Welcome to the first piece of Stats Perform’s new Legends Series.
This series examines the historical greats of the game through a combination of Stats Perform’s traditional and advanced analytical metrics. Each instalment will study a different legend of the game and make use of metrics such as Expected Goals (xG) and sequences to bring more meaningful and applicable insights to the fore when analysing player performance. This series aims to demonstrate the value of these insights to those covering sport from broadcasters to digital media outlets.
The first legend under the microscope is the cool-headed Andrea Pirlo.
The Midfield Maestro
It was not until Pirlo made the switch to AC Milan that his true potential started to shine through. Signed by Carlo Ancelotti in 2001, he was the first manager to deploy Pirlo in his now-famous deeper-lying role, a role that would soon become his namesake.
Below we can see Pirlo’s starting positions in Serie A for Milan and Juventus. In a slight tactical shift, Pirlo played mainly in a 4-3-3 at Juventus but his role was still primarily as a deep-lying playmaker.
For our analysis in this piece we are specifically interested in Pirlo’s peak years – 2005/06 to 2014/15.
Looking at Pirlo’s career at Milan, particularly in the aftermath of their infamous Champions League final defeat to Liverpool in Istanbul, is key to defining his legend status. Sitting deep, he was the conductor of the Milan orchestra; instrumental in starting attacks and heavily involved in build-up play, where he demonstrated his vision for the game.
To highlight Pirlo’s metronomic influence, the table below shows his player involvement statistics for Milan aggregated by per 90 minutes, using Opta’s sequences framework. Sequences are defined as passages of play which belong to one team and are ended by defensive actions, stoppages in play or a shot. Looking at the game through sequences allows us to more effectively analyse players’ contributions and give credit to those who might have been missed in the past.
Sequences are particularly useful in analysing Pirlo’s effectiveness. Despite Pirlo’s world-renowned creative talent, he only registered double figures for assists one season in Milan. However, looking at the number of sequences he started, and the attacking value these generated, tells a different story.
Pirlo was the heartbeat of those Milan sides, involved in both starting and ending attacking moves with a potency that dwarfs the average Serie A midfielder during this time. In four out of six of his seasons at Milan, he was ranked in the top ten in the league for the xG value generated by his involvement in build-ups. In 2010, one of the seasons he was not ranked among the top ten, he played a mere 1,139 minutes.
In comparison, defensive and central midfielders over the same time period started an average of a mere 8.9 sequences per 90. The xG value of their contributions to team build-up was only 0.13 compared to Pirlo’s 0.3. Pirlo eclipses his peers in every category.
What does a playmaker look like?
The passing network below typifies Pirlo’s role. At his peak he was always the main link, the target player to transition teams from attack to defence.
There’s a high volume of passes from Silva, Antonini, and Ambrosini to Pirlo – who much of the passing looks to revolve around – and from Pirlo, we see this converted into high value passes to Pato, Seedorf, and Antonini.
When Pirlo made the switch to Juventus in 2011, he had already established himself as one of the most talented playmakers in world football. However, the move to Turin saw Pirlo take a more direct role in attacks, playing longer passes than he had done in Milan.
Pirlo’s much-heralded long pass expertise stands up to scrutiny when looking at the number. In his first season for Juve, Pirlo attempted 13.23 long balls per 90 minutes with an astonishing 84% success rate. He used this to great effect, gaining 10 assists in that year.
This more direct passing can be demonstrated by Pirlo’s pass map for that first season in Turin. This pass map uses Expected Assists (xA), the likelihood that a given pass will become an assist, to illustrate Pirlo’s most threatening passes.
Filled blue points indicate passes that assisted goals, unfilled blue points indicate assisted shots (which were not goals). A great many of these key passes came from deep locations, highlighting his bullet-point precision.
Who could carry the Pirlo torch?
Taking some of Pirlo’s stand out metrics, we looked at players who were under 24 years of age at the start of the 2019/20 season to see if there were any players who exhibited the same skill set as the Italian legend.
Stefano Sensi came out at number one on our list. Looking at the last three seasons, Sensi averages 0.19 assists per 90, provides 2.28 key passes per 90, and makes 6.15 long balls per 90, which have an accuracy of 68%.
As well as similarities in match-level stats, Sensi also performs similarly based on our player involvement reports. He shows a similar number of sequences started to Pirlo, also highlighting a respectable Expected Goals output for sequences started. We can see he is also integral in build-up with a high number of involvements in build-ups per 90.
Rumours abound that Barcelona are interested in the 24-year-old. Having just entered his peak years, it will be fascinating to follow his career. He may someday be as good, but there will only ever be one Midfield Maestro.