Nascent

Hi, Euan. What do you mean when you say, "blogrolls from blogs tagged in Nature.com Blogs"? You mean you went to each Nature.com Blog, looked at the blogs in each one's blogroll, went to each of *those* blogs, and then tabulated the blogs listed in *their* blogrolls?

Hey Amos,

Sorry, should have been a bit clearer about the methodology, which was basically:

1) download list of blogs (and their subject areas) from Nature Blogs

2) visit each one, pull out its blogroll (by downloading half a dozen posts published on different dates, working out which part of the pages never changed and then pulling out links from there)

3) filtering out non-blogs and blogs not listed in the Nature Blogs index

There was quite a lot of munging of URIs involved (so that scienceblogs.com/ matched www.scienceblogs.com , for example).

Nice.

But Neurotopia and Of Two Minds are not physics blogs, they are neuroscience blogs (more life science than anything else).

THere may be a serious flaw in your methodology. If you search for parts of the page that do not change you will never include any rotating blogroll. My blog roll includes about 300 sites, but displays only 50 at a time.

Can I claim my blogs yet?

Is this affected by the fall in popularity of the blogroll (static) in favour of imported widgets for RSS readers (now standard on Blogger (Google) blogs but not, sadly, on Typepad as yet). I think bloggers and others are tending nowadays to subscribe and unsubscribe to blogs in their RSS readers and not bothering to update their static blogrolls. In the "old days of blogging", blogrolls were ways to get bloggers to refer to each other, etc -- but now, many coversations go on separately at Friendfeed, twitter, Nature Network, etc, as well as at the blog itself.

Must be hard to capture a lot of this ephemeral activity for analytical purposes.

Another point: Nature Network blogs don't even have blogrolls. Will be interesting to see what happens to your analysis when they do. But for the moment, they are presumably disproportionately excluded from a blogroll-based analysis such as this?

Maxine: there's nothing to stop Network blogs being blogrolled by others which would make them show up here - OTOH I bet you're right in that you're far less likely to be blogrolled if you don't maintain a blogroll yourself.

Greg: you're right, the methodology definitely can't capture rotating blogrolls. I don't think it's too serious a flaw if the numbers of non-rotating blogrolls are big enough but it's a good point and something that would definitely be worth tackling properly if we wanted to use ranks like these for anything formal. It'd be dumb to not use all the data available.

What Bora mentioned is definitely a serious flaw: those aren't physics blogs! Must have a tag gremlin... will correct tomorrow.