Most squares

Sometimes you try really hard to ask interesting questions, so that people think that you're clever and interesting too, that you actually don't realise how dumb you're being just for asking those very same questions.

For example, a few years ago, Marta's then boss invited us to see her daughter performing with her string quartet. As we sat down, waiting for the concert to start, only half (well, maybe I should say only $\frac{1}{n}-$th thinking), I asked how many people were playing in the quartet $-$ of course, just as I was speaking, I immediately realised how stupid the question was, but sadly it was too late to take it back...

I thought of this glorious moment again the other day, during one of the PSMR course lunches. I was having a chat with the other lecturers and we were commenting that this year's audience were quite good (according to the feedback forms, the feeling was mutual). 

Of course this is not always the case: Gareth, for example, remembered that a few years back he was teaching linear regression, explaining that the least square line minimises the overall difference between an observed value and the fitted value provided by a model. At that point, one person in the audience raised their hand and with a straight face asked: "why not taking the maximum?"