THIS IS A DRAFT. Posting this in the hope that someone else can do it better– like someone who researches for a living. The goal here is to try and determine “leading indicators” of online political writers, and find out who we should be reading more, and trying to avoid the trap of the”power law”, (see Shirky) which holds that people read the writers largely because other people read them.

Part of comparative studies of blogs and media. This is a new stab at trying to get some statistical data about online political writers.

Sources:

Truth Laid Bear (TLB) – “Ecosystem” rankings, Jan. 19th

The Power and Politics of Blogs by Daniel Drezner and Henry Farrell, July 2004

Obviously, there’s a horrible discrimination in the data that near-blogs, community-sites, and non-blogs are exempt from the data collection. We can compare bloggers to bloggers but not to anybody else.

Legend:

Frequency is # of postings per week. Averated over last quarter of 2003.
MR – Number of “media readers” according to Drezner & Farell
EMR – Number of “elite media” readers according to Drezner & Farrell
MM – Number of media mentions according to Drezner & Farell
TLBT – TruthLaidBear Traffic, average daily page views
V/P – Views per post
T# – TLB Tra
TLBL – TBL Inbound links
L# – TBL Inbound Link rankings
SL – Links to stories, as opposed to main page
Story% – Percent of links that are to stories, of all links.
V/L – Number of views of a blog before someone creates a link to it
ESL – Links to stories by 100 top bloggers.
Top 10 – Add up the number of citations in the top 10 list.
CR– SL divided by Frequency.
ECR – ECL divided by Frequency.

Problems with the data collection:

1. The frequency was taken from the last quarter of 2003. Malkin’s ramped up her writing tremendously, I just patched in her number for the week (*)
2. Drezner and Farrell’s data is from July 2004
3. TLBL data is not matching when I download it from TLB.
4. The Top 10 data does not exist for many blogs
5. CR/ECR; these ratios are not normalized against anything (other than each other)
6. The “frequency” numbers for Kos may be off by a couple orders of magnitude. I could only count posts on that made the top– which may be 91 over last fall, but there are up to a thousand diary posts each week. Perhaps most of the story links are to top-stories, but still.

Otherwise, we can start to recognize some trends.

Political Spectrum

The Media Five: Sullivan, Reynolds, National Review, Kaus, TalkingPoints. 1 liberal, 4 conservatives (Sullivan defected from Bush)

The People’s Ten: Powerline, Daily Kos, Little Green Footballs, Reynolds, Atrios, Hewitt, Marshall, Sullivan, Drum, Malkin. 3 liberals, 7 conservatives (again, including Sullivan)

The prominence of conservatives in the top 5, 10, and perhaps 100 may explain the high Elite Cite Ratios. The best among liberals is TPM at .22, and Tapped getting up there at .26. But Tapped hasn’t reached the masses yet, despite posting 10x a day and having the high ECR. And having a real editor.

Consider that the liberal/conservative split in this country is fairly even; Republicans hold a slight edge in elected offices.

Is it that they just make sure to cite each other? If there’s a tight-knit concentration of conservative pundits at the top (which is the reliable stereotype of Republican party, in fact), that may explain it. More work would need to be done to determine the buddy-buddy citations.