I was saying that the 210-age formula may be the best fit formula for the average (i.e. not very fit) person in the UK, and that cyclists tend to be fitter (and that casual observations make me think that cyclists have higher maximum heart rates),

I missed this bit, max heart rate is not dependent on fitness or sport, it cannot be improved by exercise and it decreases .5 to 1 beat per year but this also varies.

And I wasn't saying it can be. It's pre-selection. I'm suggesting that people with higher maximum heart rates tend to cycle (and, in general, be more physically active) than people with lower maximum heart rates. This is based on nothing else than observation of people posting their max HRs and ages on cycling forums.

Why would anyone interested in their maximum heart rate for training purposes use any formula as there will always be an error of some sort.

Because it gives you an initial idea in the absence of any proper information. Lots of people don't have an HRM and check their pulse with their fingers. You can't get an accurate HR

_{max} measurement measuring your pulse with your fingers.

Both 210 and 235 give figures more than 5 bpm out for me which could put me in the wrong training zone and I expect that would be the same for most people.

But 210-age has been proven by various studies to be the best fit of any equation of the form '

*x*-age' for the general population. Higher order polynomials will provide an even better fit but the minute you put a little

^{2} above one term then most people's eyes glazeth over.