I CONSIDERED myself to be reasonably proficient in the application of probability and statistics until I discovered Stein's Paradox in a 1970s journal. This was couched as follows: everything is connected and correlations between apparently unrelated events can generally be found. This came as a surprise because the theory as taught mostly assumes (often conveniently) the reverse to be true. In fact, for most practical purposes, correlations are so weak they can be ignored. However, Stein demonstrated that car production in Detroit and the Lions football results were related. Similarly, the stock prices can be linked to the Olympic athletics results, and so on. How come?

Imagine being a production-line worker and your favourite football team wins 4:0 against the league champions. The next day you and your co-workers go into work in a positive mood. The production line is a buzz of excitement reliving the match, work takes no effort, and output is well above average. It's as simple as that. We are influenced positively or negatively by events visible to a significant number of people.

Stein's Paradox recently came to mind again at the end of a bad technology day. It started when I discovered my mobile phone battery was down to 10 per cent charge because it had not been seated correctly in the holster. My train was late, and during the journey I discovered a new software bug on my laptop. It was raining in London and I had to wait for a cab, which meant I was late for my first meeting. I managed to find someone with a mobile phone charger but, while our phones looked identical, my "rescuer" had a newer model with a smaller diameter power jack. So I had to resort to a fixed connection, only to find the server was down.

I tried to call my secretary, but someone had put a back hoe digger through the cable feeding my office. To top it all, my wife called to say the washing machine and vacuum cleaner were broken. And all this before midday.

Quantum theory suggests all matter is clustered. But it also seems to be true of events. Car accidents, having children, deaths in families and technology failures seem to support this view. Perhaps there is a quantum theory of life. When I have a bad technology day, I find that careful examination tends to put me, or other people, in the causal driving seat. An equipment failure can make us tense and prone to error, which leads to mistakes that promote more failures. This catastrophic feedback I can understand and quantify approximately, and it is empathic with Stein's Paradox.

But I suspect other mechanisms are also at work. Buying all our domestic appliances when we get first get married and changing vehicles and computers at regular intervals all invoke a rhythm of failure. Most technology is designed for a mean time between failure of about seven years. So this is an identifiable mechanism for synchronism.

What is puzzling is how diverse, geographically dispersed and unrelated technologies conspire to create a nightmare of a day about twice a year. This is Stein with a vengeance. Surely the machines are not empathic, are they? I think not, but soon, when they are truly intelligent and online, they may be.

Peter Cochrane holds the Collier Chair for the Public Understanding of Science & Technology at the University of Bristol. His home page is:
http://cochrane.org.uk