Pagerank explained by Matt Cutts
Matt Cutts has recently written an interesting new article on how Pagerank works in Google. There is no question that link analysis has gone through extensive changes since the original introduction of Pagerank by the Google founders and it’s refreshing to see the head of the webspam team weigh in on how things work. I’ve posted a link to the original article, but I’ll comment on the most interesting parts. Basically, Matt explains page rank as a flow of points from one page to the next which are equally divided among outgoing links. If a page has accumulated 50 points from external sites and there are two outbound links, then those two outbound links will get 25 points. Then, he continues to explain that nofollow links do not pass pagerank. This is not new. What IS new Matt’s introduction of a “decay” concept.
This just makes logical sense since otherwise, it is very easy to get into infinite loops.
The question is… When a series of pages gets into these loops, do they continue to pass on points to the pages until the decay factors it out? If it is smart, it will remember what pages link to an external page and factor it out, but remember the scale at which google is doing this. It would take a lot of memory and disk space to keep a continuous log of all the incoming links for every page during indexing. That means that most likely, it is NOT keeping such a log.
The difficulty with testing this is that Google updates Pageranks VERY infrequently (on the order of 1-2 times a year). Further testing is warranted, but in the end, the question that comes up is … does pagerank matter? The truth is that pagerank does not matter unless you are selling paid links or advertising. It is the only public indication besides your weblogs of how trustworthy your site is in Google’s eyes.
The next question is… does it matter for your SERPS? In our experience, no. It is possible to rank well for competitive keywords without high pagerank and there are many examples of it. A couple of interesting points is that when we tested the nofollow attributes, we did notice that in spite of the fact that Google specifically said that nofollow had no effect on SERPS, that in fact it did. Matt explains it as bugs in the initial indexing, but as an SEO, with so much misinformation on the internet, we are trained to not believe everything that is said, but to do empirical tests to see if that is actually what is going on.
Popularity: 25% [?]