That start owes much to the calibre of the finishing. All of Spurs’ six scorers this season are outperforming their xG and Heung-Min Son is not merely the joint top scorer in the division; he is its greatest expected goals overperformer with eight from an xG of just 3.50. Unsurprisingly, therefore, Harry Kane has the greatest difference between his expected assists (3.80) and his actual total (eight).

Nearly all of that dates back to the first five games. In the last three, Kane’s xA is 0.59 and Son’s xG 0.94. In that subsequent period, each is averaging fewer shots per 90 minutes and fewer key passes. But the shift to a more defensive approach is shown by the reality Spurs have only had five efforts by defenders - four from Sergio Reguilon and one from Toby Alderweireld – in those three matches. They had 12 from defenders before then. Attacking has been left to fewer players.

Son’s numbers look unsustainable in one respect. The South Korean has eight goals from just 11 shots on target so far, and few end a season scoring from 72 percent of their efforts on target, just as Son, who has averaged between 37 and 48 percent of his efforts on target, is unlikely to carry on having 61 percent on target. In another, however, they are not.

Son is such a fine finisher he has ‘beaten’ his expected goals in each of the last four seasons, and often by sizeable margins: by 3.64 in 2018-19 and 6.23 in 2016-17. Across his Premier League, he has outperformed his xG with his right foot, his left foot and his head. And as he is actually averaging fewer shots per game this season, there is the scope for him to score more.