Backlinks from same page

<a href="xx.htm#fff">

[Stefan]

I just realized that one of my pages shows a backlink for itself because of a <a href="xx.htm#fff">(continued below)</a> going to a <a name="fff">fff</a> on the same page. The page is PR5 so it shows up in the backlinks on www.site.org/xx.htm as www.site.org/xx.htm.

Now.... the page is picking up a PR5 backlink.... if I stuck a few more of those name anchors on the page <a name="foo></a> would I get more backlinks or would it stick at the one listed from the first? And also, should I have a name anchor like that used on every page to get an extra backlink for all the PR4-PR5 pages?

I'm wondering if I've stumbled across a perpetual PR machine.

[DerekH]

Stefan wrote

I'm wondering if I've stumbled across a perpetual PR machine.

<chuckle>

If you had, I think you'd find your PR would increase and increase till you had the only PR100 site in Google...

If you've the skill and patience to read the PageRank Uncovered paper and then make an Excel spreadsheet to model the behaviour that you describe, you'll see that all you're doing is funnelling the same amount of PR in different ways. The more backlinks you put to your own page, the less each one contributes to the total.

The Google algorithm as was first explained, is inherently stable, and PR can neither be created or destroyed when PR is moved about. All you are doing is changing the amount that is kept on page and the amount that is exported to other pages. You can't keep more PR on the page than you own, so you certainly can't own more than have. If that makes any sense at all!

DerekH

[pegaweb]

I would think Google more than likely just ignores links that point to page they're on. :)

[doc_z]

If you had, I think you'd find your PR would increase and increase till you had the only PR100 site in Google...

No, that's not correct. Including 'self links' would lead to a well-defined mathematical consistent model (for the case that only one link to the same page is considered as well as for the case that all links are taken into account). This would only change the distribution of PR (but would 'produce' any additional PR). Also, my experience is that only different links are taken into account for PR calculation, i.e. adding more than one link to the same page has no influence.

I noticed a few weeks ago that links from the page itself are shown as backlinks. However, this doesn't mean necessarily that these links are taken into account for PR calculation (even if this would be possible).

[Stark]

all fairly interesting from a pagerank perspective, but do they count as valid anchor text for the page?

i.e. can I have a page about widgets called widgets.html and include within it <a href=widgets.html>Widgets</a> and get an effective boost on that keyword?

Surely not?

[plasma]

a link is a link is a link

[Stefan]

Derek and doc_z, good points. I thought it seemed too good to be true. On the other hand, plasma's, "a link is a link is a link", is very appealing.

I might fool around with it on a minor page and see what happens... who knows, maybe in a few months times I'll be inviting people here to check out my PR100 website... :-)