I’m a big fan of the blog Stumbling and Mumbling, and there he comments on the decision of England coach Capello to field Robert Green in goal last night, the goalkeeper who conceded the poor equalising goal.
He emphasises a point I’m always keen to make to my students, that of comparing outcomes, or more accurately comparing one outcome based on no counter-factual. When I’ve lectured on the econometrics of this, the topic is given the rather odd name of treatment effects.
Phil McNulty at the BBC has come out against Capello rather strongly in my opinion after last night’s game, and rather unfairly in my opinion – but then I guess that’s journalism for you.
The charge is that Capello got it wrong; but did he? How are we to know Joe Hart or David James wouldn’t have made errors? We can’t re-run the match with the different permutations of line-ups and see which one would have done best – that just isn’t how the world works.
But, of course, this is how journalists have always worked. Bad science rules the roost in the newspapers and magazines: Ben Goldacre’s blog is a fantastic read to expose much of this.