|On the Origins of Web Life|
Spontaneous generation of Page Rank?
If another dance is starting, lets while away the boredom and the anticipation with some creative thinking ...
In the web of 2004, a website without Page Rank is pretty much an inanimate site. A PR0 site is destined for the byte bucket, at least on Planet Google.
Page Rank is the essense of traffic, and therefore web life itself.
On a micro scale, we know that a site needs inbound links for it to build Page Rank.
On a macro scale, I ask where is Page Rank created? It it created locally, in some small quantity, each time a new website is created, or does all Page Rank eminate from Google itself, with it's PR10, to be distributed throughout the web.
Mathematicians, please bail me out on this one so I can get some sleep!
There's a very detailed paper called "Pagerank Explained" by a chap called Phil Craven. Even Google puts it #1. An interesting part of which is how you can focus your PR on one or a few pages to make the most of what you create in your site. Its an essential read along with Brett's success on Google alone and Themes Pyramids.
<gilb unsubstatiated one liner>
Who says PageRank is so important now, I thought it was anchor text and themes that did the trick.</gilb unsubstatiated one liner>
"where is Page Rank created?"
A good question, perhaps a better one would be what would the www look like today had we all not been fixated on those green bars? What is that green bar anyway that determines good vs bad www neighborhoods... the first thing looked at even before looking at the sites content... the driving force behind link farms and link swapping schemes. And to think we don't even know where it comes from. I see alot of self deprecating talk here about how webmasters screw up the internet with seo techniques... just food for thought as we head down a parallel path with LSI.
Am I wrong or PR is one of the least important criteria regarding the position of the page in the search results listings?
And is it just about how often and how deep Gb will spider the site?
"A PR0 site is destined for the byte bucket, at least on Planet Google. "
It surely depends on what the competition is.
Agains PR5 there is not too much difficulty in being No.1 against e.g. 1,000,000 with a PR0.
However against a 7 or 8, one has to wait a while until one achieves a PR3.
It may be wishful thinking that Page Rank plays little part in how your page ranks with Google, but all my experience indicates it's at least a significant factor . I doubt if anyone here can quantify just how significant it is, but it's significant enough that I care about it.
Phil Craven's theories suggest that Page Rank is inherited from a site's inbound links but can also be created locally.
In practice, I've tried creating Page Rank locally by adding pages and it doesn't seem to work, but I'm not 100% sure, and that's why I started this thread.
>On a macro scale, I ask where is Page Rank created? It it created locally, in some small quantity, each time a new website is created, or does all Page Rank eminate from Google itself, with it's PR10, to be distributed throughout the web.
It's created locally, in some small quantity, each time a new website is created. Each page starts out with a seed value.
>In practice, I've tried creating Page Rank locally by adding pages and it doesn't seem to work, but I'm not 100% sure, and that's why I started this thread.
My belief is that Google, when dealing with internal links, subtracts the seed value each page starts with. Thus, seed value can pass to other sites, but not your own.
Even if PR has 0 effect on where you place, it's still a good mesure of factors which do count.
That's an interesting concept about seeding, rfgdxm1.
Do you think all the pages on the web are assigned a uniform seed value of PR, or do you think all the websites on the web are assigned a uniform seed value which is divided among the site's pages? Or do you believe that some characteristics of the website or the webpages determine the seed value assigned to them?
|or does all Page Rank eminate from Google itself, with it's PR10, to be distributed throughout the web. |
Mathematicians, please bail me out on this one so I can get some sleep!
What a novel way to put it! Sounds like a job for Joseph Campbell. ;)
In the beginning was darkness, and Google said, "Let there be PageRank" and PageRank filled the internet.
In a certain sense Google is leading webmasters to the Buddhist ideal of Nirvana, "the Extinguishing of the Threefold Fire of Desire, Hostility, and Delusion."
Desire: the anti-spammer tweaks.
Hostility: the demotion (in adwords and occasionally the serps) of sites that are anti-something.
Delusion: The idea that you actually belong in the top five results ;) Y Aaaack!
Google Updates can be likened to the cosmic cycle of the universe. The Aztecs segmented this into the four basic elements of earth, air, fire, and water. Water ends the internet world in a flood (of spam?). Earth ends like an earthquake (Florida shook things up). Fire is destruction by a flame (Brandy is flammable), and air is destruction by wind.
The answer to these questions lie in "The Anatomy of a Large-Scale Hypertextual Web Search Engine" and "The PageRank Citation Ranking: Bringing Order to the Web". Both are by Sergey Brin and Lawrence Page, Google's founders.
Imagine that each page in Google's index generates a little (raw) PageRank, equivalent to 10% of the average PageRank of all pages in Google. This is the rank source.
Then imagine that at each iteration, 90% of the PageRank of a page is distributed among the pages that the page links to. The other 10% becomes the rank source.
After enough iterations, the system approaches equilibrium.
It's inspired, but really not very complicated. Sometimes people think that they can predict and analyse PageRank flow without understanding the Toolbar scale (or at least developing a reliable model). In my experience, that is futile.
I built a site with several hundred pages and pointed each page on the site to home page, and the home page in turn pointed to a set of directories which pointed back to the individual pages.
I also added a links page and pointed to a few typical outside sites. A few typical sites (like PR3) pointed in at my site. Most of these links were relevant, and non-reciprocal.
The page rank (toolbar rank) this site developed reflected that of the sites pointing at it (PR2). All those internal pages pointing at the home page seemed to be of little help in boosting the lowly page rank of the home page.
If every page has a seed rank I would think the home page of this site should have developed a decent page rank, but it didn't. Should I up the bar to several thousand pages?
|If every page has a seed rank I would think the home page of this site should have developed a decent page rank, but it didn't. Should I up the bar to several thousand pages? |
The PageRank you see in the toolbar and the Google directory listings is logarithmically related to the PageRank that is used in the calculations. I don't remember what the logarithm base is, but if it were 10 you'd need ten times as many pages to make a noticeable difference just by creating fresh PR.
This is something that I have been pondering as well. I have a site that was only revelent for a very short period (company changed it's name) the site is very small on maybe ten pages or so. It has a pr of 3 as displayed in the google bar. There are no backlinks to it. I have checked multiple search engines for backlinks and have found none. I have searched for backlinks including any that might be below the low pr cutoff. As far as I can tell that pr of 3 has materialized out of thin air. And it has remained in the google bar now at three for a long time. Don't get me wrong I am not complaining at all. I do not get any search engine results for this site because as I said above the focus and name of the company completely changed and so it is just floating out there. But I find it strange.
> All those internal pages pointing at the home page seemed to be of little help in boosting the lowly page rank of the home page.
PR feedback loops can give up to (very roughly) half a notch of Toolbar PR, assuming that some PR is coming in from one or more links.
> If every page has a seed rank I would think the home page of this site should have developed a decent page rank, but it didn't. Should I up the bar to several thousand pages?
The seed is tiny; I tried to devise a measurement but at the time I couldn't think of a way to get such a large site spidered with no PR from outside. Recently, a method has been suggested...
> [...] the site is very small on maybe ten pages or so. It has a pr of 3 as displayed in the google bar. There are no backlinks to it
I suspect that one of these may apply:
* there's a link that you haven't found
* you haven't waited long enough for the PR to disappear
* there's a duplicate page, Google merged the two and shows the PR from the other one
I can't imagine a ten page site generating PR1 or more.