The numbers that show Arsenal are better than last season

Declan Rice celebrates during Arsenal's Premier League win over Liverpool

Can a team have improved when they are two places lower in the Premier League table than the equivalent stage of the previous campaign, with fewer points and fewer goals scored? It sounds implausible but evidence is mounting that it is true of Mikel Arteta’s Arsenal.

The suggestion would have been ridiculed a month ago, but should Arsenal maintain their statistical dominance of matches and add a string of wins to last weekend’s victory over Liverpool, it will start to be taken seriously.

Neutrals have found Arsenal less enthralling to watch, even boring, while the club’s supporters found themselves tense, agitated and unsure of the team’s tactical direction after last season’s surprise title challenge. A fixture in the team, Granit Xhaka, was replaced by Kai Havertz. Crowd favourite Aaron Ramsdale was replaced by David Raya in goal. Some of these emotions were impulsive reactions to changing expectations, but some anxieties reflected very real teething problems on the pitch.

There was more to last season’s Arsenal than the memorable 14 matches before the Qatar World Cup, of which they won 12. Arsenal won 50 points from the season’s first 19 games, the best halfway tally in the club’s history, but just 34 points from the remaining 19 games. Despite both halves of the season carrying equal weight, Arteta’s team this season have found themselves compared to the dashing Arsenal of pre-February 2023, rather than the more skittish version that followed.

After the 23 games of last season, which encompassed Arsenal’s very best football and form, they had 54 points with 51 goals scored and 23 conceded. This season, Arsenal have 49 points with 47 goals scored and 22 conceded. However, their attacking output judged on expected goals is almost identical, while their underlying defensive numbers have improved. Aside from points, Arsenal did not lead the league in any of these headline metrics last season, always trailing Manchester City, whereas this season they lead or co-lead the league for goals conceded, expected goals against and expected goal difference. Points won is of course the most important measure, but might not always be the best predictor of what is to follow.

Looking at equivalent fixtures rather than games played, the difference between last season’s Arsenal and this season’s Arsenal is clear: shot conversion. From the equivalent fixtures, minus the three involving the promoted clubs, Arsenal scored 49 goals from just 37.9 expected goals last season. Players such as Martin Odegaard and Gabriel Martinelli enjoyed hot finishing spells. In the same games this season, Arsenal have scored just 35 goals from 38.1 expected goals. Their expected goals conceded, and therefore their expected goal difference, is better this season than last season in these fixtures. If Arsenal want to win the league, though, they need more positive variance in front of goal.

Arsenal’s defensive record has been the most impressive facet of their play, and they were expected to be sturdier and less vulnerable after adding Declan Rice. Their ability to suffocate opponents means Arteta’s team look far more stable in the biggest games.

While a league season is about harvesting points against every type of opponent, this bodes well for a run-in which includes trips to City, Brighton, Tottenham and Manchester United as well as potential Champions League ties. Arsenal conceded just one expected goal at Anfield in December, having conceded 8.2 expected goals across their previous two league trips there. In league meetings with Liverpool and City at the Emirates, Arsenal have faced less than one expected goal combined. City had four shots, the fewest shots a Pep Guardiola side has had in a top-flight game since April 2010.

Prior to the Liverpool game at the Emirates, Arsenal maybe lacked some attacking threat of their own against the best. However, these are fixtures where living on fine margins is less of a problem. It could prove effective in Europe, with knockout competitions rewarding teams who can avoid defeat.

If Arsenal are better, why has it not felt that way?

Game state

Arsenal’s more successful shot conversion last season put them in the ascendancy more frequently, and did so earlier in matches. Naturally, fans are more relaxed once their team is in front, especially in the case of Arsenal, who face teams looking to stifle them.

After 23 Premier League games last season, Arsenal had scored 11 goals inside the first 20 minutes. This season it is just eight, with three of those goals coming in Arsenal’s last four matches.

Across the whole of last season, Arsenal spent 47 per cent of Premier League minutes leading, and in 24 per cent of total minutes they were two or more goals in front. This season, Arsenal have spent just 35 per cent of Premier League minutes leading, and have been two goals or more in front just 16 per cent of the time.

The majority of Arsenal’s expected goals have accumulated with the scores level, which has been the case in 48 per cent of minutes compared to 37 per cent last year.

So while Arsenal have been racking up excellent numbers on the stats sheet, they have done so in game states that keep fans nervous. Due to finishing variance or other factors, Arteta’s team have been less effective at ripping games open early.

A more functional in-possession style

One possible reason why Arsenal are spending more time level on scores is their deliberate approach in possession. Whether this has been a tactical shift from Arteta or a natural consequence of replacing Thomas Partey with Rice at the hub of the team is up for debate.

Arsenal’s defensive numbers have improved because they are conceding fewer transitions and counter-attacking opportunities, and when they do concede them, Rice is there to throw a damp cloth over the flames. The trade-off is that opposing teams have longer to set in their defensive shape.

Arsenal have been more wary of forcing passes through the middle of the pitch, something which led to goals conceded in pivotal matches against West Ham and Southampton last April. Partey likes to receive the ball in congested areas, even with back to goal, trusting himself to wriggle out of the pressure. Rice prefers to move to the game’s extremities and see the game ahead of him. It is a stylistic difference of which Arteta will be well aware.

Last season, Partey was playing 4.45 progressive passes per 90 minutes whereas Rice is playing 3.52 per 90 minutes this season. The Englishman is a crisp progressor of the ball and his numbers are generally strong, but he does so with less disguise on his passes than Partey.

The result is Arsenal have not sliced through the heart of teams quite as much, leaving some fans feeling the football is staid. According to StatsBomb data released on X, formerly known as Twitter, Arsenal have just two players averaging more than six ‘line-breaking passes’ per 90 minutes in the Premier League this season (Rice and Oleksandr Zinchenko). Man City and Liverpool each have four players doing so. Rodri completes roughly double the number of Rice.

Declan Rice tackles Alexis Mac Allister during Arsenal's win over Liverpool
Declan Rice has made Arsenal more stable...but possibly less incisive from central positions - Getty Images/Shaun Botterill

Set-piece reliance

Only Chelsea have scored more penalties than Arsenal’s six this season, while Arteta’s team have scored 10 goals from corners which is more than any team. This has led to concerns that Arsenal are too reliant on set-pieces and not creating enough from open play. Even after a 5-0 win over Crystal Palace, there was fretting about the fact the first two goals had come from Gabriel headers. Had Arsenal’s goalscoring concerns really been abated?

Arsenal still trail Liverpool and, believe it or not, Newcastle United when it comes to non-penalty expected goals, so there is room for improvement before they can claim to have the league’s best attack as well as defence.

The hinge moment these concerns miss...

Some of the scepticism about Arsenal’s attack appears based on a stodgy first few months, rather than what they are currently producing.

Up to and including a 1-0 win at Brentford in late November, Arsenal ranked 12th for open-play expected goals generated, behind teams such as Wolves and Everton. This was, unquestionably, a problem.

In the 10 Premier League matches since however, Arsenal rank first for open-play expected goals generated, and that period includes two matches against Liverpool, a trip to Villa Park and fixtures Arsenal lost last season such as Brighton at home and Nottingham Forest away.

What changed? The form and fitness of Martin Odegaard could well be linked to the upturn, with the Arsenal captain having taken on a far more comprehensive midfield role since he got over a hip niggle.

In the first nine league games of the season, Odegaard was operating more like a second striker. His shot and expected-goals numbers have decreased since, but his creative statistics have gone through the roof.

In the early part of the season, Odegaard was averaging just one and 1.6 touches per 90 in the two central zones either side of the halfway line. Since, he is averaging more than five touches per game in both zones, as well as far more touches on the left side of the pitch.

Before that trip to Brentford, Odegaard was averaging 35.9 successful passes per 90 minutes, just 0.13 expected assists and only 1.7 chances created. Since then, Odegaard has averaged 55 successful passes per 90 minutes, 0.41 expected assists and is creating 3.6 chances. This is a transformation not a marginal gain.

Should Arsenal’s chance-creation trajectory continue, then the debate over whether this season’s team is better than last could soon be settled.