In 2010 the Fenway Sports Group purchased English Premier League side Liverpool Football Club. From the outset, their intentions were simple. They would apply statistics-based approaches to the running of Liverpool in terms of player acquisitions and team tactics.
The method, popularised in the Michael Lewis book Moneyball about the Oakland A’s baseball team, had not yet been utilised in soccer but Fenway set about adapting the approach.
Liverpool’s early signings signalled their new philosophy – They paid £20m for Stewart Downing, the former Aston Villa player heralded for his ability to cross the ball. They paid £7m for Charlie Adam, a playmaker known for his pinpoint delivery and whose ability to take corners Sir Alex Ferguson had once claimed “was worth £10m alone”. To complete the package, Liverpool paid £50m to Newcastle United for Andy Carroll, a twenty-one year old striker who was strong in the air and had scored eleven league goals in the first nineteen games of the season.
Many scoffed at the fees being paid by Liverpool and the very notion that traditional soccer, a game of skill, bravery and occasional luck, was reducible to mathematical scenarios. Soccer after all, is a chaotic game played at savage intensity. As if to demonstrate the ‘chaos’ of the game, Liverpool had lost to Sunderland in October 2009 after an opponent’s shot was deflected into the net by a beach ball thrown from the crowd.
The old adage of stoic English soccer was ever-present – how could statistics predict what could happen on a cold, rainy night in Stoke?
The problem with adapting the Moneyball approach from baseball is determining in a completely different sport the type of statistics that might be worthwhile.
The earliest and most rudimentary incarnation was based on goals and assists. This was hardly earth shattering or innovative. Nor was it originated by the Fenway Sports Group at Liverpool, but the signings of Downing and Carroll were the first clues into the machinations of their new approach to business.
The utilisation of statistics in soccer has always been beset with difficulty. Debates abound regarding the greatest players of all time; Pele or Best, Maradona or Messi.
Even the debate into the greatest Argentinian player of all time is likely never to be settled; least of all by the record books which show that Messi has scored 398 league goals to Maradona’s mere 108 league goals.
Even the goal considered Maradona’s greatest ever; his winner against England in 1986 was more about the skill, agility and control to beat the English side rather than the ultimate finish.
So what would account for Messi’s vastly superior record? He is a product of the modern era for one. He has played more games, though his goals to game ratio is also better. In his book The Sports Gene, journalist David Epstein argues that in athletics and swimming, advancements in technology and understanding have allowed former world records to be smashed. The Marathon World Record is now 53 minutes faster in 2019 than it was in 1908 due to changes in physiology, diet, training and equipment.
A similar argument could be made for soccer.
Soccer balls have been redesigned to swerve and favour attacking players. Importantly, rules have been changed to protect attacking players and prevent cynical play.
Pitches have also improved since the era of Pele and Maradona. That infamous Northampton Town divot covered pitch on which George Best scored six goals in 1971 is not a sight which greets many modern-day footballers.
Returning to Liverpool; the Merseyside club’s new statistically informed signings failed to perform aswell at Anfield as they had at lesser sides. The club then changed tack and pinned their hopes on a new manager.
Brendan Rodgers was less focused on statistical approaches to soccer, but he did rate one particular statistic above all others: possession.
When you’ve got the ball 65-70% of the time, it’s a football death for the other team. We’re not at that stage yet, but that’s what we will get to. It’s death by football.
Rodgers signed Joe Allen, nicknamed ‘the Welch Xavi’ for his passing ability, to partner Steven Gerrard in midfield and subsequently sold Downing, Carroll and Adam as their skillsets were no longer suited to the Liverpool philosophy.
The problem for Rodgers however was Liverpool’s impressive possession statistics did not always translate into goals nor guarantee success on the pitch. In early 2014, Liverpool amassed 74% possession away to West Bromwich Albion but still only drew 1-1.
And then there’s the uncomfortable question – does this superior possession equate to good football? The BBC match report did not highlight any apparent superiority from the Merseyside outfit in that game despite having the majority of the play.
Earlier this season, the Premier League published an article on their website celebrating Jorginho’s performance in Chelsea’s 2-1 win over Newcastle. Jorginho equalled a Premier League record of 186 touches of the ball during the game, in what was described as a ‘dominant display’ by the Italian international.
The problem however is that Jorginho equalled a record set by Fernandinho in March 2018 in a game between Manchester City and Everton. Following that game, an English national newspaper credited Fernandinho with making vital interceptions but rated his performance as “6/10”.
A further examination of passing records would reveal that Arsenal’s Santi Cazorla is fourth on the Premier League’s all-time list. Cazorla registered 178 touches in Arsenal’s 0-0 draw with Sunderland in May 2015 yet the BBC match report failed to mention Cazorla entirely, something which is hardly indicative of a domineering display.
It returns to that old question – what is ‘good football’? Does statistical success of one sphere necessarily equate to footballing success or enjoyment for the spectators? Barcelona famously had 71% possession away to Inter Milan in 2010 yet lost 3-1 in one of the most memorable Champions League games of the last decade. Inter Milan performed a defensive, counter attacking tour-de-force, something which was the true narrative of the game, and one which could not be explained in numbers alone.
Herein remains the fundamental problem with statistics. Despite the arrival of newer, more advanced statistical summaries like Expected Goals (xG) whereby the number of goals a team or player is expected to score based on their chances and possession, the use of statistics is still an attempt to explain a largely qualitative game in quantitative terms; terms which may not reflect the true narrative in which the game evolved.
Liverpool now sit atop the 2018-19 English Premier League table at its half way point, though their approach has required significant refinement and nuance before it began to pay dividends.
In 2015 the team appointed Jurgen Klopp, a manager whose teams are built around fast, attacking football with relentless pressing and harrying of opposition players. Klopp’s football brand, ‘heavy metal football’, represents something akin to organised chaos rather than statistical reverence.
It appears as though Liverpool have consigned the ‘Moneyball’ approach to the past yet there is no doubt that players chosen for the current incarnation of the Liverpool – Joe Allen has long been sold – were selected in part based on statistics relevant to Klopp’s approach – factors such as sprints per ninety minutes – but it is now considered as part of a balanced ensemble approach.
What the soccer world has learned since the Fenway Sports Groups’ initial foray is that cold, hard statistics and soccer don’t sit entirely comfortably together. Statistics can be highly useful, but something more is clearly needed. The old football men are not to be thrown out just yet. A collective gasp from the crowd at something special under the lights at the Allianz Arena cannot be understated.
The Zidane Roulette could never be quantified in facts and figures, nor can a Messi stepover, and they are some of the most special elements of the game, capable of transfixing supporters and opponents alike. It would be naïve to suggest that sport cannot benefit from measurement, analysis and statistical support but statistics clearly don’t tell us everything.
The Republic of Ireland famously beat a star-studded Netherlands side in Dublin in 2001 to set up qualification for the 2002 World Cup. Jason McAteer scored the decisive goal, but the stage was truly set on 33 seconds when Roy Keane hacked down Marc Overmars with a ferocious tackle from behind that was subsequently described as a ‘reducer’ in the national media. The Dutch players were irate. The crowd went wild. Patrick Kluivert was clearly rattled, and the tackle served to turn up the atmosphere in an already cauldron-like stadium.
There is no statistical equation to reflect either the intensity of that tackle or the manner in which it galvanised and inspired the Irish team, but they proceeded to then produce one of their finest ever displays on the international scene.
For some things, you just need to see it with your own eyes.