Moneyball Part 2: Brad Pitt Moves to LiverpoolPosted: September 18, 2011 | Author: brucem | Filed under: Measurement and Analytics | Tags: Applying Statistics to Soccer, baseball statistics, Brad Pitt, John Henry, Liverpool football team, Moneyball, Premier League, statistical thinking, Tom Werner | Leave a comment
Just kidding. But in a recent article by Matthew Futterman and Jonathan Clegg, titled “The Red Sox Invade Liverpool: Team’s New Owners Take a Statistical Approach to Its Roster,” we read that the new owners of the Liverpool soccer team, John Henry and Tom Werner, are applying the statistical and analytical methods employed by Billy Beane with the Oakland A’s and then employed by others with the Boston Red Sox. But who knows? If the film Moneyball does well at the box office, Hollywood, given its penchant for sequels might not be able to resist plopping Pitt down in Liverpool.
Writes Futterman and Clegg:
Broadly, Liverpool’s spending spree crystallizes all the risks and mysteries associated with adapting analytics to a game where the jury is still out on their value. After all, there are so many scouts, would-be agents and lay observers on the lookout for talent that the idea of fielding a team with undiscovered value would appear absurd.
Every Premier League club has a team of number crunchers, but it’s unclear how much influence they have or which of the numbers they produce are actually useful. Even Henry, who owes much of his fortune to his mathematical approach to the financial markets, sees a limit to the use of data in a free-flowing game like soccer, as opposed to baseball, where events can be isolated and statistically adjusted.
“Determining what goes into a goal is much more complicated than determining what goes into a home run,” Henry wrote in an email this week.
There is widespread agreement that no one has come up with the algorithm that reveals a player’s value to a team. “I’m not aware of anyone who has cracked the code,” said Nelson Rodriguez, executive vice-president of competition for Major League Soccer, who consults regularly with European clubs.
There may be too much information. Data-tracking companies collect some 300 measurements per game on about 2,500 player movements. Still, Rodriguez said clubs continue to search for a winning formula, taking their cues from the data-obsessed franchises in U.S. pro sports.
To cut through the confusion, Liverpool seems to have gone at the problem with a rather simplistic—if expensive—approach that speaks to the owners’ experience with the data-centric Epstein-led Red Sox.
The great discovery made by Red Sox senior adviser and stats guru Bill James a generation ago was that baseball teams could score more runs if they could get more runners on base, in whatever fashion, and thereby create more scoring chances.
Liverpool finished sixth in the Premier League last season partly because of its record of scoring just 59 goals in 38 games. That was far off the pace of Manchester United, which scored 78 goals. In four of the past five seasons, the top four teams in the standings were the four highest-scoring teams, the lone exception being 2009-10, when Manchester City finished fifth but scored more than Tottenham, which finished fourth.
To solve the shortfall, Liverpool has gone on a mission to create more scoring opportunities—to get more runners on base—even if that has meant betting the house on a group of players who, other than Suárez, have little international or Champions League experience.
For instance, Downing, Adam and Henderson were all among the top-eight chance-creators in the Premier League last season, according to statistics provider Opta Sports. The average Premier League midfielder creates roughly 1.21 chances per game, but Adam created 2.06 chances per game last season on average, while Downing and Henderson made 2.24 and 2.22 chances per game respectively.
Those three also produced some 750 crosses—centering passes that create scoring opportunities—last season, with an accuracy rate of more than 24%, slightly ahead of the league average. Five of Adam’s eight assists last season came from crosses, while no Premier League player has made more successful crosses the last three seasons than Downing. Since the start of the 2004-05 season, only four players have created more chances than Downing in Premier League games—Frank Lampard, Cesc Fabregas, Steven Gerrard and Ryan Giggs.
Conventional wisdom suggests those numbers should improve if the trio is crossing to a classier collection of forwards and midfielders than on their previous clubs. The hope is that a good portion of the crosses will connect with the 6-foot-3 Carroll, whose specialty is heading the ball into the net.
Still, there are those who question whether bombarding an opponents’ area with aerial crosses is an effective tactic. The rate of completed crosses that produce a goal rarely exceeds one in four. The probability of scoring from a cross is just 5%.