Hull City 2-0 Manchester United: 21 shots, no goals

Manchester United played 654 passes away from home, held the ball for 72% of the match and took twenty-one shots. They scored none. Hull City replied with eight shots, 252 passes and two goals, both before the interval, both from defenders. The final 2-0 is accurate; the numbers behind it explain rather more.

Two defenders and a single corner

Ajayi opened the scoring on seventeen minutes: a centre-back, wearing number six, starting the match. Mendy doubled the lead on thirty-eight, another defender. Hull scored nothing else, and the rest of their numbers suggest they were not chasing much more - one corner across ninety minutes, no offsides, eight shots in total.

Jakirovic withdrew both scorers at the same moment, sending on McNair for Ajayi and Herrington for Mendy on sixty-four. By then the hosts had spent forty minutes in a 5-4-1 with barely any of the ball, and the double change looked like fresh legs for the line carrying the load.

Twenty-one shots, eight of which never arrived

Eight of United's twenty-one attempts were blocked - thirty-eight percent. Close to one shot in three was struck with a defender still between the ball and the goal. The remaining thirteen reached the keeper's end: five on target, eight wide or over. Tzolakis made five saves and kept the clean sheet.

Fifteen of United's shots came from inside the box. That volume normally converts into two or three goals. Here it produced 1.81 expected goals and none at all.

Possession flattered nobody

A 72-28 split usually means one side played while the other defended. Divide passes by shots, though, and the gap disappears: United needed roughly 31 passes per shot, Hull roughly 31 as well.

Control of the ball gave the visitors more attempts for the simple reason that they had more of the ball. It did not make them better at turning possession into shots: Hull spent the same number of passes on each of their attempts, they just had two and a half times fewer of them.

Chance quality, not chance volume

Hull's average shot was worth 0.16 expected goals. United's was worth 0.09. The hosts shot about half as often, but each time from a position that goes in noticeably more often.

Hence the closing gap: 1.26 xG for the side that scored twice, 1.81 for the side that did not score at all. Hull finished 0.74 goals ahead of their model, United 1.81 behind theirs - two and a half goals of combined deviation inside one match.

What Carrick changed

Rashford came on straight after the break, at 0-2. On sixty-seven Carrick introduced Mainoo and Sesko together, removing Tielemans and Andrey Santos, changing both midfield and attack at once. Dalot and Lacey followed by the eightieth.

Five substitutions, three of them attacking, and no goal across the forty-five minutes after the interval. This is the case where refreshing the personnel did not change the nature of the chances: the shots kept arriving from the same positions where they had been blocked in the first half.

What to take from the match

The clubs had not met since 1 February 2017, a goalless draw. Across their last four meetings United have two wins, Hull one, with one draw.

For Hull this was not pure defensive endurance. Eight shots and 1.26 xG on 28% possession is respectable output going forward: they picked their moments rather than sitting back entirely, and carried almost all of them into the box.

For United the reading is harsher. The issue is not the volume created - 1.81 xG away from home is a normal return, one that usually collects points. The issue is that a third of the shots never reached the goal, and those that did were struck from positions worth roughly a one-in-twelve chance. Against five defenders in a line, that did not work once in ninety minutes.