Sir Alex Ferguson once said that “attack wins you games, but defence wins you titles.” In the first Premier League season, United won the title, conceding nine fewer goals than the next best team defensively, but having been outscored by Blackburn Rovers by one. And Liverpool’s acquisitions of Virgil van Dijk and Alisson have been credited with contributing hugely to their title win and to last season’s Champions League, tightening up a defensive unit that had often looked shaky. But was Sir Alex, right? Looking at the Premier League since it began in 1992/93, not including the current season, a few things are notable: the team with the highest total of goals scored won 17 of 27 titles; the team with the lowest total of goals conceded won 11 titles; and the team with the best goal difference won 19 titles. Only six times did a team manage to top all three categories, and they are generally considered all-time great league teams: Manchester United 2000/01 Arsenal 2003/04 Chelsea 2005/06 (tied with Manchester United for goals scored) Manchester United 2007/08 Manchester City 2011/12 Manchester City 2017/18 Only three times have a side won the league and not been best in any of the three categories: Arsenal 1997/98 Leicester City 2015/16 Chelsea 2016/17 The team with the best defensive record winning the title is a more recent phenomenon: while it applied to that 1992/93 United side, it only happened another three times up to and including 2003/04 – but after that, it happened a further seven times. It’s also worth noting just how important goal difference is – it’s much better to be great at both ends, as one would expect. For example, in 2014/15, Chelsea won the title conceding only 32 goals with a goal difference of 41; the next best defensive team was Southampton, who conceded only 33 times, but they finished in seventh with a goal difference of 21 because they found scoring hard.
To sum it up: “If you're a very good defensive team, it means matches are lower-scoring, and therefore you are more vulnerable to dropping points in draws against weaker teams. Whereas if you're an attacking team, it is easier to blow away weaker opponents.” Other studies have born this out. When the Numbers Game authors Chris Anderson and David Sally looked at data from the Premier League, Serie A, La Liga, and the Bundesliga, they found that between 1991-2010, the “best defence will pick up a championship 46% of the time”. The best attack, meanwhile, won the championship 51% of the time. There was variety among the leagues, too: the Bundesliga and Serie A saw a greater number of seasons won by the best defence compared to the best attack, while La Liga and the Premier League saw the reverse. But in short, while it’s marginal, it’s better to have a good attack. As the book points out, the flaw with just looking at output is that you don’t know how the team won points – did they attack or look to keep it tight? So, they investigated. They conducted analysis of Premier League matches between 2001/02 and 2010/11 to compare the value of goals scored and goals conceded.
And what is noticeable from their work is that when looking at avoiding defeat, rather than winning, “the goals clubs didn’t concede were each 33 per cent more valuable than the goals they scored”. This means that when looking at winning, scoring goals and not conceding goals are of roughly the same value; when considering how not to drop points by losing, it’s more valuable to focus on defense. It’s very hard to work out a value in points or wins or avoiding losses that each player brings. This is because football is a hugely complicated game where so many thousands of actions, some measurable, some not, contribute to a performance.
However, we can look Liverpool’s goalkeeping situation and infer a few things: Comparing the last three seasons, it’s very clear that even if Alisson cannot be credited specifically with winning games or not losing them, he is a huge improvement on what has gone before. Not only does he have the lowest goals conceded per 90 of any of Liverpool’s five ‘keepers in the last three seasons, which is, of course, partly a team effort, he’s also got the best save percentage. And, crucially, when considering his post shot expected goals plus or minus, which is the post shot expected goals value minus actual goals conceded, he’s exceptional. In simple terms, in the last two seasons Alisson has prevented 8.6 goals that based on their post shot xG would be expected, on average, to go in. Mignolet and Adrian, on the other hand, let in more goals than on average one would expect from the shot’s expected goals value. Here, at least, Alisson’s value is clear. And, as we’ve seen from Anderson and Sally’s work, when avoiding defeat, these goals prevented were worth a third more than goals scored at the other end.
So, in Liverpool’s case, upgrading their defence and goalkeeper will certainly have helped them avoid defeat. And this put them in the hunt for the title because upgrading a pretty poor defence and making it world beating then provided a platform for their superb attack. Not only could they now achieve victory in games, they could avoid losing them. That balance, really, is what wins titles.