When August ended, Arsenal were bottom of more than one table. They propped up the one that mattered, below even Norwich by virtue of scoring fewer goals - none in fact.
But the still uglier standings came in terms of expected goals against (xGA). After three games, Arsenal’s was already 8.81. They had faced 20 shots on target.
By every marker, Arsenal’s defence has done dramatically better in the subsequent six matches. After conceding 10 goals in three games, they have let in four in six. Their xGA in September and October is a respectable 6.22.
It is bettered only by Manchester City. They have gone from 20th to second in that respect.
In terms of goals conceded, they have gone from worst to third, with only City and Chelsea proving more frugal.
![Arsenal xG](/images/news/690x388/7b3c04fa-a0d1-44c8-9377-4c536510909a.jpg)
Arsenal have now faced 40 shots on target this season: 20 in August, 20 since.
One explanation is that they faced two of the top three teams in August and they have encountered four of the seven lowest scorers since then, but it also reflects the transformation of their defence.
Personnel change behind defensive improvement
Aaron Ramsdale, Takehiro Tomiyasu and Gabriel Magalhaes did not feature in the three August games (the Japanese had not even been signed) and have been ever-presents since.
Ben White began only one of the opening three defeats but has played in each of a six-game unbeaten run.
Only Kieran Tierney has really straddled the bad and the good, while Cedric Soares, Rob Holding, Calum Chambers, Pablo Mari, Sead Kolasinac and Bernd Leno have only started in losses.
![Ramsdale v Leno](/images/news/945x532/5b942bfb-0161-484d-8b15-bdc1dda90076.jpg)
The goalkeeping change has made a major difference.
Ramsdale’s save percentage of 80% puts him third in the league, behind only Edouard Mendy and Jose Sa.
Leno’s ratio of saves – 50% – leaves him second from bottom, ahead of just Freddie Woodman, who has also been dropped.
Ramsdale’s clean sheet percentage of 50% is bettered only by Mendy, Alisson and Ederson; Leno is one of five keepers on 0% and the other four all play for Newcastle or Watford.
Tomiyasu has added a different dimension. He has won 22 aerial duels already.
The only defenders to win more have all spent more time on the pitch and, even though he did not feature in the first three matches, no right-back has won more at all. It means he is a right-back with a centre-back’s skillset.
He has already made more tackles, interceptions and clearances than Tierney, a more attack-minded player, has on the other flank. The difference in their duties is apparent in the attacking numbers: the Scot has completed nine passes into the penalty area, the Japanese just one.
Understandably, it makes sense to look towards the manager when a team is beginning to produce more consistent performances.
What have they changed, is their philosophy finally beginning to bear fruit?
Mikel Arteta would like to think so, and it may very well be true. But for now at least, Arsenal simply have a better team than they did at the start of the season because it contains better players.
But the system remains the same.
Leicester's attacking revolution
As for Leicester, we can look squarely at the manager for their reason behind their uptick, and the Foxes may very well present the stiffest test of Arteta’s new defence when the teams meet at Saturday lunchtime.
Brendan Rodgers’ side are the best opponent Arsenal will have faced since Manchester City, in terms of both last season's finishing position (fifth), and goals scored and xG created in the period since that 5-0 thrashing on August 28.
![Leicester post-international](/images/news/945x532/cb59f640-e1cf-49fd-8de5-df8e15f458e1.jpg)
Rodgers switched to a 3-4-1-2 formation during the most recent international break, and in the two subsequent matches his team have been transformed.
Leicester are averaging 16 shots per game since changing shape, compared to 10.9 before, and have been rewarded with six league goals.
Those six goals have contained four outstanding chances, each with an xG of at least 0.46. Arsenal have only conceded one such clear chance, to Aston Villa’s Ollie Watkins, in their last six matches.
Rodgers' tweak to the system appears to have freed up his two most creative midfielders.
Youri Tielemans has averaged 5.50 shot-creating actions over the last two games, compared to 2.57 before then, while James Maddison has had more shots in the last two games (six) than in the seven prior (five).
🚀 Youri Tielemans with the most 'Unexpected Goal' of the Premier League weekend.
— Sporting Life Football & Infogol (@InfogolApp) October 25, 2021
📊 Our xG model gave the Leicester star just a 3% chance of scoring this belter...@LCFC #LCFC pic.twitter.com/XiygVLh5Ei
Fielding a system with two strikers has also made it easier for Rodgers to get Patson Daka on the pitch.
The Zambian only played 27 minutes in the first seven league games but has been an influential substitute, with a goal and an assist, in the last two - on top of scoring four goals away at Spartak Moscow in the Europa League.
It may be a shape that gets Jamie Vardy on the ball less: he had only six touches in 45 minutes at Brentford and 19 in 90 against Manchester United, but it is a formation that has worked.
Leicester only won two of seven Premier League games this season with a back four. They have triumphed in both with a back three.
Saturday will see a meeting of two sides, and two managers, who seem to have overcome their false starts to the season in very different ways. Change the system to suit the players or change the players to suit the system?
Really, does it matter if you find the right formula?