logical consequence of the calculations as defined by Brin and Page. Academic interest only. As I said earlier "...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
Michael sounds like he is talking about Schroedingers Cat. 'PR leakage does occur its just that when you try to measure it with anything you destroy the validity of the experiment'.
LOL That is funny mnemtsas, made all the more so because I was a Physics major, thank you for introducing some levity into what is a rather dry subject! To be clear though RAW PR is not like Schrödinger's cat or even Heisenberg's uncertainty principle RAW PR is simply a variable that is calculated by substituting numbers into a formula in the way that Brin and Page described. The actual value of this variable will depend on the data set that the original numbers are extracted from. Anyone can perform the Brin and Page calculations on any suitable data set and those that use the same data set will produce identical answers. However we do not have access to Google's data set and therefore cannot use it to calculate RAW PR. So we can never know the value of the RAW PR of any of the pages in the Google data set and consequently we cannot perform experiments on this particular data set and observe the corresponding changes in RAW PR. What we can do though is choose another data set over which we have complete control and use exactly the same calculation that Brin and Page have described. Because we have control of the data set we can then make changes to it and observe the effects of these changes. This is exactly what Rogers has done on paper and Hagstrøm has done on a spreadsheet and the results are there for everyone to see. - Michael
Hi Minstrel, Umm I don't if this is a PR linkage or not, would you be able to tell me why this is hapens... http://www.hhdirect.co.uk/ All ibls are pointed at index page (internal pages and other websites). It has PR2, some of internal pages have PR3. Thanks
You could because there would be no PR transfer and I am sure a lot of people will. It adds yet another method to the long list of ways for link partners to cheat - Michael
It's not PR leakage. There have been a number of threads here and on other forums about this -- it's one of the outcomes of the latest Google update. What it means isn't clear, except (1) the datacenters are still not all synchronized, (2) this was an unusual update which may be still being tweaked, and (3) I don't believe any publicly reported PR values any more.
As I said I don't disagree with their explanation of the calculation mechanism. But the data sets they are using are artifical and biased to show as large a PR dilution as possible. That type of structure just doesn't exist in the real web world and several people have already reported pages that shouldn't have any PR based on this principle, however they still do have substantial PR. My biggest point is to try and counter the riduclous obsession some people have with this entire subject and the heroic lengths they go to to try and avoid "leakage".
Of course the data sets are artificial and as a result we can deduce nothing at all about the magnitude of the effect. It may be small or microscopic in the real Google universe but its existence is the point at issue here. The PR you are referring to here is presumably toolbar PR? It is not possible to observe the flow of RAW PR by observing the toolbar PR. Not least because we have no way of knowing the relationship between the two but also because toolbar PR is a crude and unreliable indicator in itself. I agree but the key to this is education and imo denying its existence does nothing to further understanding of this difficult concept. If you really want "to try and counter the ridiculous obsession" then maybe confirming that it exists but explaining that the effect is probably negligible and why it should be ignored would be a more effective counter, rather than declaring "Some people will never believe that PR Leakage doesn't exist" as you did earlier. - Michael
Michael, Does "go fuck yourself" have any resonance with you? Show me where I ever said "Some people will never believe that PR Leakage doesn't exist".
Well said, compar. Michael, why are you continuing to waste bandwidth with something that cannot be measured or seen, has no practical significance or relevance, and can not be shown to exist except in a hypothetical data set in a hypothetical universe? Give it a rest.