Tuesday, 11 December 2007 - 10:38 AM GMT

Not quite. A certain percentage from each age range will die per year through accidents and suicide as well as murder. The set of teenagers will, on average, be less than 16.

Is it heartless for me to point out your error?

Tuesday, 11 December 2007 - 1:45 PM GMT

There are seven 'teen' years. Therefore the average teenager is 16 and a half. The deaths you mention will only reduce the average by a week or so, so the average teenager is still '16'.

Tuesday, 11 December 2007 - 2:21 PM GMT

I'm not sure it's obvious that you'd necessarily expect likelihood of being murdered as a teenager to be independent of age is it?

Tuesday, 11 December 2007 - 5:00 PM GMT

Perhaps not, Juliette. But I don't think that the Standard would have led with the heading: "Average age of teenager murder victims in London equal to average age of teenagers not murdered' or 'age not a factor in killings'. I suspect instead that they gave no thought to whether their 'finding' meant anything at all. They saw a number and thought "there has to be a story in here somewhere".

I recall that one of the tabloids ran with a 'shock' story about a particular method of contraception. Apparently every year five per cent of women who used it got pregnant despite the claim that it was 94 per cent effective! It didn't seem to occur to them that this was in fact a good thing, since the expectation would be for six per cent to get pregnant.

The most interesting thing in the Standard article is that murders have gone down in London despite the increase in the ease with which people are able to get guns. That deserves to be the angle if you ask me. Is it an 'outlier' result or could it even be the case that murder has fallen because more people have guns! I'm not sure I'd go that far, but it looks like a case could be made that gun availability is not a significant factor in the London murder rate.