PR Based on more than Links?

Forum Moderators: Robert Charlton & goodroi

Message Too Old, No Replies

PR Based on more than Links?

A controlled experiment yields interesting results

Murdoch

3:48 pm on Oct 4, 2005 (gmt 0)

Okay, so there are two pages on my site that have no outside incoming links and are linked on my site only from their previous page (same page). Page names are based on cities and are listed in alphabetical order.

Page 1 starts with an H (meaning it is listed first) and has a PR of 0.

Page 2 starts with an L and has a PageRank of 4!

The reason I mentioned where they are listed on the page is I wanted to rule out that G only went to part of my page and left before assigning PR. In reality the links are VERY close to each other.

So if PR (and yes this is toolbar PR) is based on link strength what could be the reason?

Note that both pages are the EXACT same age.

The only thing I can think of is that the L city is typically more popular (searches for single keyword city name yields 9,470,000 results over 1,730,000 results).

Comments are appreciated.

tedster

6:28 pm on Oct 4, 2005 (gmt 0)

no outside incoming links

If you are certain about this -- and you know you can't count on Google's link: operator -- then something is very odd. Yes, PR is only about links and the PR of the linking page. Search popularity does not enter into Page Rank. But you said " L city is typically more popular" and that makes me suspect that there may be some inbound link that you are not aware of.

You can and should validate the page, just to ensure that googlebot is not thrown into an error recovery routine and skipping some of the content.

Murdoch

6:41 pm on Oct 4, 2005 (gmt 0)

Actually I used Yahoo's link: tool but I can't say for certain that is 100% correct either (though I'm guessing it would be close enough to make an accurate assessment).

tedster

6:45 pm on Oct 4, 2005 (gmt 0)

If you had, say, a PR-6 or PR-7 link to the PR-4 page, I'd imagine you would see some direct traffic in your server logs from it. Interesting situation -- keep us posted.

Wizard

6:46 pm on Oct 4, 2005 (gmt 0)

I agree with tedster, how can you prove your page has no inbound links? Yahoo and MSN may not know about links Google know, and Google won't say.

There is one more possibility. Theoretically, if the page, which currently shows PR4, used to be 301 or JavaScript or meta refresh redirected to any PR4 URL, it will show PR4 in toolbar until next update, even if its real PR is not so high.

For example, if your site has PR4 on main page, and you redirect wrong URLs to main page, and it happened that this particular page didn't exist but there were link to its URL, and Google crawled it before last toolbar update, it would result in exactly what you observe.

Sharper

2:25 am on Oct 5, 2005 (gmt 0)

Just to sort out the obvious.

Since it's PR 0, is the H page cached in Google's index? It's obviously possible that the L page was picked to be spidered before the H page was...

SimmoAka

8:27 am on Oct 5, 2005 (gmt 0)

Doing a link: in G! without the colon may help to check up more on incoming links - will probably show a great deal more info...

Wizard

9:10 am on Oct 5, 2005 (gmt 0)

Since it's PR 0, is the H page cached in Google's index?

There are plenty pages with PR0 cached in Google, so I don't think that'd be the matter.

Doing a link: in G! without the colon may help to check up more on incoming links - will probably show a great deal more info...

A deal more junk, I'd say. The same you gonna get just writing "mydomain.com" without 'link:' or 'link;' and in any case there's still a risk of missing many links Google knows and refrains to confess!