Ronald Fisher, the statistician who popularised the p-value and the idea of ‘statistical significance’, worked out that complex traits must be massively polygenic – that is, must be related to many thousands of small-effect genes – as far back as 1918.63