Friday, May 20, 2011

On-Line Opinion and The Conversation or On distributions and variance

A colleague recently suggested that I should consider writing for On-Line Opinion as a way of getting my views more coverage. Separately Josh Gans has tweeted about The Conversation. So today I've had a bit of a look at both.

On-Line Opinion is self described as "Australia's e-journal of social and policy debate". It has been going for about six years now and is "professionally" edited though contributions are unpaid. The Conversation is very different having been started by former Age editor Andrew Jaspan, and drawing exclusively from academics for content and even using an *.edu.au domain.

Both carry the kind of material usually reserved for the Opinion pages of major newspapers, but reflect the fact that these are becoming mono-voice regions (follow the party line) and are declining in availability anyway.

They vary from the on-line "news" sources like Business Spectator, or Crikey in being exclusively opinion.

Anyhow, two items from On-Line opinion caught my attention, one questions whether markets can provide food security, while the other asserts that the world can't rely on alternative energy sources.

Ultimately both these items rest on the observation of the unpredictable variability of the weather. Food security can't be guaranteed by markets for the simple reason that the amount that is globally planned to be grown might not match the demand globally produced because of weather events. That problem exists, however, with or without international food trade. It just means that the impact of weather events is different and may not be felt in the location of the event, but elsewhere. A loss of production in a wealthy country results in more exports from a poorer country resulting in food shortages.

The energy case is equally built around the argument that the wind doesn't always blow and the sun doesn't always shine. In common with most such arguments it makes assumptions that existing storage technologies are as energy efficient and as price efficient as they'll ever get - they aren't of course. It also dismisses each storage solution because it alone could not handle the storage task, whereas in reality the combined use could.

But variability and distributions are indeed everywhere. Elsewhere today there have been reports about the float of LinkedIn and surprise about comments reported from their prospectus that "...a substantial majority of our members do not visit our website on a monthly basis, and a substantial majority of our page views are generated by a minority of our members." Indeed there is a small minority of users that make up the dominant usage. One could say .... "derr".

The question though is exactly what kinds of distributions they are. Lots of people have started to realise that the normal or Gaussian distribution is an abstraction that isn't particularly normal at all. So books like Black Swan and The Long Tail talk about other kinds of distributions. Unfortunately these too readily make assumptions about other kinds of distributions (notably Pareto or Power Law) ignoring that the small deviations from these that occur in things like log-normal distributions also happen in the hard to measure tail.

Meanwhile The Conversation had a learned article on Julia Gillard's voice and speech. It in part concludes that one of the difficulties is that the PM controls her speech too well, and it lacks sufficient variability to be engaging and convincing. Because the mode of speech is so unusual it is regarded as being artificial.

I've added these two to my daily reading list but I suspect I'm going to find more of these kinds of stories than things that provide any new insights.




Novae Meridianae Demetae Dexter delenda est

No comments: