The Last Man Is the First Story

Finishing is the most honest act in sport. You cannot argue with a ball that has crossed the line. And yet, at the Institute of Applied Finishing, our entire discipline is the study of the shot that should not have gone in.

This month gave us two. A striker who won the Golden Boot from a sofa in Essex, and a goalkeeper who left his post in the ninety-seventh minute to settle a World Cup final with his head. Our software filed, in the second case, a resignation letter. I have chosen not to accept it.

The lesson is not that models are useless. The lesson is that finishing lives in the exact place models fear — the moment when a professional, against all incentive and instruction, decides to try the impossible thing anyway. We can measure everything about that moment except the only part that matters.

So we keep our expected-goals maps, and we keep a second file, unlabelled, for the goals that expected nothing. It is, quietly, the larger file. In Timeline B it is the only one.