James O'Neill's Blog

September 11, 2007

The scientific reason why social networks don’t do it for me.

Filed under: General musings — jamesone111 @ 7:04 pm

Somewhere in my school science lessons we were introduced to something that I know as the law of sheer cussedness  (Cussed, [Kússid] causing annoyance and anger, especially by being uncooperative). Basically if a chemical reaction which emits heat has reached equilibrium cooling it down makes it react more. The current that flows in a spinning dynamo wants to make it a motor spinning in the other direction… and so on. I thought it was Gay-Lussac’s law but it turns out that it wasn’t so I went to look up Scientific Laws named after people on Wikipedia. Skimming down by field I saw “Computer Science” against Bradford’s law. This was a new one on me. So I looked it up

Bradford’s law is a pattern first described by Samuel C. Bradford in 1934 that estimates the exponentially diminishing returns of extending a search for references in science journals. One formulation is that if we sort journals in a field by number of articles into three groups, each with about one-third of all articles, then the number of journals into each group will be proportional to 1:n:n². There are a number of related formulations of the principle.In economics this pattern is called a Pareto distribution.[Known sometimes as the 80:20 rule]

This explains so much. Why TV is rubbish for example, or why my supermarket carries 200 kinds of mineral water would be another. And why I know instinctively that facebook et al are of no interest to me.

Here’s how Wikipedia puts it.

Suppose that a researcher has 5 core scientific journals for their subject. Suppose that in a month there are 12 articles of interest in those journals. Suppose further that in order to find another dozen articles of interest, the researcher would have to go to 10 journals. Then that researcher’s Bradford multiplier bm is 2 (ie 10/5). For each new dozen articles, that researcher will need to look in bm times as many journals. After looking in 5, 10, 20, 40, … journals, most researchers quickly realize that there is little point in looking further.

Why my supermarket carries [many] brands of [something] . Everyone knows there are only 2 kinds of bottled water: with bubbles and without. If they only sold their own brand of locally bottled water they might sell to 100 customers. Introduce Perrier, Evian, Ty-Nant and 5 more to put 10 kinds on the shelf, and they sell to 200 customers. That’s a Bradford multiplier of 4. Want to sell to 300 customers ? They have to add 32 more kinds (that’s 42 in total). You see where this is heading…

Why TV is rubbish.  my favourite TV channel might serve up 4 hours of (what I consider to be) really good TV over the 30 hours or so I might watch during a week. The other 4 “main” UK channels might serve up another 4 hours between them. That gives me a Bradford multiplier of 4. Another 16 channels from cable/satellite/DTT might give me 4 more good hours. The amount of good TV doesn’t increase linearly with the amount of number of channels. The proportion of so-so and “crud” increase exponentially. Ever heard of Sturgeons law ? “Ninety Percent of everything is crud” – it was a defence of Sci-fi made by Theodore Sturgeon who accepted that “Ninety percent of science fiction is crud” but that’s because “ninety percent of everything is crud.”.  If you take a large enough sample, that’s probably true. Give me a another 64 and another 256 channels of TV so I have 1+4+16+64+256=341 channels and I only find 20 hours of good TV out of 10,000. 0.2% is good. 9.8% is so-so. 90% is crud. Maybe it’s not a 90:10 rule but an 80:20 rule (see Pareto)  and 19.8% is so-so but the principles are the same.

Blogs, mailing lists and social networks.  What’s true for TV works for other media. If my top ten blogs/lists/forums/other sites produce a dozen items worth the time it takes to read them each day, it might take another , it might take the next 20 to produce another dozen worthwhile items. The next dozen items might need me to read 40. Hugh Macleod coined Hugh’s law  “All online social networks eventually turn into a swampy mush of spam.”  Robert Scoble talked about it here 

Out of 1,048 items on my link blog in the past 30 days only 490 came from the top 35 blogs.
So, more than half of the value of that link blog came from the B, C, D … Z list of my 772 feeds.

So Scoble got 490 linkable posts from the first 35 blogs and 558 from the other 737. That’s close enough to saying 500 from the first 35 and the next 500 from the next 700 (less than one linkable post per feed/month). That gives him a Bradford multiplier of 20 ! If Bradford’s law holds he’d have to read 14,000 more blogs for a month to find another 500 decent posts

Of course anyone who writes a blog with any noticeable number of regular readers (I include myself here) would do well to remind themselves from time to time that those readers have decided the blog is part of the small percentage of stuff out there which isn’t crud. Thank you, whoever you are.

This post originally appeared on my technet blog.


Blog at WordPress.com.

%d bloggers like this: