Why football’s goalkeepers are cheap and unheralded

By T.A.W.

FOOTBALL fans have become used to seeing transfer records broken, as Europe’s top clubs have enjoyed a decade of rapid growth in revenues. The latest transfer window, a month-long mid-season affair which closed on February 3rd, offered further proof of a bull market. The teams in Europe’s “big five” leagues, in England, Spain, Germany, France and Italy, spent an unprecedented £815m ($1.15bn) on acquiring new players. The three most expensive transfers ever have all been completed in the last six months. Yet despite this prolonged spree, one long-standing milestone has yet to be passed. The record for the most expensive goalkeeper is still unmoved after 16 years.

That player, Gianluigi Buffon (pictured), is now 40 years old. His €53m ($66m) move to Juventus in 2001 is one of only three goalkeeper purchases among the 200 priciest sales of all time, according to Transfermarkt, a football statistics website. Widely regarded as the greatest goalie of his era, he is one of just five to have a top-three finish in voting for the Ballon d’Or, an annual award given to the world’s best player. Lev Yashin remains the only keeper to have won the prize in its 62-year history, claiming the title in 1963. Although flying goalkeepers are one of football's most thrilling sights, they are cheap and unheralded when compared to their super-star teammates. Why?

One of the biggest obstacles that keepers have faced in the battle for recognition has been a lack of data. Despite spotty records, an interested fan can quickly discover how many goals celebrated attackers like Pelé or Diego Maradona scored, with roughly 800 and 350 respectively. But there is no estimate of how many goals Yashin prevented in his career, or even how many saves he amassed. Although football fans and teams are gaining in statistical sophistication, they still lag far behind their counterparts in other sports. The game’s most comprehensive stock of match data, Whoscored.com, only has information for goalkeeping saves going back to 2009.

For more recent seasons, quantitative analysis is becoming increasingly reliable. Post-game interactive diagrams that show the minute and location of every significant event in a match, such as shots, passes and tackles, are now available for free online. This raw data enables analysts to produce rough estimates of each player′s contribution. The most common, “expected goals”, shows how often a league-average player would score from a shot given its location, the part of the body used and the nature of the preceding pass, among other variables. It is an intuitive way to assess whether a team is getting better results than usual from the chances it creates and allows.

Expected goals can measure striking prowess, but also shot-stopping ability. Colin Trainor, an analyst, has published “post-shot expected goals” figures for a handful of keepers since 2010, which also account for the placement of the attempts. The numbers confirm that Mr Buffon is an exceptional shot-stopper. Even during the twilight of his career, he has conceded 20% fewer goals than an average goalie in a major European league would have, given the shots that he faced. Manuel Neuer (aged 31) and David de Gea (aged 27), two of his most talented heirs, have both conceded 17% fewer goals than expected over the last eight seasons.

Those statistics must seem like vindication for long-maligned goalies. Yet a 20% advantage over a mediocre keeper makes little difference to an already frugal defence. Messrs Buffon, Neuer and de Gea play respectively for Juventus, Bayern Munich and Manchester United, three of the world’s strongest teams. On average, their acrobatics have saved their clubs between four to six goals each season relative to a league-average counterpart. Those are handy contributions. But how do they compare to the 30-odd goals smashed in by the best strikers?

The fairest way to assess goalkeepers′ impact, as well as that of outfield players, is to measure how many league points they earned. This idea is most familiar to fans of baseball and basketball. Because baseball is primarily a series of one-on-one confrontations, each player can essentially be granted full credit for their teams′ performances when they are at bat or on the pitching mound. Similarly, in basketball, lengthy fixture lists and frequent personnel changes make it possible to work out how a team fares with and without a certain player, and therefore how much value he adds. In both cases, these contributions are usually benchmarked against the expected performance of a typical veteran free agent earning near the minimum league salary.

Replicating this method for football is tricky. For a start, there is no minimum wage. Moreover, measuring performance with and without a player is difficult—especially for keepers, since the number-one choice at each club will typically play more than 90% of league minutes. Nonetheless, squad rotation, injuries, transfers, cup competitions and international fixtures mean that most players will be deployed in various combinations with teammates.

The 21st Club, a football consultancy firm, has developed a model that uses these variables to estimate every player′s value (see chart). By their reckoning, both Mr de Gea and Mr Neuer are worth about four points per season more than the average player in their position in the Premier League, which has been used as the benchmark. That is the equivalent of turning two draws (which are worth one point) into wins (which are worth three). In fact, the keeper rated most highly by the model is Atlético Madrid′s Jan Oblak, a comparative youngster at the age of 25, who has contributed nearly five extra points a season.

Yet these gains are dwarfed by those created by the best attackers. Lionel Messi adds 11.3 more points to Barcelona’s total than would be expected of an average Premier League player in his position. The next star in the 21st Club’s ranking is not Cristiano Ronaldo, who is generally considered to be Mr Messi’s rival for the title of world’s best player, but Luis Suárez, another of Barcelona’s attackers. Not even Mr Oblak would find himself among the list of the top 15 outfield players. Across the strongest ten clubs in each of the big five leagues, the 21st Club calculates that the typical starting striker is worth two points per season above the Premier League average in his position, whereas goalkeepers add 0.7 points. In the past four transfer windows, strikers bought by those teams have cost €17.1m on average, at a rate of €8.5m a point. Goalkeepers have cost €6.5m per head, at a slightly higher rate of €9.4m a point. That suggests that keepers cost less than forwards not because they are underappreciated, but because they are less valuable.

The 21st Club notes some important distinctions between shot-stoppers and outfielders. Goalkeepers are the only players that offer almost no attacking benefit. Of the 50 players who have completed the most passes in this season’s Premier League, none are keepers, whereas roughly half are defenders, who have become increasingly involved in offensive play. Between them, the keepers have contributed two assists—passes that lead directly to a goal—while the defenders have chipped in 89. By contrast, even the most tackle-shy attackers contribute defensively, since holding the ball reduces an opponent’s scoring chances. It has also become increasingly fashionable for teams to deploy a “pressing” style, in which even the strikers harry opposition players who have the ball. Few fans would pick flamboyant Barcelona as the world’s stingiest team. But no club has a higher defensive rating on the global index maintained by FiveThirtyEight, a statistics website.

Not all statisticians agree that goalies are fairly priced, noting that keepers—including Mr Buffon, who could captain Juventus to a seventh straight Serie A title this year—tend to have far more longevity and durability than players in other positions. Stefan Szymanski, an economist and author of “Soccernomics”, a football analytics book, believes that goalkeepers may be undervalued, though a lack of data on players’ salaries makes it difficult to be sure. Clubs often disclose the transfer fees paid for a star, but rarely his wages. However, Nick Harris of Sporting Intelligence, a sports website, claims that the typical goalkeeping salary in the Premier League has dropped from 79% of the league average in 2006 to 69% last season. Increasing discounts for goalies would surely annoy the countless footballers who choose to play between the posts for fun (among their number have been Albert Camus, Vladimir Nabokov, Luciano Pavarotti, Che Guevara and Pope John Paul II). But the data suggest that the boot is indeed more valuable than the glove.