Week 27 has come and gone and Man City have no less of a grip on 1st place than before. They are looking to set all kinds of records if they continue on this pace and when you read the reports all that the punditocracy can talk about is how good they are going forward. They are the best attacking team I have ever seen play football.
Matchday 27 against Leicester City, the team who won the Premier League just two years ago, Man City started without six of their first team players: Mendy, Jesus, Delph, Sane, Stones, and Silva. And they still scored five goals.
But what was most impressive against Leicester City wasn’t that they scored five goals, it’s that they kept Leicester to just two shots! Vardy scored one but Leicester really never threatened Man City. It was a combination of possession, Man City dominated 70-30, but Leicester have been in those kinds of games before and are experts at getting out on the counter and getting shots.
That defensive acumen is what made me predict that they would win the League this year. That was before any summer spending. City didn’t just control possession, they killed off their opponent’s chances to score when they did win the ball back.
Here, possession is just passing. Ok? And City average 727 passes per game. That’s a lot. But Arsenal are 2nd in the League this year in possession/passing and they average 635 passes per game. 66.5% possession for City, 58.8% for Arsenal. But what’s incredible about City is that they have allowed the fewest Big Chances against, just 23. While Arsenal have more than double that number with 48 Big Chances allowed.
City is on the way to break a record for goals scored but they have also conceded the fewest goals, the fewest shots, the fewest big chances, and the fewest shots in prime. That’s why they are walking away with this Premier League title.
Here is the data for the League through week 27 (click to embiggen). I’ve sorted it by Big Chances allowed (big chances that the other team took).
Arsenal still create plenty but as you can see they are nearly as terrible at defending as Crystal Palace, Swansea, and Bournemouth.
As for conversion numbers (click to embiggen):
Arsenal are the 6th worst team in the League in defensive conversion rates. Let’s say that 10% is the baseline* – Burnley and Man U are massively overperforming; Watford, Huddersfield, and Liverpool are underperforming; and then Arsenal are in the sort of second tier of underperformers. It might not seem like much but if Arsenal were just average in defense they would drop from 36 goals conceded to 29. That’s just 7 goals and again doesn’t seem like much but that’s a draw against Stoke, maybe a win over Watford, Bournemouth, and Swansea.
Arsenal’s inflated defensive conversion rates are almost certainly a product of the fact that Arsenal have conceded so many big chances and that Arsenal have the League leading sloppiest defense – conceding 11 goals off 24 errors. Cech leads all players in goals off errors with 4, Bellerin leads all defenders with errors with 4 (only 1 goal conceded), and Granit Xhaka is third in the League in errors (3) but is joint leader of the League in goals from errors with 2.
A 180 degree turnaround in defending is needed to get Arsenal back into the top four. It’s possible (anything is possible!) because typically a lot of teams head into these last 10 games of the season fairly certain of their place in the League and they tend to take their foot off the gas a bit. The problem is that this season seems a bit different because any of the teams from 10th place Bournemouth to 16th place Huddersfield could slip into the relegation zone.
Arsenal didn’t buy any players to address the problems with defense; Arsene Wenger could change Arsenal’s shape and approach to games and maybe make Arsenal more defensive but that seems highly unlikely and difficult to do with 10 games left. So, my prediction for the next five matches is more of the same from Arsenal – sloppy defending, leading to many big chances conceded, leading to frustration and a 6th place finish.
But the difference between Arsenal’s possession and defending and Man City’s possession and defending does show that it’s not all about possession for Pep – he is also getting his team to defend together as a unit, to kill off counters, and preventing the opposition any real chances at goal.
*10% is kind of stats nerd shorthand for average scoring rates. It varies from season to season but typically hangs out around 0.095-0.1.
Sources: Opta for the raw data and my database for everything else