GP is being hyperbolic. He's saying that, "Just because you can't guess the trend in the data (eg. the track of a soccer ball in yesterday's match) doesn't mea

GP is being hyperbolic. He's saying that, "Just because you can't guess the trend in the data (eg. the track of a soccer ball in yesterday's match) doesn't mean that something is wrong with the data; you often need to double-check both your hypothesis and your data. And in this case your hypothesis was probably at fault due to inherent bias rather than inaccuracies in the data."