Marcus du Sautoy examines a sample of dover sole from a day's catch, and by measuring the weight of this small number of fish, explores how the bell-curve of the Normal Distribution allows us to predict what the largest fish in the population is likely to weigh, even without catching it. The calculations are not shown in full detail, but his prediction is backed up by a conversation with a fisherman. Teacher notes: Use as a practical example when looking at standard deviation and measures of location / measures of spread. Would also work well as part of a Statistics course. For those who won’t be doing the actual calculation, students can still explore how the chance of getting a certain measurement changes with the number of standard deviations away from the mean. Students can collect continuous data of their own – for example, heights of students in a particular year – and see if it fits this model.