Stat Attack: Man City v AFCB

Written by Matt Stevenson

Here’s the xG from the City game. We all knew what a tough one this would be, but I’m heartened that in 4 of the other 6 games played on Saturday, a team had higher xG than City did. We didn’t create much but defended better than the scoreline suggested restricting City to 3 big chances. The first came in the 9th minute when Travers saved well from Foden (rated 1 in 4). The second was the opening goal, Haaland was slightly fortunate to get a second chance to play in Gundogan but reacted quickly and it was a good finish low to Travers’ left (1 in 3). Twelve minutes later, De Bruyne showed great individual skill to send Kelly the wrong way with a body swerve and picking out the far-left corner with the outside of his right boot from the edge of the box. On 37 minutes, the game was effectively over as a contest when Foden scored, possibly with a deflection off Kelly making it a harder save for Travers (1 in 6). Before then, Pearson had a shot from the edge of the box (rated 5%) and Smith had an effort (1 in 10) which City broke from for De Bruyne’s goal. Foden’s goal was also after an AFCB attack, showing the problems that will be faced by the weaker teams when falling behind against Man City. This may have shaped the second half as AFCB created little and City was content being comfortable. The only goal was unfortunate, with the ball hitting Lerma and ricocheting into the net. Infogol doesn’t handle own goals well, I’ve derived the probability (1 in 5) from the difference in the total xG and that from the sum of individual shots.

In terms of average positions unsurprisingly we were less advanced than against Villa. One tactical change was that Lerma played in the middle of the back three and Mepham on the right, which reversed the line-up at Villa. Given how little Haaland was in the game this may well have been effective.

What caught my eye is the clumping of players: Stacey and Kelly; Smith with Cook/ Billing, plus Tavernier; and Pearson and Christie. I’ve tried to add more detail by showing subs before 88 mins. Orange denotes substituted and grey denotes being brought on. Billing for Cook (64 mins) was a like-for-like swap, whilst Stanislas played further forward than Christie (63 mins). Senesi’s debut allowed Lerma to move back into midfield, although Tavernier was withdrawn (80 mins). Tavernier was the more forward midfielder, which could explain why Stanislas was more advanced than Christie. Lowe came on for Moore in the 90th minute but didn’t touch the ball.

Defenders dominated the number of touches, with Lerma leading with 60. The exception was Stacey, who was noticeably below the others, possibly because we played it less through him as he was out of position.

For the second consecutive week, our central defender in the back 3 had the most passes. Mepham was second and Pearson third. Special mention to Tavernier who had most passes in the opposition half and who had a 100% completion rate with Pearson and Lerma also recording over 90%.

City very much play the ball on the ground meaning the main aerial battles were when we hit it long to Moore. Only Smith apart from Moore had more than one duel. I suspect this may be the only game where Cook has a 100% success rate on aerial duels.

We made fewer tackles than against Villa probably showcasing Man City’s ability to pass the ball accurately.

City’s slick passing meant that there were few interceptions. There were more recoveries, with Smith, Christie and Tavernier leading the way amongst outfield players. Unsurprisingly most clearances came from the defenders. There weren’t many blocks, again possibly indicating City worked the ball for better chances.