The TechMeme Bikini

Some stats regarding the distribution of headlines on TechMeme between A-listers and others seems to be getting some attention. Attention, but little real thought. The basic observation is that while 70% of the headlines on TechMeme are accounted for by the top 100 ranked sources (according to the leaderboard), 30% is from the long tail.

For the sake of argument, let's assume that the data is static - that is to say, that the leader board 100 is always the same (it isn't). Let's also assume that TechMeme crawls 10k weblogs (I don't know that it crawls that many). Let's make some more assumptions: that every weblog posts 1 post a day and that there are 10 headlines per day on TechMeme. Thus, there are 100 posts per day from the top 100 sources and 7 of them will appear on TechMeme. Thus, each of the top 100 sources has a 7% chance of producing a headline on any given day.

The other 3 headlines come from the remaining 9, 900 sources. Thus, if they are also producing 1 post a day, each source has a 3/9, 900 =  0.03 % chance of getting noticed. So while the 2:1 ratio of A-listers to others sounds good, for any individual, it actually translates to 233:1 odds (7/0.03).

Of course, the assumptions above are a little rough and there is absolutely no accounting for how network effects really get things done in the blogosphere. The point is, there is a 2 orders of magnitude difference in these numbers between what an individual can expect and what the groups (A-listers/others) can expect.