Thursday 25 February 2016

The Birth of the Aging Striker?

During the Champions League round-of-16 first legs, one thing worth noticing was that the ones to open the score for the all-star teams of Real Madrid and PSG were “old-timers” and club top scorers Cristiano Ronaldo (31) and Zlatan Ibrahimovic (34), respectively. What is more, Zlatan has now scored in each of his last nine(!) appearances for PSG, including Champions League, Ligue 1 and French Cup. In the only one of their last ten encounters PSG did not manage to score, the Swedish veteran was not fielded. 

Source: telegraaf.nl

Meanwhile, Jamie Vardy, aged 29, is leading the Premier-League’s "pichichi". An unlikely feat, particularly given that, at the age of 24 – generally considered already quite “old” for a striker – Vardy seemed to have reached his terminal in the fifth tier of English football. During the 2014 FIFA World Cup, the BBC published an article claiming that, in accordance with popular wisdom, a football player peaks at the age of 27.5. Naturally, goalkeepers are expected to peak above and strikers below this average. Are Vardy, Zlatan and others defying logic then? Or could it be the logic rather than these players that is dated? 

As you probably rightfully noticed, all of the above examples I carefully cherry-picked. For instance, there was no mention of the opening goals by youngsters Dybala and Draxler, both 22 years of age. In order to see if something of interest may actually be going on, let’s be slightly more scientific. As of currently, the top-10 scorers active in the five main European leagues, including any goals they may have scored in the Champions or Europa Leagues and with penalties counting for half a field goal, are, in order: Cristiano Ronaldo, Suárez, Lewandowski, Higuaín, Aubameyang, Ibrahimovic, Benzema, Müller, Aduriz and Messi. Their average age is slightly above 29 years and 6 months (median age: 28.44). For strikers – moreover, the top ones in Europe at this very moment – that seems pretty old. It begs the question, “are top strikers – as is the general population – actually getting older?”

Source: marca.com

In order to investigate this question, I turn to the “Golden Shoe” (or "Golden Boot"), an annual trophy awarded to the most prolific scorer in any national league within Europe (since 1997, adjusted by a coefficient depending on the strength of the league). The Golden Shoe is being awarded since the 1967-1968 season, when Eusébio was the first to try it on. For each year since, I computed the average age of the scorers that made it to the podium (top-3). Importantly, the winner of this award is not elected but rather results from having scored most goals. Hence, any changes that may appear over time won’t resemble changing preferences but rather an evolution of the game and/or scorers as such. For the running season, I include the current top-3 in the Golden-Shoe ranking: Barça’s 29-year-old Suárez, Napoli’s 28-year-old Higuaín and Benfica’s 31-year-old Jonas. The below chart indicates the average age of each season’s top-3.

Figure 1: Average age of "Golden Shoe" award top-3 (years on X-axis; age on Y-axis)

At first sight, there is no trend clearly emerging and what mainly grabs the eye are the many fluctuations. I next include a linear trend-line, hinting to some general trend. The important thing to note is the trend-line's upward slope, indicative of an increase in average age over time.

Figure 2: Average age of "Golden Shoe" award top-3 (years on X-axis; age on Y-axis).
Actual data in blue; linear trend-line in red.

The next graph depicts the non-linear trend-line that is closest to the data and which is readily supported by Excel. This trend-line starts off at slightly below 25 years of age and increases - let it be not monotonically - to nearly 29 years. 
Figure 3: Average age of "Golden Shoe" award top-3 (years on X-axis; age on Y-axis).
Actual data in blue; non-linear trend-line in red.

For those readers who know about statistics, the R2-values neither of the linear nor of the non-linear trend-line are high. However, the Spearman rank coefficient of season and age is 0.29 and significantly different from 0, suggesting that there is indeed an upward trend in topscorers' age over time. Even if from a strictly scientific viewpoint the evidence is limited, I could draw a parallel with global warming: the possibility of a few years difference over a less-than-50-year period may still be worth some consideration, wouldn’t you think? Here's what Aritz Aduriz (35) has to say about that.