Forum Moderators: open
For example, you have 10 links on a page, and 2 of them point to your home page. Would it pass (Passed PR) / 9 or would it pass ((passed PR) / 10) * 2?
Right now it is just an academic question for me. But I imagine that if applied properly, you could use it to help shape the PR distribution among your pages.
Another question related to PR and passing it on:
If I have a site with a main page of PR 5 (let's assume that it is a one page site). Then I build 6 additional pages and link them all together. Would my main site possibly go down to a 4 while raising the others?
onebaldguy - my suspicion (and it is only a suspicion) is that they would not mess with the PR algorithm when it comes to template factors, they are more likely to apply a template filter to other factors in their ranking, such as anchor text.
Those navigation bar links are actually some of your most important links on your page. They should pass on PR. But their anchor text is not as likely to be as revealing as anchor text from the content part of a page.
Passing on PR does not reduce the PR of a page. In fact, if those pages link back to your home page, you will get a very small boost in PR from the intrinsic value of those 6 pages, divided byt their outgoing link count.
I don't know about linking to 2 different URLs that have been merged due to duplication or redirection.
What makes you say that the two links are counted as one?
This is very intresting.
I recently had to add an external link to the home page of a site and did not want to use a trick, to make the link pass on no PR. - So I decided to add 6 links to the same page in my site so the external link only receives 1/7th of the homepage PR.
By what your saying it's going half to the external and half to the internal!
Are you sure about this ciml - what about different anchor text, will this make them count as multiple links!
Yes.
Dino, my view is based on experimentation. The PageRank transfered is exactly the same as if the links from one page to the same URL are merged; the anchor text boost looks to be at least very similar.
I haven't tried with different anchor text, though.
I was guessing that someone actually tried it.
Did you use the exact same URL both times? What I mean is, does "http://widgets.com/widget.html" "/widget.html" and "../widget.html" count as the same URL when linked from a file in the second level directory?
I have noticed that Google does seem to follow complete URLs more readily than pathnames, that makes sense as it helps them find new servers. But it is possible that it is handled differently while passing PR too.
But I can tell you that during a crawl, they are more likely to follow full URLs. The first month, Google only picked up my home page. While my home page was still gray bar, that second month, it only crawled the pages that included the host name. Once I got PR assigned it was much more willing to follow path names.
I tried this on one other site and it behaved in a similar fashion. I have not tried it with a site that only does pathnames, to see if it will crawl those if there are no complete URLs to crawl.
Google's crawler is one of the best out there - they take into account the location and response of the server among other things.
It just wouldn't make sense for them to treat them different (at least I can't think of a reason).
Since they have more than one crawler working at a time sending them out like little soldiers making decisions based on PR, location, lag - and other things - the same site might get spidered differently at different times.
I always thought the site with 90 links would get 90% of the available Pr but this is not the case they would get half each!
WW where you learn everyday :)
A grey bar site is only going to get a limited amount of attention till the PR is calculated. But it is in Google's interest to find any other new servers that this sitem might link to. All external sites will have a complete URL, so they just toss those in the queue.
I'm not claiming it as fact, it is just what I have observed with the 2 sites that I have put up so far. It could be coincidence.
Like Dino, I thought that the page with 90 links out of 100 would get 90% of the available PR. It seems to match the 'random surfer' idea better, but it's not what happens to me.
I can only guess that someone inside Google felt that it would be best to merge outbound links from one page to another.
