So arrivederci then, Gianluigi Buffon. Assuming there is no careering handbrake turn away from retirement, the match against Verona on Sunday will prove his last. It seems vaguely astonishing that he made his debut in 1995 – a teenage action hero clad in pink and black for Parma, fearlessly facing down George Weah, Roberto Baggio and the rest of Milan’s all-stars. Nearly 900 games later, he will hang up his Puma One Grips having averaged a trophy for each year of his career.
And along with a World Cup, Uefa Cup, five Italian Cups, six Italian Super Cups, a European Under-21 title and multiple Scudetti, Buffon holds another honour. For he remains the only keeper in the 50 biggest transfers in history according to Transfermarkt (which tracks deals in Euros) – even though his €52m move to Juventus took place 17 years ago.
Back in 2001 that price seemed crazy. Now it looks like one of the bargains of the century. Yet few clubs have dared follow the Old Lady’s lead. Incredibly, Transfermarkt lists only 11 keepers who have cost more than €15m (£13.2m). In a world where Theo Walcott and Guido Carrillo both cost £20m that seems bizarre. But according to Nick Harris’s excellent Sporting Intelligence website it illustrates a wider trend. His numbers show that keepers are also paid less than defenders, midfielders and strikers – and have become less valued relative to outfield players too.
In the 2005-06 season, for instance, Premier League keepers were paid £533,000 on average – 79% of the average basic annual salary of all players, which came to £676,000. By 2016-17, keepers in first-team squads earned £1.68m on average – only 69% of the average Premier League basic salary of £2.4m.
It seems obvious that some keepers are massively undervalued. The question is can we prove it?
Ted Knutson, who worked on player recruitment at Brentford and the Danish club FC Midtjylland and heads the football consultancy StatsBomb, believes so. As he explained during a presentation at South Bank University last week, keepers are often harder to evaluate than other positions. They need to sweep up, distribute the ball accurately and start attacks as well as keep clean sheets. Yet the data isn’t always there to properly assess their strengths and weaknesses.
Save percentage, for instance, matters little if every shot goes down a keeper’s throat. And while looking at how a keeper performs compared with the expected goals (xG) they are predicted to concede is more robust, it doesn’t take into account defensive pressure or the power of a shot.
Knutson recalled a chat with Bob Bradley, the former coach of Swansea, while interviewing him for the Midtjylland manager’s job. While amenable to using data, Bradley pointed out an obvious flaw with xG. “You can’t tell me that if I have two men on a guy having a header from six yards out that is a good chance,” he said. “I know for a fact it is a very hard to score.”
Knutson conceded he had a point. “But I have to look across 30 leagues across the world to find undervalued players,” he replied. “And I cannot duplicate your eyes across 20,000 players and multiple seasons.”
Now, however, Knutson believes he has a more reliable way of assessing chances and keepers. A key breakthrough is that the velocity of every Premier League shot can now be tracked (unsurprisingly Riyad Mahrez and Harry Kane lead the way in attempts from distance) – and his data also shows the exact position of each player when the ball was hit, and whether the keeper was moving, set, or on the ground.
This gives the scout or analyst a wealth of information. It means they can assess a keeper’s reaction time; how good their positioning is compared with other keepers across multiple leagues; and, ultimately, how good their saves are. It could, suggested Knutson, be a game changer.
Using this data, his StatsBomb colleague Derrick Yam then ranked keepers in the Premier League in 2017-18. Unsurprisingly David de Gea was right up there, conceding eight goals fewer than an average keeper would based on the shots he has faced. Arsenal’s Petr Cech ranked last, having conceded six goals more than expected (Liverpool’s Simon Mignolet was not far behind).
We all know De Gea is far better than Cech. Yet having such numbers gives us a much better idea of their respective worth. As Knutson says: “Conceding eight goals less a year than average is a massive amount. Flip it around and say an average striker scores 10 goals a year and is worth £20m. Depending on age and other factors, an extra eight league goals could make a striker three times as valuable.”
And if De Gea’s exceptional performances are repeatable across multiple seasons – and the evidence suggests they are – he and other top young keepers should be worth a minimum of £50m or £60m.
Longevity should be another factor in their favour. We know that players decline physically as they wade into their 30s. Yet keepers appear to cast off their powers at a slower rate, and what they lose in reactions they often make up for by reading the game better.
Certainly no one at Juventus will be counting the cost of signing Buffon all those years ago. Especially after he paraded a seventh successive Scudetto in front of an adoring Curva Sud Scirea on Sunday.