One issue that comes up a fair amount in analytical chemistry (of which I’ve done a fair bit over the years) is the concept of signal to noise ratios. The electrical engineers amongst us won’t need that explained, but basically if you have a very weak signal, then it can be obscured by background random signals. So you’re better off having a strong signal, rather than amplifying a weak one. Where it all gets tricky is when you start looking at single molecule detection - which is possible, but is kind of tricky (think Avogadros number of noise and you’ll get the picture.

Anyway, this post is really about sites related to my weblog. I tried this after reading my server logs (again) and noticing that the urbex.org.uk guestbook was ‘related’ to my front page. Which is a little strange, since I’ve never linked to that page (before now) and it hasn’t been linked to me - moreover, it’s my weblog that repeatedly links to urbex’s log. Anyway, the links seem solid as concrete if you try to work out how my weblog is related to "Cricket talks fail to break deadlock", a BBC News Politics article, or even an open source object-oriented extension to the Prolog programming language. Now if I were google, I would try and explain it away as signal to noise problems, rather than admitting that their idea of ‘related’ doesn’t necessarily imply that there is some sort of common link between things. Hmm.

(Bear in mind that the google links will change over time, so unless you look at it today, it might be completely different.)