Contact us for a free demo Contact us
for a free demo
StatsBombIQ StatsBomb Data

Arsenal 2017-18 Season Preview

By Euan Dewar | August 6, 2017 | Analytics


Friends, it’s that time of year again. Le Professeur Arsène Wenger is back for his 22nd(!) season as manager of Arsenal. The 2016/17 campaign ended with a strange cocktail of feelings as another FA Cup win was shaded by missing out on the top-4 for the first time in Wenger's tenure. Despite this the club has renewed his contract through 2019, giving him another shot at winning the league title that has eluded him for going on 13 years now.

Arsenal’s failure to win the Premier League in an open 2015/16 season brought much consternation but a look under the hood showed positive signs. Their attack was humming to a quite impressive extent, producing some of the best expected goals per shot numbers the league has seen in recent seasons. 2016/17, by contrast, was a dysfunctional downturn.

Shot volume stayed around the same but the league-leading, almost Barca-esque expected goals numbers in attack dropped off from 0.131 per shot in 15/16 to 0.115 in 16/17. In defence they allowed a handful of fewer shots but the quality of them sharply rose, going from 0.081 to 0.100. 

Arsenal's attack in 15/16 vs 16/17.

The problems in attack - usually Arsenal's bread and butter - are particularly worrying, with no one clear answer as to their origin. Santi Cazorla's absence due to injury will have hurt as he played fewer league minutes than in 15/16 (619 minutes vs 1292) but he still wasn't ever-present back then either. Aaron Ramsey similarly played around 1400 fewer minutes. New signing Granit Xhaka sopped up most of their minutes and, while he has his strengths, he is not the same sort of fit. Cazorla posts between 2.3-2.6 key passes per 90 when he plays, whereas Xhaka has never breached 1.3 in his career. Mesut Özil also didn't have the same kind of barnstorming creative season, with his 4.3 key passes per90/19 assists in 15/16 dropping to 3.2 KP/9 assists in 16/17.

Özil's dropoff came alongside Wenger's experiments with the striker situation. In the 15/16 season Özil to Olivier Giroud was a very effective combo - Özil assisted Giroud 7 times overall, the joint top assist to goalscorer tally in the league, with the two of them producing some excellent shots in central locations. The frenchman's ability to get up for good headers is dream for someone with Özil's ball-playing abilities.

In 16/17 Wenger elected to make Alexis Sánchez the main striker, a move that made some sense on paper but in reality had unintended results. Not only did it put a dampener on Özil's partnership with Giroud it also neutered the best shots Özil was able to create for Sanchez, as well as having similarly odd effects on the team's output.

The 3-4-3

Between the 31st of January and the 10th of April Arsenal went on a damaging run of form. They racked up 7 losses, 1 draw and 4 wins in all competitions, with 2 of those wins being in the FA Cup against non-league teams. That run included two 5-1 thumping at the hands of Bayern and deserved losses against Watford, West Brom and Crystal Palace. Seemingly as a response to this (and likely influenced by Antonio Conte’s success) Wenger debuted a change to a 3-4-3 formation away at Middlesbrough. From there until the end of the season they lost just 1 and won 9 including, of course, winning another FA Cup. A stroke of genius?

Well, a glimpse below the surface suggests otherwise. In the 8 league games they played with the 3-4-3 they scored 15 goals from open play and only conceded 3. However expected goals numbers over that timeframe suggest they should have scored 14 and conceded 10. Even if you ignore this slice of variance, the base xG values here in both attack and defence just aren’t very good to begin with. Their expected goal difference over this final stretch of the season was only 7th best in the league. Clearly the results masked some serious teething problems.

None of this is to say the 3-4-3 will never work. They’ve continued to use it in pre-season and perhaps a full summer of training in it will lead to a more cohesive version in 2017/18. It just isn’t the panacea that it might seem to be at face value from such a small sample size. 

Although they finished 5th last season their final points tally was actually higher than it was when they finished 2nd in 15/16 (71 points to 75) but that is obfuscated by a pronounced underperformance of xG in that season and an overperformance in 16/17. Even if you take the points totals at face value it shows that the Premier League’s top 6 is stacked with more high-level player/coaching talent than ever and Arsenal don't have the room any more for issues like these. For the first time in years they are not just struggling to compete for the title, they now find themselves on the lower end of the upper class. Managerial change is clearly not on the table, so gains need to be made elsewhere.


Striker is perhaps the only position that has provided Arsenal supporters with as much angst as Defensive Midfield in recent seasons. After many half-answers the club have brought in a more definitive solution in the form of Alexandre Lacazette. The Frenchman has been tearing it up in Ligue 1 for some time now, currently hot off of scoring 18 open play goals (+10 penalties) and laying on 3 assists. A good tally, even with the penalties removed and especially so when you consider he only played 2408 minutes due to injury (Sanchez with his 22 non-pen goals played 3224 minutes, for comparison).


Last season he dropped off somewhat in terms of high-value shots in central positions with a lot more coming just wide of the six-yard box. However some of that is down to Lyon's at times disjoined attack, which didn't really get going until later in the campaign (he's also weirdly good, season upon season, at scoring from those wide positions). There’s an overperformance of slightly modest xG numbers but he has been doing that for a while now. The lovely Marek Kwiatkowski‏ recently took a look at quantifying finishing skill and Lacazette came out looking very good indeed. His shot volume hovers around 3 per90 most seasons but what sticks out from recent years is how sharply he has cut the truly poor shots out of his diet. A couple of years ago he was taking around 30 shots from outside the box in a season. Now that’s down to about 10 shots without his general shot volume taking a hit.

The overall skillset he has presents some interesting prospects for combining with Arsenal’s other key players. Lyon sometimes used him on the wings and Wenger has also experimented with him there in pre-season, playing alongside/behind Giroud in addition to a more conventional lone striker role. He’s a solid dribbler - completing 1.8 per90 last season - and racks up very good expected assist numbers for his position. Having him as a focal point, interchanging with Alexis and terrorizing opposition backlines with Özil feeding them might rectify some of their striker-related issues and revive a semblance of their excellent 15/16 attack.

Sead Kolašinac is a very smart addition who fills a need now and in the future. Despite registering 5 assists for Schalke in the Bundesliga last season (joint-most of any player on the team) the left-back’s underlying attacking output is more modest than that headline stat suggests. In his full seasons he produces about 0.7 completed dribbles and 1 key pass per 90. These numbers are actually quite similar to Nacho Monreal's, in fewer minutes.

He profiles as more of a defensive fullback, which Wenger seems to like these days - at least on the left flank. He’s built like an original Xbox and only just turned 24 so has room to improve and should form a nice rotational partnership with the much older Monreal (this likely means that Kieran Gibbs, having played just over 1000 minutes in the last two seasons combined, is surely on the out). Schalke spent time last season in a 3-at-the-back formation so he should slot in just fine.


Squad Concerns

Arsenal's biggest xG contributors last season were Alexis Sánchez (turns 29 in December, likely will leave the club at the end of the season), Theo Walcott (28, injury prone), Olivier Giroud (turns 31 in September), Mesut Özil (turns 29 in October, also is on last year of his contract) and Aaron Ramsey (injury prone, always undershoots his xG numbers). Santi Cazorla (32) would likely rank highly were he not injured.

How Arsenal's xG output was shared last season

Outside of the top contributors there are myriad other uncertain situations spread throughout the squad. Petr Čech is 35, with his closest possible successor Wojciech Szczęsny having just been sold to Juventus after spending two seasons out of sight and out of mind on loan at Roma. Per Mertesacker is retiring at the end of the season and his best partner Laurent Koscielny is 31. Jack Wilshere is clearly not long for this club. Calum Chambers, despite being only 22, doesn't seem to be in Wenger's plans in any serious way.

These examples could go on but you get the point: there's a lot of deferred outcomes and holes that will soon need filling in the current squad. Özil and Sánchez could both very well leave at the end of the season, meaning the club would somehow need to recruit two elite players out of nowhere if they desire to retain their current level. Convincing them to stay, if that's even a possibility, also seems pretty unappetising. Do you really want to be giving pricey long-term contracts to two 29-year-olds? Especially considering Sánchez has played international football every summer dating back to 2014 and is notable for the fact that he is always running? These problems would be lessened if there were comparable talents waiting in the wings to inherit the responsibilities but there just aren't many. Finding such talents is obviously difficult yet Arsenal have avoided making preparations for too long and will likely suffer for it.

The addition of Lacazette should help paper over a few cracks but his presence is not a cure-all. Between the decision to stand pat with the current squad and the purchase of a striker on a similar age timeline it feels like the club are putting an unreasonable amount of eggs in the 17/18 basket. This would maybe be understandable if they were agonisingly close to glory last season, but they weren't. Two seasons ago they had their best chance in a wide-open league and were ultimately pipped to it by Leicester. Now they've kicked the can down the road with the same flawed squad plus a couple of new faces, all being handled by the same flawed manager. Except now their top-6 competition is as strong as it's ever been. Unless everything clicks together in a way that is unanticipated Arsenal are in real danger of missing out on the Champions League for two seasons in a row.

Article by Euan Dewar