Comments on: How Obama lost the healthcare debate online

By: Dave

Dave — Fri, 11 Sep 2009 14:16:06 +0000

The graph shows how one blog affects another. To show actual influence on US voters, you need to factor in the US readership of each site – or more strictly, the number of page views of the relevant articles.

By: Mark

Mark — Fri, 11 Sep 2009 04:44:21 +0000

Hi Gray,

The presence (and prominence) of UK stakeholders derives from the framing of the topic in the context of the British National Health Service.

By: Gray

Gray — Thu, 10 Sep 2009 12:23:10 +0000

To clarify my last point: The Torygraph and the Guardian may be more influential in the UK, I don’t know. But not globally, and certainly not in the US! And, remember, this study allegedly is about the “US Healthcare debate”. But the algorithm doesn’t seem to be limited to influence in the US. Quite to the contrary, it seems to be that you put unduly weight on UK sources when you did “crawl the internet looking for pages which are about the topic” and “track mutual references between people, institutions, entities mentioned in the context”.

This doesn’t make any sense when the focus of the study should have been the “US Healthcare debate”. It’s simply delusional to believe a British newspaper is more influential IN THE US than the NYT. Sry, but this is obvious nonsense.

By: Gray

Gray — Thu, 10 Sep 2009 12:12:45 +0000

One other point that should have rung your alarm bells: According to this study, the Torygraph is more influential than the NYT! Now, come on, really. In your wet dreams maybe! This is obvious evidence that there is something seriously wrong with the algorith used for computing “influence”.

By: Gray

Gray — Thu, 10 Sep 2009 12:05:55 +0000

This chart is total nonsense. You show Twitter and “Daily Telegraph Bloggers” as single points in the graph, even though they are only providing a service for a hodgepodge of users with different opinions. If you would instead have shown single Twitter users , or single bloggers at DT, they would have shown to be so uninfluential that they are located at the lower end of the graph. If you were consistent in your misguided approach, you would at least have treated blogger.com in the same way, which would then have been at the upper end of the chart, too. Sry, but obviously your approach not only doesn’t make any sense, it is really dishonest.

One more point: The “liberal blogosphere” may look uninfluential, but its strenth is in the numbers. If you add three or four blogs with similar opinions up, their aggregated influence passes even the most powerful news publications. And that at a fraction of the costs involved. Not bad at all.