Forum Moderators: open

Message Too Old, No Replies

PR Numbers and Specfics?

quoted their page rank as 6.7 and 5.4.

         

webdev

9:09 am on Feb 20, 2003 (gmt 0)

10+ Year Member



I have recently done a link exchange with a fellow site and as part of the initial email I was sent the webmaster quoted their page rank as 6.7 and 5.4.

How can anyone find this out, I'd like to know so I know how close I am to the next level of PR.

I did email the chap and ask them how they knew these exact figures but they never replied which has led me to believe they are lying about the exact numbers.

Any one else had any experience with this.

Thanks

ukgimp

9:14 am on Feb 20, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



Cant find it now but I read that you could go to the dmoz and save the gif that represents the TBPR and then measure it and you could work out a relative scale. How accurate that is I dont know

webdev

9:16 am on Feb 20, 2003 (gmt 0)

10+ Year Member



Unfortunately I'm not in Dmoz so I can't do that..I read that post too a while ago.

cwebb

2:41 pm on Feb 20, 2003 (gmt 0)

10+ Year Member



You can't download the gif as it's simply a green gif scaled to a certain length, so you have to look in the html code, but you can only decide if you're high or low in a PR range if you have twi different results, you can't tell to a decimal point!

Chris_R

2:51 pm on Feb 20, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



[searchnerd.com...]

I doubt they are lying. People that misunderstand PR outnumber the dishonest people 10 to 1.

You can come up with a pretty good idea - if you are in a category with lots of other people.

Still this is all relative - and you can't ecpect someone out of the blue using a perhaps different method to understand what you mean by 6.7 and 5.4.

aspdesigner

7:22 am on Feb 21, 2003 (gmt 0)

10+ Year Member



Unless they work for Google, I can't see how they could find this out. Even the internal data sent to the Toolbar expresses it as an integer.

Google Directory values can narrow this down slightly ("high" 5 vs. a "low" 5), but certainly nothing that would approach accuracy within .1

The very fact that they are stating it like "6.7" would lead me to believe that they don't know what they are talking about. Remember that the Toolbar PR is not even your actual PageRank, it is only a rough approximation of PageRank values displayed as a 0-10 scale. Actual PageRank increases exponentially for each Toolbar step, so what they meant by "6.7" would not even be clear.

Ask them if the ".7" part of this value was determined linearly or exponentially, and what is the Toolbar exponential factor they are using that allows them to calculate this so precisely. Then watch the confused blank look on their faces. ;)

While Chris_R may be right and they may just be very confused, given that they refused to answer your question, I think they were just trying to B.S. you.

daroz

7:25 am on Feb 21, 2003 (gmt 0)

10+ Year Member



I believe the datastream for the toolbar gives a decimal number in X.X form.

(I could be wrong, but I did check one or two links in other forums that listed an example. Check the Google Toolbar forum here for PR info.)

Chris_R

9:25 am on Feb 21, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



The datastream gives it as a whole number - not a decimal.

aspdesigner

9:43 am on Feb 21, 2003 (gmt 0)

10+ Year Member



Last time I checked, it was still an integer.

When I first got the Toolbar, I packet sniffed the data myself to check for this possibility. Back then it was XML, and was an integer value.

Since then, it has changed to the new toolbarqueries approach, but it is still an integer. Where you may be getting confused is the new approach contains several #'s with colons between them ":", but last I looked the returned toolbar PR is still an integer value.

Given that this is a calculated value who's only purpose is to support the toolbar's 0-10 display, this makes sense. It would be a waste bandwidth to send extra unused digits, when you stop to think that Google has to transmit this data for every page that every person who has the toolbar visits.

If you are aware of a way to retrieve this with greater precision (or better yet, the REAL PageRank value, rather than a 0-10 display), I would love to hear about it!

ciml

9:01 pm on Feb 21, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



I don't see a problem with 6.7 and 5.4 as long as they're using the same Toolbar PR scale.

For example, if someone was to say "my page has 50% more raw PageRank than a Toolbar PR6 page therefore It's PR6.5" then I'd disagree. On the other hand, if someone was to say "my page is half way between PR6 on the Toolbar and PR7 on the Toolbar therefore It's PR6.5" then I'd only wonder if the methodology was OK.

The Toolbar resolution can be improved upon considerably when analysing simple link structures, but it's no easy task when looking at a typical Web site.

If anyone can get better resolution than +/1 0.016 of a Toolbar notch then I'm interested. :-)

aspdesigner

11:45 am on Feb 22, 2003 (gmt 0)

10+ Year Member



But how could they possibly know?

They're not going to find that kind of resolution from the toolbar, either from the display or even from the raw data.

And there is no way they could calculate the PR to that resolution themselves.

Calculating PR requires MASSIVE interative calulations which involve every link across the Internet in the Google index! The size of this task is one of the reasons PR is only updated once a month.

I seriously doubt that any SEO or webmaster is going to have either the data or the raw horsepower to attempt to do this themselves.

Even if you were to short-cut the process and simply use the results of the final interation (i.e. - the final PR of all the pages linking to yours), you would still need the actual accurate, internal PR values for each of these pages in order to get a precise result.

If all you have for inputs are the imprecise integer toolbar values for the pages linking to you, this lack of precision in your data makes it impossible to expect greater precision in your results.

Or, as the old computer adage goes - "Garbage in - Garbage Out"

Whether they were just confused or outright lying to him is another story. But given their refusal to answer his question as to how they determined these values, my guess is the later.

ciml

12:11 pm on Feb 22, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



It's not impossible aspdesigner; if you happen to have an appropriate link structure then you can get more precise figures with considerable accuracy.

I don't suggest that this is likely to be the case with webdev's email correspondents, as Chris_R points out most people misunderstand PR (and/or the Toolbar).

aspdesigner

12:51 pm on Feb 22, 2003 (gmt 0)

10+ Year Member



But they are still limitted by the inaccuracy of the input data. Remember that PR is based on the sum of the PR credit from all of the inbound links. If you only have a rough idea of the PR of each of these inbound links, the results of summing these is going to be less precise, not more so.

Now, if they had the actual internal PageRank values of all the pages linking to theirs, knew what the precise exponental factor Google uses to translate these values into toolbar values, and what damping factor they are using, then I would grant it would be possible to calculate this with great accuracy, and be able to make such bold statements like "my PR is exactly 6.7"

But if all they have to go on is the toolbar display, I can't see how they can claim that their PR is 6.7 (and not 6.6 or 6.8) with any reasonable degree of accuracy.

Tapolyai

1:31 pm on Feb 22, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



I second Chris_R's note regarding the "alternate" PR calculation method. I think that's were, or similar source your webmaster got those fractions.

According to some, including NerdRank, the DMOZ directory within Google displays a 40 pixel Google PageRank bar. Some people have noticed this to vary within a single PR (for example, my PR5 page is a 22 pixels, so it's closer to a PR6)

The question is - does that .7 or .4 worth anything? I think that it nice as a curio, but of minimal value in the long run.

hakre

1:53 pm on Feb 22, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



to Tapolyai:
the .7 or .4 does count a lot if you're interested in your own pr after the linking. even 'the long way' :)

Tapolyai

8:31 pm on Feb 22, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



hakre, my thinking with fractions of PR on a long run is that they fluctuate with other sites' ranking-change with the same key-phrase, the key-phrase neighborhood changes, site copywriting changes, etc.

In essence, I take fraction as just the noise within the equation. Of course this depends what your time frame is.

aspdesigner

2:31 am on Feb 23, 2003 (gmt 0)

10+ Year Member



PR is a fixed value calculated for each page independent from other ranking criteria, such as keywords. It doesn't "change" depending on what search you do.

Even if were possible to determine them, fractional toolbar values would be of limited usefulness. I would MUCH rather have the actual (internal) PageRank value that Google uses to calculate rankings!

Remember that the toolbar display is NOT your actual PageRank, it is just a rough attempt to display all of the possible PageRank values on a 0-10 scale.

This scale is not even linear, the actual PageRank increases exponentially for each toolbar level (i.e. - the difference in actual PageRank between a "7" and an "8" on the toolbar is MUCH greater than between a "2" and a "3"!)

Each toolbar level covers a wide range of actual PageRank values. While any change in your actual PageRank can effect your rankings, you won't see it on the toolbar unless it is large enough to move you to the next "level". You can often notice a change in PageRank by a change in your rankings, long before that change is large enough to see on the toolbar.

The higher your PageRank goes, the worse this problem becomes, due to the exponential increase in the size of each range.

NickCoons

4:14 am on Feb 23, 2003 (gmt 0)

10+ Year Member



aspdesigner,

<This scale is not even linear, the actual PageRank increases exponentially for each toolbar level (i.e. - the difference in actual PageRank between a "7" and an "8" on the toolbar is MUCH greater than between a "2" and a "3"!)>

Another way to put this "exponentialness" into perspective is to realize that having a page with toolbar PR5.5 (if you could know that) would have about three times the actual PR of a toolbar PR5 page.

aspdesigner

4:38 am on Feb 23, 2003 (gmt 0)

10+ Year Member



The exact amount of that would also be dependent on the precise exponential factor Google was using.

NickCoons

9:33 am on Feb 23, 2003 (gmt 0)

10+ Year Member



Most of the reports I've read say that it is somewhere around 6 to 8, but I don't believe I have any way of knowing that for sure.