Jürgen Klopp’s side coasted to the title yet were the league’s second-best team according to the popular metric. Are
Liverpool baffling the boffins – or are we missing the point?
There was just the faintest whiff of schadenfreude about it. A table had been published showing how the
Premier League would have finished if it were decided by expected goals. “Look,” a colleague said, “according to xG
Manchester City should have won the title by 13 points.” In fact, over the past two seasons, Liverpool should have had 39 points fewer. It pained the colleague to say it, but might Jürgen Klopp have exposed the limits of nerdery?
Seeking to show whether a given shot is a good chance, whether it should have gone in, xG is one of many metrics used in
Football analytics. It is, however, the first to move into the mainstream. It is featured (after the corner count) on Match of the Day, it pops up in analysis articles and is widespread on
Twitter. It has also become a focal point in the debate about how far the magic of football can be distilled into numerical form. So when an xG league table looks so far off the real thing, people start asking questions.