I’m very proud of the Healthcare 100, because I think it has helped a lot of people (bloggers and non-bloggers alike) discover valuable new sources of information about health and medicine, healthcare policy, and other important issues.
I also think our ranking algorithm works pretty well. When we have issues with a blog’s place in the rankings, as we sometimes do, it’s usually not the fault of the algorithm; the problem is that, in attempting to attain up-to-date data from Technorati, Bloglines, and Google, we sometimes hit a snag. Often, it’s a problem on their end — not ours. Nonetheless, this sometimes results in rankings that we must go back and manually correct.
Such has been the case recently with Medgadget and Pharma Giles. In the first instance, the Medgadget folks sent an e-mail pointing out to us that their blog had fallen to #16 in the rankings, which seemed low to them. They were right; we were getting incorrect data from Technorati. Now, they’re back at #3 where they belong.
Pharma Giles, meanwhile, was pleasantly surprised to see that it ranked #30 on the Healthcare 100 on Friday. We were, too, since the Technorati and Bloglines scores looked high to us. We checked it out and they were high. So, for the moment, Pharma Giles is out of the top 100, but we’re sure they’ll be back soon.
We’ll keep working to keep the inaccuracies to a minimum — and you can help by dropping us a line if you come across anything that seems off to you.
Thanks!
[...] Last month MedBlog.nl collected data from 593 english written medblogs. Number 1 of August is KevinMD. Runner up this month is Medgadget.com, yeah those guys. The solution of Technorati problems probably causes their steep advance from 137 to 2. Third on the list is Bad Science. Number 1 of the sneak preview Healthbolt dropped three positions to rank 4. Gruntdoc at position 46 makes the biggest improvement (+336). The number between the brackets displays the rise or decrease of rank (1). Below you can see the top 100. [...]
Cary,
I’ve just come across Healthcare 100 and it’s clearly a valuable resource.
Maybe this has been said before, but let me say it anyway. Your algorithm adds up four scores with different maxima (10, 20, 30 and 10) and either 0 or 1 as a minimum. Surely that means that the component score with a maximum of 30 (Technorati) is three times more influential on the Total Score than the two with maxima of 10 (Google and edrugsearch), with the Bloglines score having an intermediate influence?
Is there an implicit value judgment in the algorithm that Technorati is more reliable than the others?
John,
The algorithm does weight Technorati more heavily, not because its data is more reliable — but because what Technorati measures (recent inbound blog links) is the best single indication of influence.
That said, measurements like subscriber totals and PageRank help fill out the picture.