1 of 1 link from pr6, or 1 of 8 from pr7

which will provide more PR?

[HackingLawyer]

The question stated in the subject and meta should be an easy for this board

[ciml]

In terms of PageRank transfer, one link of eight from a PR7.0 page is worth more than the only link from a PR6.0

However, one link of eight from a PR7.0 page is worth less than the only link from a PR6.4 page. So it depends.

Also, a link from a high PR page may not pass PR on. This seems to happen now quite often when there are a bunch of paid links, but I don't know if that's by design or whether PR sellers are getting the can-have-pr-but-not-pass-it-on penalty by linking to bad neighbourhoods.

> should be an easy for this board

I don't know how many people know the numbers, but it took me ten months to find them. :-)

Currently, I don't have an accurate answer to the same question from a link text boost perspective.

[taxpod]

I suppose there are a couple questions here. First off as cml pointed out, since toolbar PR is logarithmic, the first question is how strong of a PR6/7 are we talking about. It is possible for a strong 6 to be almost as strong as a weak 7. Then the question is what is the logarithmic base since if you divide the outbounds by more than that from a PR7, by definition you can't get any more than a single outbound from a PR6.

[HackingLawyer]

"by definition you can't get any more than a single outbound from a PR6."

I'm sorry, could you elaborate on that statement? I don't understand what you mean.

Thank you

[ciml]

The number of links on a page, where each linked page gets PageRank of exactly one notch less on the Toolbar if it has no other links to it.

In other words, PageA has exactly PR7.0, so how many links can PageA have that will give PR6.0 to the pages it links to?

Note, 2 links to the same URL count as one. The relationship between number of links and PageRank transferred is logarithmic, but breaks down at high numbers of links per page.

[taxpod]

HackingLawyer,

Sorry bit of a mental hickup there. What I'm saying is that a link's value would depend partly on the number of outbounds from the page and partly on the logarithmic base, I think. Like (excluding dampening factor):

Assuming log base of 4:

A) 1 outbound from a PR 6 would be worth (4^6)/1=4096

B) 1 of 7 outbounds from a PR 7 would be worth (4^7)/7=2341

C) 1 of 4 outbounds from a PR 7 would be worth (4^7)/4=4096

But if the base were 8, you get a different result:

A) 1 outbound from a PR 6 would be worth (8^6)/1=262144

B) 1 of 7 outbounds from a PR 7 would be worth (8^7)/7=299593

C) 1 of 4 outbounds from a PR 7 would be worth (8^7)/4=542288

[rfgdxm1]

>In terms of PageRank transfer, one link of eight from a PR7.0 page is worth more than the only link from a PR6.0

How do you know the base of PR function? I'd rather have the one link from a PR6.0 than one link of eight from a PR7. You must guess a higher base than I do.

[HackingLawyer]

Thanks for the responses. Is there a simple formula to determine whe PR transmitted ( assuming the tool bar number ) with the number of inbound links - using the number of links and the toolbar pr as the variables?

[ciml]

HackingLawyer, I think you're looking for (log_b(n) + x), where n is the number of links on the page, x is the Toolbar PR decay due to the rank source and b is a figure that people don't agree on. It works nicely until some point beyond which PageRank is lost more rapidly. Google suggest [google.com] using no more than 100 links per page so that seems a good bet.

RFG, yes I do use a higher figure for b than most people.

[HackingLawyer]

In order to figure the equivalant decay of a link to a pr7 with x inbound links and a pr6; can you derive a formula to arive ant the number of links on the pr7 wherby a link on the pr7 would transfer the same pr as a pr6 with only one link?

I tried to do this but quite frankly did not understand your formula.

If x is the decay, then one would assume an equal decay in order to derive the new equation. n, the number of links could be figured if a value for b is assumed.

Thanks for the help!

[HackingLawyer]

Now that the update is on, can someone take a moment derive the equation referenced in the post above?

[ciml]

HackingLawyer, I think you summed it up perfectly. x is the Toolbar figure for the amount of PageRank not passed on from a page to the pages it links to.

This is represented by (d) or more usually (1-d) depending on how you write the PageRank equations. It is a multiplier when used on the raw PageRank, but on a log scale it becomes a number to add (well, subtract) and is the the difference in Toolbar PR between a page with just one link and the page it links to, assuming that the page it links to has no other incoming links.

Note that log_b(1) is zero, no matter what b is (hence constant x).

> can you derive a formula to arive ant the number of links on the pr7 wherby a link on the pr7 would transfer the same pr as a pr6 with only one link

In raw PageRank, it's this (where PR7 is the raw PR that equals Toolbar PR7):

(1-d) . PR7 / n = (1-d) . PR6 / 1

On the Toolbar, it's this:

log_b((1-d) b⁷ / n) = log_b((1-d) b⁶ / 1)

I think that ends up as something like n = b⁷ / (b^(log_b((1-d) b⁶ / 1)) / (1-d))

n is the number you're looking for. If you can find b and d the equation is easy to solve.

There is a range of combinations of b and d that works for the example you gave, but only one combination should work for any set where n is less than some number (assuming that PR is calculated in the same way at the top and bottom ends of the scale).

[pielich]

about pagerank : is it possible that page rank works on log. scale ---> 10 pr 6 links have the same weight of 1 pr 7 links (on a same number of outbound links page)?

[ciml]

Hello pielich, yes that's what we're saying. The big question is the log base. Is it ten (as in your example), 6 (as many think), twenty-something (as I suspect) or 42 as GoogleGuy recently suggested (I think he was joking).

[HackingLawyer]

CML, I must admit to a lack of comfort with the numbers, its been quite a few years since agebra, calculus, DE and the like.

If someone is trying to decide whether to "buy a link" on a pr 7 or a link on the pr 6, the pr 7 has 8 links, and yours would be the only link on the 6 - and the cost being equal, clicks through being equal, exposure being equal-page rank transfer the only consideration in the decision - would the pr6 or the pr7 site be the better buy! Can this formula help with such a decision?

Thanks for enduring with this thread!

[HackingLawyer]

CIML, Is there a number or approximation that will work?

[ciml]

Yes, the PR from the PR6 page is:
log_b((1-d) b⁷ / 8)

The PR from the PR7 page is:
log_b((1-d) b⁶ / 1)

So if b is less than 8 then the link from the PR6.0 page with 1 link gives more PageRank, if b is more than 8 then the link from the PR7.0 page with 8 links gives more PageRank. We don't all agree on b, so this doesn't help you much I'm afraid.

However: If the PR6 page was really PR6.9 and the PR7 was PR7.0 then things would look different. Also, some pages have a 'can have PR but not pass it on' penalty. They rank well, they show plenty of PageRank, but buying PR is pointless. I think GoogleGuy wrote something that can be paraphrased as 'buyer beware', I would certainly agree.

[HackingLawyer]

Thank you CIML, your posts were very helpful. Is there any way to avoid caveat emptor( buyer beware ) by determining whether there is a "pass on penalty" referenced in your earlier post.

[shady]

I would think the easiest way is as follows:

Look at the PR of the sites which are being linked to

If the linking site's PR is 8 and some of the linked to sites are less than PR6 (and they list the linking site in their backlinks), I would think you can assume that the links are not worth bothering with from a PR point of view.

[ciml]

It used to be much easier, but I agree with shady. If you (think you) know b and d in those equations above, then you can tell that a page has less PR than it should have when looking at its backlinks.

Personally, I think that very few people do realise when this is happening. This is good for PR sellers, but bad for PR buyers.

[shady]

I must admit, I was highly disappointed to have got two PR8 links recently, which did not pass on any PR to my sites! At least they were virtually free!