Forum Moderators: Robert Charlton & goodroi
But I might be close to a solution, which makes checking real PR possible, moreover, not just in 0-10 scale, but distinguishing between PR5.0 and PR 5.9.
I'm aware it might be wrong way, as sometimes I get unreliable results, but in most cases it seems to work. Maybe someone of you will find a way to improve it.
The idea is the following: let's do a kind of search, where the order of the results will be not determined by any factors but PageRank. For example, "inurl:www.site1.tld OR inurl:site2.tld OR inurl:site3.tld". It is not unlikely, that the results will appear sorted by PageRank.
So take a few sites to check, and a few reference sites, and make a query that returns them all sorted by PageRank. If a checked site will be placed between two PR5 reference sites, it may mean it has also PR5.
What reference sites could be? There are sites with very stable PR, for example, www.microsoft.com can be good reference of PR10 site, just as www.google.com, and www.google.co.uk is quite good reference of PR9. I know a few sites that have PageRanks 3, 4 and 5 constantly for a year, so they are pretty good references for these PR values.
The problem is, that even in most cases my query returns pages sorted by PR, sometimes it doesn't. So I'm not sure about the reliability of this method. But maybe you'll be able to improve it.
[edited by: ciml at 11:13 am (utc) on June 8, 2005]
[edit reason] Examplified [/edit]
But that said, it could explain why sometimes you see sites with a apparent lower PR sandwiched between others (higher pr sites).... perhaps on that occassion the actual PR of the said site is higher!
just some thoughts
Interesting find.
I tried this with various sites in my sector and im not sure if its just the results i got but every site was in order of age!. (older sites first)
Some clear low PR sites ranking above others and the other way around.
You might be onto something here but i think its a starting factor based on age prior to any algo filters being applied.
It is a problem to find good reference sites indeed, but I know established informational sites in my niche with stable tPR, I mean their tPR didn't change during many updates, so it's better than nothing.
Ordered by age? That's interesting, I have to check it out! Maybe this method cannot tell much about PR, but can predict when a site will be freed from the sandbox? Anyway, it would be a valuable information.
But what if you take old site with very small PR and young one with very big?
I genuinely think that older sites no matter what content start from a better ranking position. I have seen loads of sites over say 4 yrs old with very poor content using black hat seo and ranking very high in google whilst a new site using these techniques would be picked up very quickly and sandboxed.
Its almost like the algo favours older sites and is more relaxed on the content issues than newer sites.
Also, i dont think you will find a new site with a high PR - i would go so far as to say that if a site carries a PR8 for example it will have had to have been at least a few years old on the net. Im convinced that site age is a factor in PR despite content and links
wizard: GoogleGuy said PageRank is not changing constantly, and toolbar is updated periodically, so I thing we can take it as a fact.
Do you have a link to this?
It's fact that PR is not updating constantly when it comes to toolbar and directory.... but clearly to calculate pagerank a page ranking measure is used over a period of time in order to produce the visible pagrank. Perhaps the termininology I'm using clashes, but the fact is sites do rank differently constantly - google have always said the ranking is based on many things, hence the search engine placements for results. I know this is 2 seperate types of ranking (pagerank and search engine placement).
Anyway, I do see the idea behind this and agree there may be something to it... but i personally do not see much use for it. I just look at the rank placement order in the google directory.... if 5 PR6 sites appear, then the one at the top is a higher PR6 than the one at the bottom.
But at the end of the day what value can this offer?
Actually I believe he said the exact opposite:
We have a bank of machines that computes PageRank continuously
message 5 [webmasterworld.com]
--
Anyhow what is the point of all this effort to calculate allegedly "real" PageRank?
If I add more links, the link pop goes up, and my net PageRank increases. Right now I have PR 4 pages outranking PR 6 pages.
It's not how much you have, it's what you do with it that matters. In the 30 minutes, or hour or however long that you spend doing this, think of how many link requests you could have made, or better, how much content you could have added to your site that would have resulted in other sites linking to you on their own.
wizard: GoogleGuy said PageRank is not changing constantly, and toolbar is updated periodically, so I thing we can take it as a fact.
Do you have a link to this?
[webmasterworld.com...]
Read the last paragraph in post #5.
This was what I referred to, and maybe I expressed my thoughts wrong way, saying "constantly" I meant "all the time" or "continuously".
Wizard: GoogleGuy said PageRank is not changing constantly
Wizard - I'm not sure how you are misunderstanding this, as it seems very clear, but I'll rephrase it:
PageRank, as visible on the Google Toolbar is only updated every 3-4 months, and represents a "snapshot in time" of PageRank at the moment it was calculated and updated on the Toolbar.
Real PageRank, however, it being updated continuously.
There is no way to accurately calculate this, unless you can hack in to Google's PR caculation network, but the usefulness of this activity still escapes me.
But he also said that they were going to update the PR again over the next week or few weeks during this summer.
Usefullness of checking real PR, not only the value updated on last toolbar, is that we check a true value, not outdated one.
And the method described in this thread is an attempt to do this, which can be proven wrong but I thought worth checking.
