Forum Moderators: open
i tried to find the answer on the following question on the web and on this site but couldn't:
what does google consider a "link"?
as far as i understand google calculates a page's PR taking into account the following factors (simplified):
- number of inbound links in combination of the linking pages' PR
- number of outbound links
so, regarding ourbound links the question is "what is a link?" i can think of the following possibilities (while there are probably more):
- normal anchor tag <a href=..>
- javascript anchor tag without href <a onclick=..> (or onmouseover, etc.)
- form action <form> in combination with <input>
- any other javascript link, eg. lets say <img onclick=dowhatever>
what does google take into account as "links"? only standard <a href> or also other types? wouldn't i start trying to "javascript-mask" all links leaving my site to keep PR within my own site?
and would i risk being banned by google using any masking?
best regards from vienna,
muesli
You could, of course, mask your outbound links using javascript, but then you might find that those linking back to you would quit since your links would not really be reciprocal.
as far as i understand google calculates a page's PR taking into account the following factors (simplified):
- number of inbound links in combination of the linking pages' PR
- number of outbound links
... is an accurate description of the situation. I'd modify that instead to:
- number of inbound links in combination of the linking pages' PR and the number of outbound links on those linking pages.
That is, each page has some "amount" of PR that it can pass on to other pages, and the strength of that PR is distributed among all of the page's outbound links. The more outbound links there are, the less each will be helped by a link from that page. That is the "detrimental" effect that a high number of outbound links may have.
But I do not believe it to be the case that an individual page's PR is lowered directly by the PageRank forumula based on how many outbound links are on that page.
But outbound links matter in the overall PR collection you have on your site in terms of leaking PR out to your links. A smart system of navigation should help that. And since a site that doesn't have outbound links may have a little trouble getting inbounds, you have to do some of that. And even if Googlebot can read and follow JS links (I don't think it can), most people think it can't so linking via JS is probably a really bad political move.
ok, i understand.
so google most probably cannot read javascript links. any sources that exactly proof what the bot can and what not? anybody who has exactly analyzed what the bot did on his website, how it crawled?
one more issue: basically my question was "what does google consider a link?" in contrast to "which links does the bot follow?". eg. the bot
- doesn't follow links with question marks on pages that have a question mark in their URL
- (probably) doesn't follow javascript or <form> links.
- (probably) doesn't follow textmark links within this very page.
but does google CONSIDER such a dynamic or js link as a link when calculating the PR page A passes on to page B per link? maybe google counts the number of "<a " on a page when calculating pagerank?
i.e. JayC, you mention the importance of the number of outbound links on those linking pages. this means it is well important to know what google considers a link.
one more question (JayC corrected me so that i'm not sure anymore): DO outbound links influence the linking page's PR? do i lower the PR of this very page (i'm now NOT talking about the PR lost for my entire site) if i - say - add one more link?
best regards from vienna, austria,
muesli
Googlebot can index pages with ? in the URL, and can follow links from such pages, but with less likelihood (the main factor seems to be PageRank).
I don't know what textmark links are.
> DO outbound links influence the linking page's PR?
No, unless you link to a 'bad neighbourhood' and pick up a rare penalty.
thanks for the information. i'll try to research further on the issue, this is an interesting topic. will post results here.
>I don't know what textmark links are.
textmark links are links to text parts on the same page, in the format <a href="#paragraph2">, often used eg. in FAQ-pages.
>Googlebot can index pages with ? in the URL, and can follow links from such pages, but with less likelihood (the main factor seems to be PageRank).
it's hard for me to believe this. i agree (at least i consider it possible) that google uses PR to decide on how deep it crawls, but this must be true for dynamic and static URLs. a question mark in a URL doesn't say anything about the quality or relevance of the page.
dynamic pages bear the risk that the bot might get trapped in a loop. most search engines therefor avoid dynamic pages, google and some others to my knowlegde avoids getting trapped by simply not following any links on the dynamic page. this way all dynamic pages reachable by a static (looking) one get indexed, the others don't.
i'm quite confident that the decision on crawling depth is an independent issue, as a high PR wouldn't reduce the risk of getting trapped.
muesli
