Forum Moderators: open

Message Too Old, No Replies

PR different for same category in DMOZ and Google

How Come?

         

SkinnyJoe

11:13 am on Apr 8, 2003 (gmt 0)

10+ Year Member



Why is it that the category page "Widgets" in the DMOZ directory has a PR6, while the same category page in Google directory has PR4? Shouldn't they be [almost] the same?

Probably a lot of issues here.

Thanks,
SkinnyJoe

takagi

12:16 pm on Apr 8, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



Hi SkinnyJoe, there are a few things that help explain the difference:

1. Google Directory (directory.google.com) starts with PR10 and DMOZ (www.dmoz.org) starts with PR9.

2. The internal linking is different.
Just as an example; if you go in ODP to top/world/<language1>/<cat_a>/<subcat_b> you can see a link to top/world/<language2>/<cat_a>/<subcat_b>. In Google Directory you don't see these links.

3. Different incoming links on the categories.
Both directories have links directly to sub categories. A link doesn't always go to the homepage of a site. But pages link to a category in ODP and not to the same category in the Google Directory (or to ODP but not to Google Directory).

4. Google Directory links to ODP profiles
On a lot of pages, there is a link to the profile of the editor. The editor's profiles exist only on ODP. From these profiles, there are links into ODP. There are no links from ODP to Google Directory (okay, there is 1 in 'Sites Using ODP Data').

5. Through the profiles some categories get a boost
Some editors, who are allowed to edit high categories, also have a link to a very low category for some personal reason.

Maybe I miss some other reasons, but I think these 5 already explains why the PR isn't always the same.

[edited by: takagi at 12:33 pm (utc) on April 8, 2003]

Dynamoo

12:25 pm on Apr 8, 2003 (gmt 0)

10+ Year Member



(Posts crossed over.. spent too much time typing it!)

The internal linking structure between the Google Directory and DMOZ are different, so PageRank can vary between the two.

Here's a couple of examples.. in dmoz.org, the existence of the editor profile pages changes the structure, the front page is different, the pages could have backlinks from outside the directory itself.

So, to give an example, look at a page with a link from the front page of dmoz.org (PR9/10ish), but not directory.google.com (PR10+), in this case the UK directory, Regional/Europe/United_Kingdom/ - in dmoz.org, the front page PR9 gets passed straight down to that category as a PR8, in Google only a PR7...

..now click on the last editor name in that category (note, this is a random choice). You'll notice that particular page has a PR7. Click on the SECOND link in the category list and you'll see that the resulting page has a PR6, which is actually one higher that the parent category.

dmoz.org has been around for nearly five years, so it will also have more backlinks to it than directory.google.com. According to Alltheweb, dmoz.org has 716,519 backlinks, and directory.google.com 283,824. Where those are deeplinks, this will add PR to the lower levels of the dmoz.org directory.

I guess I have been spending waaaaay too much time looking at PR in dmoz.org :)

Rhadamanthus

7:30 pm on Apr 10, 2003 (gmt 0)

10+ Year Member



Does it even really matter? If I were Google, and I knew that one of these sites was essentially a mirror of the other, I'd only crawl one (probably the original - DMOZ). Otherwise, you're getting double the influence for any link in there.

So my question is, does Google actually spider the Google directory AND the ODP, or just one? If just one, which one? If it's only spidering the ODP, then this thread is largely irrelevant, because the PR in the Google directory won't affect anything.

Mohamed_E

7:47 pm on Apr 10, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



> So my question is, does Google actually spider the Google directory AND the ODP, or just one?

Since both sets of pages have PR I assume that both are spidered.

I am pretty sure that the Google directory is spidered, as it is not excluded in Google's rather long robots.txt.

The ODP must be spidered, since new sites added to it while the RDF dump was not working were found by Google.

SkinnyJoe

8:46 pm on Apr 10, 2003 (gmt 0)

10+ Year Member



I was thinking that the different PRs for the same category could be an indication that a category is growing or shrinking rapidly in importance, inasmuch as PR is a measure of "importance."

If PR measures importance, it seems problematic that pages with such obvious similarity would have greatly different importance.

The PR discrepency could be an Undervalued/Overvalued indicator, where searchers are oversupplied or undersupplied with content.

Mohamed_E

9:06 pm on Apr 10, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



Google's definition of "importance" is based 100% on links, not on content. Thus a obscure DMOZ clone can have a very low PR for the entire cloned directory, while Google has a PR of 10 for the same, equally cloned, directory.

<added>Content is read by humans, who choose to link, or not link, to the page. People decide how "important" or otherwise a page is, Google just observes their linking behavior. Of course, reciprocal linking plays havoc with the fundamental assumption :). </added>

steveb

1:53 am on Apr 11, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



Look at the front page of dmoz and the front page of the Google directory. Different pages are linked off the first page. For example with Dmoz:
Sports
Baseball, Soccer, Basketball...
With Google:
Basketball, Football, Soccer...

Dmoz will give higher pagerank to baseball sites, while Google will diminish baseball but value football sites.

Dynamoo

1:20 pm on Apr 11, 2003 (gmt 0)

10+ Year Member



It's an academic point I guess, but the way PageRank goes through these highly-structured directories is actually a good way to learn about how PageRank flows.

dmoz.org and directory.google.com are ideal because of their high PageRank.