Forum Moderators: open
I think there is a pr glass ceiling for mom and pops at pr7
3dflags receives over a million hits a day based on information at crames. Ah-ha, thats the answer, Google is tracking click thrus. If that site is getting that many hits a day, how many are clicking through to crames? Wow, impressive! The sheer number of click thrus is the reason for the PR8, am I correct?
P.S. There are five links on that 3dflags page, one of them goes to crames. There are 4,580 links pointing to 3dflags. I'd refer to that as PR contamination in a positive way. There is so much there, that one link leading to crames has quite a bit of power behind it!
dont really want to mention the name but it's a.com and is pretty "rotten"
very cryptic
Not many PR9/10 pages to be linked from, and....supposedly if they have such PR they will have traffic to back it up.
Paying for the traffic would inevitably mean acquiring PR from any link you bought.
But if there are any sites that openly link to other websites and have a PR9 / 10 on their page then someone sticky me with it before you start a stampede in here ;)
Ive always thought that PR from links aside, there is some sort of PR boost for uniqueness, ie. unique info, or valuable resource (like w.m.w) but thats just a hunch (and a hopeful one at that)
/just realised im a tad OT, but I do not know of any "mom and pop" sites with PR8 apart from those already now mentioned
Thanks google! Actually thanks to all the folks who linked to me ;)
[stallman.org...]
So, I say the heck with PR .
I agree. I think that anything pr5 or above usually does the trick. I don't think that PR is "ranking" as high as it used to.
I find one Mom & Pop commercial affiliate site [sunfinder.com...] ... Most of the links are clicks to ads or affiliates on this site yet it has a PR8 ...
Has anyone had a look at Google's 'backwards links' to this page? The pages that I have checked don't have links to it at all.
Is that right or am I being a bit of a melon?
Cy
seems ok to me, just the hollywood stuff keeps on interfering with pop-ups and on the first check my IE browser chocked on the googleback-link for it. On the second check the first sites do have back-links.
Pageoneresults,
you are absolutely right, on Olwen's find of crames.com, there is no substantial feed of Pagerank to this PR8(index)page! 3d-flags has a PR of 7. The crames subpages, as far as a quick check showed, have no external feed as well, so its not a case of a high PR7 linking to a low PR8, topped up with internal pagerank gifts from high internal pages.
So where does he get his pagerank from?
Am I missing something, or is this one of the few Pageranks I cannot fathom?
Your click-through theory could be something. Now that his 3dflags site is under construction (and for a while? see the cache) he is probably getting more clicks than normal and there are no other outward links to go to or to choose from. It probably also helps to have an outward link from an index page...
Anyone else ideas?
people.ee.ethz.ch/~oetiker/
It's a private page with lot's of usefull stuff.
And it has PR8
IndividualsRalf S. Engelschall - [engelschall.com...]
Jakob Nielsen - [useit.com...]
Tobi Oetiker - [people.ee.ethz.ch...]
Richard Stallman - [stallman.org...](I can see a pattern there!)
Individuals affiliated with large organisations
Tim Berners-Lee - [w3.org...]
Bill Gates - [microsoft.com...](pattern continues)
Small (?) businesses
[crames.com...]
[sunfinder.com...](better than Brett?)
[edited by: danny at 10:40 am (utc) on July 2, 2002]
The rest are nonprofit or net celebs (trying to find any small bizes that have scaled the pr mountain).
1. I thought clickthoughs are not really involved in calculating PR or are they in a major way?
2. When I did google backlinks on his site it came out with only two pages of links, most of which were his own domain with only one PR7 link, doesn't the site with so few links need to have at least one PR9 link to get a PR8 rank or at least a good number of PR7, PR6 and PR5 links?
3. I have about 400 backlinks in Google to one of my pages and about 10 of them have PR6 and this page only has a PR6, it is hard for me to understand how he can get PR8? Does he have 1000s of < PR3 links that don't show in backlinks?
I am in the same confused boat (see my message #44)
Normally speaking a high PR7 (possibly page3d-flags) with only one outbound link (which is the case) could yield a high PR7 page. If the sub-pages would add PR by linking to the index of Crames it could get a low PR8. The problem is that the subpages are getting all their pagerank from Crame's index, not from external sources, so they actually do not contribute anything substantially extra.
Also there are no 20.000 sub-PR4 page links not showing in Google but appearing in Alltheweb.
It is a nice mystery, a glitch, or indeed some new pagerank calculation taking other things such as click-through into consideration.
Its like here is this little bucket. Someone sticks a large funnel in the bucket knowing that anything poured into it is going to overflow. Those 4,000 + links to 3dflags are like overflow. As soon as someone clicks on the link to crames, wham, there goes a release of PR, and a big one!
Since the link to crames is the only place to go from 3dflags, its getting a lot of click thrus. It was mentioned on crames that the 3dflags site gets millions of hits per day. If I take a small percentage of those million hits and they clicked through to crames, then I'd have to logically assume that crames is getting a boost for click thrus. I could be wrong, but there is nothing else there that would really explain the PR8.
The other thing I was thinking about are how many users that visit the 3dflags site have the Google Toolbar installed. Since the toolbar tracks history, that would be one of the ways where click thrus could be calculated into the PR.
P.S. I'm thinking the same thing is happening with Sunfinder. They have a very high volume of click thrus from a variety of high PR properties due to their advertising.
The effect of inbound links is larger than one may think. If a website has no outbound links and receives an additional inbound link from page X, the PR boost for the whole site is (d/(1-d))*(PR(X)/C(X)).
If page X has only one outbound link and d equals 0.85, the PageRank boost for the whole site is 5.67 times the PageRank of page X. This PageRank can then be distributed to the pages of the site. The page which receives the additional inbound link always just gets a minimum PR boost of d(PR(X)/C(X)). The rest is PageRank feedback from internal pages. But at a damping factor of 0.85, the PageRank boost is most likely higher than the PageRank of the linking site if it has only one outbound link.
(As far as I can see, there is one outbound link on the crames site to 3dflags. This reduces the PageRank boost for crames, but not very much, since the link is buried in the site.)
If you don´t believe me ;), a little example:
We have a two pages website consisting of page A and page B, which link to each other. Without inbound links both pages have a PR of 1. If page A receives and inbound link from a page that has a PR of 10 and no other outbound links, the PR of pages A and B is at a damping factor of 0.75 given as follows:
PR(A) = 0.25 + 0.75 (10 + PR(B))
PR(B) = 0.25 + 0.75 (PR(A))
If you take a pencil, you get the following PRs:
PR(A) = 127/7
PR(B) = 97/7
The total PR of the site is now 224/7 = 32. So the total PR boost is 30. And 30 equals (0.75/0.25)*10.
But the PR of page A is about 18.14, which is definitely higher than 10. Just a little bit of maths. And sorry for the long post.
PR(A) = logn(0.25 + 0.75 (n^10 + n^PR(B)))
PR(B) = logn(0.25 + 0.75 (n^PR(A)))
Instead of about 13 and 17, you get about 10.1 and 10.2 It makes a difference, but not as big a difference.
Internal linking should be worth roughly a half a notch on the Toolbar. By my calculation, two interlinked pages A and B, with a link to A from one PR7.5ish page could leave A with a PageRank of about 8.0 after iterating; assuming d=0.85 and assuming the original formula.
I just wanted give an easy to calculate example that shows that the page which receives the link having a higher PR than the linking page complies with the original algo. By just looking at the algo you easily underestimate the effect of inbound links.
Of course, in terms of toolbar PR, the effect is not that noticable. But the crames home page having a low PR8 and the 3dflags page having a high PR7 can absolutely be explained by the original algo. There is no need for clickthroughs.
[edited by: namniboose at 8:03 pm (utc) on July 2, 2002]
Indeed seems they look to be border-line PR7/PR8.
Ciml's numbers seem to add up, however I assumed (but never iterated/calculated) that the effect of the interlinking subpages (B) of Crames (+ one external link) would have more of a dampening effect on the final Pagerank kick-back effect towards Crames'index page(A) than explained above?
Given the fact that the direct subpages of Crames' still have PR8, I would also guess the dampening & dilution factor from an index page to be very limited, at least these are effects I see and try to understand from my own site.
