Pr Update!?

Mel: I just have not seen anything on the internet that holds to Dodger's view. Not even Fabien who's article he cites. What do you mean by a ranking of 4/7?

 

It appears to me that there are seven ranking steps in the directory greenbar as compared to the ten steps in the toolbar PR.

 

There are entries in the Google directory that are not ranked yet. They would be new to the directory of course and hard to come up with any examples, because as you know the it was just updated.

But here is one example. Personal Home Page Directory which is in a directory section, that, if you go into any of the alphabitized areas of this section -- toward the end you will find a whole slew of unranked pages. They are ALL blanks.

As time progresses and more sites are added to the directory, you will find more of these examples. After the directory has a PR update, those blank entries will disappear -- very easy to spot a directory update has occured.

It is important to note that having a blank bar in the directory does not mean that the site has no actual PR. It is just so happens that after the directory has been updated, these sites stick out like a sore thumb.

Actually when you get down to it. All of the scales and assumptions made about them, well quite frankly, they suck. PR is more logrithmic in nature, and converting that to a linear scale of any type does not compute -- mathematically speaking.

 

I think that when you think about PR being logarithmic, Ron, you must be thinking about the way that Google obfusticates the ranking in the toolbar.

PR is a straight linear function. Look at the forumula.

 

It is a re-iterative formula which branches out exponentially. This is what I meant by "logrithmic in nature" ... but that is assuming it is base 10 (which nobody can really say).

The scale on the PR bar, while displayed in equal increments, are not truly equal. Each step is exponentially greater than the next. Sort of like the richter scale for earthquakes.

 

Dodger: So are you still holding to the view 'think of the directory guages as a "viewport" that only shows PR3 and up to part of PR10'? I just noticed you made a couple posts since I posted back some questions but didn't address my questions.

I wrote a post directed to you showing toolbar PR10 site that is directory PR 38/2, and toolbar PR9 site that is directory PR 32/8. Samples of this could be multiplied.

So which is it, if your guage theory that toolbar PR10 is dir 44/0 and toolbar PR9 is dir 38/2 what about the "www.microsoft.com" (PR10) and "www.cnn.com" (PR9) examples above? Have those dropped down to a toolbar display PR9 and a toolbar display PR8 (according to the way you read the guages) or is your position you posted to me incorrect?

Also I noted that the person that wrote the article you cite don't support your above statement. Further I asking if you had any articles that support your stance. Let me know.

 

digitalpoint: Well I relize that the directory displayed PR was updated on Aug10 and the last update for toolbar display PR was June 23. But I used 2 examples that have been PR10 and PR9 for a good number of months and there is not alot of chance they have changed.

If I am understanding your view you are putting foreth that 44/0 is a PR10, 38/2 is a PR9, 32/8 is a PR8 all the way down to a 5/35 is a PR3? Also if you view is correct then "www.microsoft.com" is now a toolbar display PR9 and "www.cnn.com" is a toolbar dispaly PR8? Or would be if the toolbar display PR was updated when the directory display PR was updated.

But of course you can say when the toolbar display PR update is done that it has gone back up and the directory displayed PR is out of date. So there is wiggle room to run around the bush on this one.

It is very clear in my mind that cnn.com has not dropped down to a PR8. To drop from a PR9 to a PR8 is like a huge huge drop. While I would be willing to admit that microsofts raw PR does not scale over to a very strong toolbar display PR10 as they only have a total of 5 toolbar display PR10's, (Adobe has over 50 toolbar display PR10s, Google has some odd 35 and Apple is around 27) it is not impossible for them to drop down to a PR9 but highly unlikely. See my PR 10 Sites page.

SEO Chat, SEO Guy and Digital Point show as 27/13 directory display PR. If I understand you correct you are saying that this is the same as a toolbar display PR7. Is that correct?

Now the current toolbar display for SEO Chat is PR8, for SEO Guy it is PR7 and for Digital Point it is a PR6. According to your view if raw or real PR was updated to the toolbar display PR at the same time the directory display PR was updated they would all read PR7's on the toolbar. That means Digital Point has gone up from PR6 to PR7 and SEO Chat has gone for PR8 to PR7. Is that what you are holding to?

Now you ask me the same thing, which I think it is fair. How do I deal with those three sites. If you look at The Handy Dandy Google Page Rank Figurin' Guide by Chris Raimondi's you will see that the 27/13 directory display PR takes in a high toolbar PR6, all of toolbar PR7 and a low toolbar PR8.

So that scale handles that example you producted but I think it is fair to say the chances of microsoft.com going to a PR9 is very small and that changes of cnn.com going to a PR8 is yet smaller still.

By the way I find these discussions very interesting and if I can be shown that the position I have taken is wrong I would be happy to change. I think discussions like these are very helpful and make us think. Thanks for you time!

 

digitalpoint: Well we are both dealing with theories and we could be both wrong. I guess we will need to wait and see. If SEO Chat drops to a toolbar display PR7, SEO Guy stays the same, Digital Point goes up to a PR7, cnn.com drops to a PR8, and microsoft drops to a PR9 then the theory you are holding to could be true.
If microsoft.com stays at a toolbar display PR10, cnn.com stays at toolbar display PR9, I would have to say that your theory is wrong. Of course microsoft.com may have gone back up to a PR10 and cnn.com gone back up to a PR9 by the next toolbar dispaly update.
I would say from the examples we looked at that Raimondi's theory seems more credible considering the examples we have looked at.

 

Yes. And guilty.

This is not my forum of choice to spend most of my time on. When I get the notifications there are many posts that took place. After I finally get around to click thru and due to the nature of vBulletin -- it skips me to another page and not to the actual post that triggered the notification. I address the posts on that page and do not backtrack, especially when the conversation does not seem to be going anywhere.

The fact is I do not support either one totally. But I support the sliding window theory more. Actually it is more like I support the other one less if you really want to know.

There are too many assumptions being made. One (Shawn just pointed out) is that the measurments being compared are taken at different points in time. Does PR magically hold still so you can make comparison notes? I highly doubt it.

The toolbar shows an approximation of what the true PR is. You do not know what it actually is. It is close enough to gauge at lower PR values (2 or 3 for instance), but when you get into the higher end of the bar those increments become exponentially greater.

Another assumption made is that the PR in the directory is computed exactly the same as the toolbar PR. This fact alone is enough to throw any and all theories out the window. It makes this whole conversation ridiculous to even argue about.

It could possibly be that the PR in the directory was a more refined value that went thru more iterations. Possibly this is why it is not updated that often in comparison to the toolbar. Maybe it is less refined and not that reliable to begin with.

Perhaps the directory uses a different process altogether, which could be the case because you can see large PR values for entries in the directory that you would not normally see with the toolbar. These entries are for sub-pages and both theories do not support the larger values in the directory either way.

I think it is silly to keep referring to the directory PR as 32/8 and 38/2. Of course, if you are talking about 7 of 9, then it is not that silly and I highly encourage it. These are pixel widths not PR values which translate to DPR7 and DPR6 respectively. Yes I just now made DPR up, unless you can come up with something better to call it. )

 

Dogder: I looked at your orginal post and to me and it seems your view has changed. I think Raimondi view makes much more sense to me than your orginal view. I will stick with the 5/35 - 38/2 scale of 7 as it seems to fit. I expect we will see when the toolbar display PR is updated that that microsoft.com stays a toolbar display PR10 and cnn.com will say a toolbar display PR9.

There is only one real or raw PR I see no support for 2.

I have found no directory displayed PR that doesn't convert over to toolbar displayed PR according to Raimondi's comparson chart.

 

I agree with Bobmutch in that IMO there is only one true PR calc, it is a time consuming calc, and I can see no reason why Google would want to recalculate PR just to display it in the directory when they have the data already avialable to them.

IMO the directory and PR toolbar values are derived from the True PR values with a simple mathematical operation, though it a appears to be different in for the directory and The toolbar.

 

Mel: Your view is the commonly accepted view.

Most hold that there is a real or raw PR that is probably on a manual scalation which follows a logarithmical scheme (Wakfer's chart uses base 5.5 and Sobek's scale is base 6).

This raw PR is displayed on the toolbar on a scale of 1-10 and in the directory on a scale with 7 steps, 2/38 to 38/2.

I hold that Google updates the real PR after every BL update and that it has not been updated on the toolbar because they have gone to quarterly updates.

 

Thats not really what I am saying Bob, I am saying that there are (at least) three different PR figures:

True PR which no one but Google knows about, and which is the number used in any ranking calculations. I see no way that this is a logarithmic function, it is a simple a times b type calc, and could have a range of between 0 and several billion.

Toolbar PR where a range of true PR values are arbitrarily defined as belonging to one of the boxes labeled 1 to 10. It appears that the selection of these ranges is based on a logarithmic scale of some sort, perhaps something as simple as plotting the log of True PR and dividing the graph into ten equally spaced divisions which are then labeled 1 thru 10.


Directory PR, similar to Toolbar PR but the number of boxes is changed from ten to eight or more, and possibly the selection process is different also.

I believe that Google has a private up-to-date database of all the links it knows about, and that database is updated at the very least several times a day as links are found and dropped. PR is calculated by taking a snapshot of that database at a certain time/date and using that to calculate PR, a process that IMO takes some time to accomplish.

Once a new PR set is calculated then the public PR and links database snapshot becomes the public links database (or at least as much of it as they are willing to show us).

BTW Bob I measure the width of the Directory toolbar as normally being 40 pixels wide, with the first division at 5 pixels, but with Microsofts home page PR10 measuring only 39 pixels wide, Googles home page PR10 measuring 45 and Froogles homepage PR10 measuring 39 pixels wide but Google Blogs PR 9 page measuring 38 pixels wide. The selection process for the directory bar may be somewhat different than that of the tool bar.

 

One PR calculation? Two more mathematical calculations? Three (at least) separate figures?

Huh?

 

Mel:

Google updates there BL about once a month. In the last 89 days we have had 4 or 5 BL updates. When a BL update has been done the backlink displayed changes. At this point people see a change in there rankings as there new backlinks have been given credit. The backlinks are NOT updated every several days. That people get credit for their new backlinks after a BL update is common knowledge.

The scale for the directory displayed PR is 2/38, 11/29, 16/24, 22/28, 27/13, 32/8, 38/2. Mircosofts home page PR10 is a 38/2 not a 39. I am not sure what you saying above. Are you saying that Froogles home page is a 39 in the directory? Where do you get the 45 for "www.google.com" , it is a 44/0. Not sure what you are meaning here.

On the other stuff its seems we agree, I think we are just saying it different. Where real or raw PR is logarithmical or times b type calc I need to look into that a bit further and will post on that later.

 

It works fine.
One is for real, the others are for show

 

There are no others. They are just different ways of displaying the same thing. The confusion in some people's minds is due to the different displays not being updated and not using the same intervals or values on the axes.

 

Yes Minstrel that is correct. There is no known method of translating the PR shown on either the toolbar or the Directory into the figures that are used in the ranking calcs.

You really don't think that the PR shown in the toolbar and/or the Google directory are the same figures that Google uses in its ranking calcs do you??

I suggest that it might be helpful to read the original Paper written by Page and Sergy which defines the PageRank equation and how it is used so that you can obtain a better understanding of PageRank.

There are those who believe that this original equation has been modified by Google over time, but I would like to suggest that this is not at all likely as the PageRank Algo is the property of Stanford University, and is used by Google under license from them. Google have recently extended this license thru 2011 on an exclusive basis, which would not seem to be necessary if they were no longer using the original patented forumula owned by Stanford.