The biggest match in week eight of the 2023/24 Premier League season is undoubtedly Arsenal’s clash with Manchester City at the Emirates Stadium. The top two from last season are going head-to-head, with the former able to overtake the latter in the league table if they can win.
History is not on the Gunners’ side. Leaving aside their victories in the Community Shield (on penalties) and FA Cup (in a deserted Wembley), Arsenal have lost their last 15 matches with City. It is little wonder Pep Guardiola’s side are Paddy Power’s favourites for the three points on Sunday.
LONDON, ENGLAND – AUGUST 06: Head Coaches Pep Guardiola of Manchester City with Mikel Arteta of Arsenal during The FA Community Shield match between Manchester City against Arsenal at Wembley Stadium on August 06, 2023 in London, England. (Photo by Robin Jones/Getty Images)
When trying to beat the bookmakers, you can look at underlying expected goal models to see if there is any value to be had in the prices. The data from 2022/23 suggests that City’s price for a victory at Arsenal should be shorter than it is, the numbers from the small sample that is this season shorter still.
But expected goal models and spreadsheets can’t possibly account for all the other factors which can have an effect. How recently did both sides last play? Is it dry and sunny or raining and windy? What the hell is going on in the VAR bunker at Stockley Park?
Most importantly of all, data is unaware of players who can’t take part. Both teams look set to be missing key men this weekend so it’s worth considering how their absence may affect the outcome of this mouthwatering clash.
Arsenal
Arsenal fans will be sweating on the fitness of Bukayo Saka after he had to be withdrawn during their 2-1 Champions League defeat to Lens on Tuesday, and with good reason.
The England international scored 14 league goals last season and provided a further 11 assists. In Premier League history there have only been four previous instances of a player who was no older than 20 at the start of a campaign hitting double figures for both statistics. If you’re in an elite attacking category which contains Michael Owen and Wayne Rooney, you’re probably doing something right.
LENS, FRANCE – OCTOBER 03: Leandro Trossard of Arsenal competes for the ball with Jonathan Gradit of Lens during the UEFA Champions League match between RC Lens and Arsenal FC at Stade Bollaert-Delelis on October 03, 2023 in Lens, France. (Photo by Franco Arland/Quality Sport Images/Getty Images)
However, a team is obviously about far more than the statistics posted by one player. One of the many numbers on FBRef is a player’s individual goal difference, as in their team’s goal difference when they are on the pitch.
Using this information, it’s possible to see whether a side performs better or worse when any player is involved.
Again, there are factors which are not accounted for, with fixture difficulty and split of home or away games hugely relevant and not incorporated. Nonetheless, Arsenal fans might be surprised to learn that the team’s goal difference has been a miniscule 0.01 per 90 minutes better when Saka has been on the pitch.
He has improved their expected goal difference to the tune of 0.29 per 90 but his presence hasn’t had the impact you might have assumed.
By these measures, Gabriel Martinelli is the bigger miss with Arsenal’s expected and actual goal difference being 0.54 and 0.47 per 90 higher respectively when he has played. At the time of writing, Saka looks the far likelier to be involved against the champions, yet the data suggests the return of the Brazilian from the opposite flank might be more pivotal.
Manchester City
Manchester City’s players are somewhat harder to assess via this method. They steamroll over so many sides in so many matches that the figures are usually spectacularly impressive no matter which 11 obscenely paid players are on the pitch.
Even near perfection can be improved upon, though, Spinal Tap wouldn’t have amps which went to 11 if it couldn’t. Kevin De Bruyne has unsurprisingly made City better, to the tune of 0.16 xG per 90 and 0.15 for goal difference.
MANCHESTER, ENGLAND – APRIL 11: Kevin De Bruyne of Manchester City is injured during the UEFA Champions League quarter final first leg match between Manchester City and FC Bayern Munich at Etihad Stadium on April 11, 2023 in Manchester, United Kingdom. (Photo by Visionhaus/Getty Images)
But while fellow absentee Rodri is a touch behind for impact in the latter (at 0.11), his presence lifts City’s underlying performance by 0.30 per 90 minutes.
That’s an enormous margin for a team which has been the best at this metric throughout his time there.
His growing influence can be seen in the figures each year too. Since 2019/20, Rodri playing has altered City’s xG difference per 90 by -0.02, then +0.11, +0.33 and finally +0.71 last season.
It’s his absence more than even that of De Bruyne which explains why City are not stronger favourites to win on Sunday.
History suggests that Martinelli would be a better returnee than Saka to boost Arsenal’s chances though. It’s always worth checking these figures before placing your bet.
*All prices are bang up to date with our snazzy widgets, while odds in copy are accurate at time of publishing but subject to change
More football tips
- Shoot on over to Paddy Power News for the best Football Betting Tips & News
- Lawro’s Predictions for the Premier League this weekend are live!
- Don’t miss this 22/1 Bet Builder for Luton v Spurs
- Here’s our 22/1 Multi-Match punt for Saturday’s 3pm Premier League games
PADDY POWER GUIDE TO SAFER GAMBLING – EVERYTHING YOU NEED TO KNOW
