Alex Keble analyses the data to see if it is defence or attack that is most important to winning the league.
“Attack wins you games, defence wins you titles.” So said Sir Alex Ferguson, and he should know. He won quite a lot of them with Manchester United.
It’s no surprise the adage has stayed in vogue, or that Ferguson’s former players would believe in its simple truth.
Nemanja Vidic was at the heart of defence for five of Ferguson’s 13 Premier League titles. Speaking about Arsenal at the weekend, he said: “They're starting to understand what it takes to be champions.
“It's not just going forward. It's about how you defend."
But in actual fact, only 60 per cent of champions in the last 10 Premier League seasons had the best defence.
Does that mean Ferguson’s famous saying is wrong? We’ve looked into the numbers to find out.
Why Arsenal have the best defence
Arsenal’s defensive record has been incredible this season.
They have the best defence in the league, shipping only 24 goals, and in 2024 have conceded only four times in 11 matches, allowing 4.98 Expected Goals Against (xGA), fewer than half as many as anybody else.
Top Premier League defences in 2024
| Team | Expected Goals Against |
|---|---|
| Arsenal | 4.98 |
| Man City | 12.28 |
| Liverpool | 14.27 |
Things are getting better over time for Arsenal, whose impenetrable centre-back partnership of Gabriel and William Saliba barely let a shot reach David Raya.
In fact, not only have Arsenal allowed the fewest shots on target, with 67, but 32.8 per cent of the shots they have faced have been blocked by defenders, which is the second-highest proportion behind Nottingham Forest.
Arsenal’s defence is immaculate. But does that mean they should be title favourites?
Attacks, not defences, have historically won the league
Looking back over the 28 Premier League seasons, since the league was reduced to 20 clubs, Ferguson’s adage does not hold true.
Only in 14 of those 28 seasons have the team with the best defence won the title – half the time - whereas on 19 occasions, or 68 per cent, the team with the best attack have lifted the Premier League Trophy.
There are some remarkable outliers in there, and funnily enough it tended to be Ferguson himself who won titles without one of the best defences.
In 1996/97, Man Utd had the seventh-best defence but still won the title. In 1999/2000 they had the sixth-best defence but were again victorious. And in 2012/13, his final year at the club, they had the fifth-best record.
There are no such anomalies when it comes to attack. Only twice ever has a team won the title without being in the top two for goals - in 1997/98 Arsenal ranked third but won the league, as did Leicester City in 2015/16, which was an outlier in every respect.
PL winners' rank for defence and attack
| Season | Champions | Defence rank | Attack rank |
|---|---|---|---|
| 95/96 | MUN | 3 | 1 |
| 96/97 | MUN | 7 | 1 |
| 97/98 | ARS | 2 | =3 |
| 98/99 | MUN | 4 | 1 |
| 99/00 | MUN | 6 | 1 |
| 00/01 | MUN | 1 | 1 |
| 01/02 | ARS | 2 | 2 |
| 02/03 | MUN | 1 | 2 |
| 03/04 | ARS | 1 | 1 |
| 04/05 | CHE | 1 | 2 |
| 05/06 | CHE | 1 | =1 |
| 06/07 | MUN | =2 | 1 |
| 07/08 | MUN | 1 | 1 |
| 08/09 | MUN | =1 | 2 |
| 09/10 | CHE | 2 | 1 |
| 10/11 | MUN | 3 | 1 |
| 11/12 | MCI | 1 | 1 |
| 12/13 | MUN | =5 | 1 |
| 13/14 | MCI | 2 | 1 |
| 14/15 | CHE | 1 | 2 |
| 15/16 | LEI | =3 | 3 |
| 16/17 | CHE | 3 | 2 |
| 17/18 | MCI | 1 | 1 |
| 18/19 | MCI | 2 | 1 |
| 19/20 | LIV | 1 | 2 |
| 20/21 | MCI | 1 | 1 |
| 21/22 | MCI | =1 | 1 |
| 22/23 | MCI | =1 | 1 |
Liverpool boast biggest goal threat
So, attack is generally more important than defence. Luckily for Arsenal they are also top of the league in this regard – but only just.
Arsenal have scored 75 goals, three more than Liverpool, but with seven matches to go there is certainly time for that to change, especially with Jurgen Klopp’s side facing a slightly easier run-in.
They have two of the top six left to play, while Arsenal have three.
What’s more, Arsenal rank third in the Premier League for Expected Goals (xG) on 61.3, behind Manchester City's 62.9 and Liverpool's 70.8, the latter being way out in front.
Liverpool won’t need to improve their finishing by very much to end the season as the best attacking team.
Does that make them title favourites then?
Defences have dominated since Pep's arrival
Not so fast. Pep Guardiola has radically altered the Premier League, and things look very different when you isolate the numbers for the last six seasons, starting with the first year that he won the title.
Since then, the title-winners have had the best defence in five out of six seasons.
And in the one exception, in 2018/19, Man City’s xGA was the best, at 24.7 compared with Liverpool’s 28.8, despite City conceding 23 goals and Liverpool conceding 22.
Champions' Expected Goals Against
| Season | Champions | Defence rank | xGA (PL rank) |
|---|---|---|---|
| 2017/18 | Man City | 1 | 23.8 (1st) |
| 2018/19 | Man City | 2 | 24.7 (1st) |
| 2019/20 | Liverpool | 1 | 37.7 (=2nd) |
| 2020/21 | Man City | 1 | 30.2 (1st) |
| 2021/22 | Man City | =1 | 24.6 (1st) |
| 2022/23 | Man City | =1 | 32.1 (1st) |
In other words, the champions in each of the past six seasons have either conceded the fewest goals or had the lowest xGA.
In that time, Man City have been the highest scorers on every occasion. But the fact that Liverpool won the title in 2019/20 by 18 clear points despite scoring 17 fewer goals than Man City further cements the idea that defence trumps attack in the modern game.
Which brings us back to Arsenal, who have the best defensive record this season with their 24 goals conceded and – the pièce de résistance – an xGA considerably better than their rivals.
Title rivals' Expected Goals Against 23/24
| Club | xGA |
|---|---|
| Arsenal | 21.3 |
| Man City | 30.1 |
| Liverpool | 36.2 |
Based on the last few seasons - long after Ferguson’s retirement - his famous saying is finally coming true, putting Arsenal in the driver’s seat.
