#FootballDataScience: Analyzing Premier League Spending
Introduction
Many have sought to examine Premier League spending in the past, especially fans of different clubs attempting to attack rivals over their recent spending or simply defending their own.
This post however will attempt to use data analysis in order to identify key spending trends in the Premier league as a whole. As the Premier League is one of the most watched leagues in the world, television revenue for PL clubs allows them to consistently outspend every other football league in Europe. As such, in many ways the Premier League’s spending habits can be seen as both symptomatic and trend-setting of wider spending trends observed in the football market today.
I used data obtained from Transfermarkt found in a public GitHub repository in order to examine and analyze Premier League transfer spending. All transfer values are obtained via Transfermarkt unless otherwise expressed. I used Python in order to process and visualise data (Pandas, Seaborn), as well as Tableau for visualising spending clubs. The associated Jupyter Notebook is available on my Github.
Defenders, defenders, defenders
Which positions are have become the most valued over time by PL clubs? These following charts attempt to visualize the evolution of mean spend (£m) on each position. The 2014/2015 and 2015/2016 seasons have seen heavy spending on wingers (nearly £40m on average for left and right wingers), with world-class wingers such as Di Maria, Mata, Alexis Sanchez, Martial and Payet all moving to PL clubs. Attacking players have been known for attracting higher transfer fees, as clubs tend to value goal-scoring players much higher than midfielders or defenders. According to Transfermarkt, 9 out of the 10 world record transfers are indeed wingers or strikers (the last being Paul Pogba, an attacking-minded midfielder).
However, recent signings made by Premier League clubs shows a distinct change in their spending strategies. Clubs have rapidly augmented their spending on defensive players, as demonstrated by the recent purchases of center-backs Virgil Van Dijk (£76m) and Harry Maguire (£78m), the plethora of fullbacks bought by Manchester City (Kyle Walker (£47m), Danilo (£33m), Joao Cancelo (£58m) and Benjamin Mendy (£52m) just to name a few) and goalkeepers such as Alisson (£56m), Kepa (£72m), and Ederson (£35m).
The emergence of players like Liverpool’s Trent Alexander-Arnold and most recently Bayern Munich’s Alphonso Davies, who both figure in the top 10 of the recently published CIES Football Observatory value list, demonstrate the increased demand for complete fullbacks as they become crucial components of the modern game. Both players are capable in both offensive and defensive duties in addition to technicality and set-piece ability.
Some may be surprised that fullbacks like Trent Alexander-Arnold (€171.1m) are more valued than star attacking players such as Marcus Rashford (€152.3m), Mohammed Salah (€144.9m) and Sadio Mane (€139.2m). However, long-term trends of PL spending on defenders may indicate otherwise, as seen in the chart below.
Finding the next Haaland
What types of players are PL clubs spending on?
Do PL clubs prefer spending heavily on players entering their prime (said to be approximately around 25–27 years of age), on more experienced league veterans or on youth?
Common belief indicates that clubs are shifting to the latter, as young players showing great potential are increasingly being sold around Europe for exorbitant amounts (such as Mbappe, sold for £130m at only 20 years old, or Joao Felix, sold at £113m at 19). However, trends in data indicates that this may not necessarily be the case. The below chart shows the difference in spending between 2010–2014 (blue) and 2015–2020 (orange).
A first insight we obtain from these visualizations is the 116.9% inflation between the two periods, with spending growing from £4.12bn to a behemoth £8.95bn, explained by the steady rise of TV revenue in the Premier League.
However, we can see that inflation across ages is different- inflation seems to have been the highest at ages 24 and 28, and the lowest at ages 17 and 35.
The below chart shows inflation in spending for each players of each age between the 2010–2014 period and the 2015–2020 period. Inflation levels for ages 32 and up have been excluded, as 2010–2014 levels were near zero and hence generated extremely high inflation levels, hereby skewing the visualization.
Despite our original instincts, it would seem that the highest inflation has in fact occurred for players aged 27 and 28. This demonstrates that clubs are not necessarily spending more on youth, as prime players seem to be commanding higher fees than their younger counterparts. Despite this, it would seem that there has also been significant and consistent inflation for 18 and 19 year old players as well as players between 22 and 24. This is due to the fact that while 18 and 19 year olds are still young enough to be developed through loan spells (as is the business model of several PL teams), 22 to 24 year olds may be considered mature enough to provide a direct impact to the first team, and hence command higher fees whilst representing better investments than older players.
However, another pattern emerged in the data, which provides nuance to this renewed investment in young players. Despite the fact that clubs are now spending high amounts on younger players, the average age of players arriving in PL clubs has risen from 23 to more than 25 between 2010/2011 and 2019/2020.
This indicates that clubs aren’t necessarily spending on higher volume of young players, but are instead choosing to invest in highly-rated, higher-quality youngsters that they would be able to quickly integrate into first-team plans, as demonstrated by successful signings like Aaron Wan-Bissaka and Wilfred Ndidi. Spending on young players may also be influenced by high-profile transfers involving international youth, such as Kepa Arrizabalaga’s arrival at Chelsea for £72m at only 23, or Christian Pulisic’s arrival at the same club for £58m at only 20.
Homegrown Talent?
Where are players being sourced from when PL clubs look to buy? Which clubs and leagues have profited the most from PL money?
Interrogating the supply-chain of talent is a crucial aspect of understanding spending patterns in the Premier league, as clubs shift their scouting and spending to players outside of the English pyramid. The two following charts show the top 20 selling clubs who have sold players to the Premier League in the 2014/2015 and 2019/2020 seasons.
Out of the top 20 selling clubs of 2014/2015, 8 are from the English Premier League and Championship, with sales to Premier League Clubs totaling £315.7m. In addition to English clubs, Spanish clubs also feature heavily, with 5 Spanish clubs (including the “Big Three”) with sales totaling £248.67m. We can see the success of Southampton’s strategy to develop and sell players for profit, with sales of the likes of Luke Shaw (£33.75m), Adam Lallana (£27.9m), Dejan Lovren (£22.77m) and Calum Chambers (£18.21m).
Out of the top 20 selling clubs in 2019/2020, only 6 are from the English leagues, with sales totaling only £224.37m, marking a 28.93% decrease of domestic transfers over 5 years. French clubs feature heavily, with 4clubs included in the top 20 and sales totaling £124.2m. Leicester has profited the most off of English clubs through a single sale, Harry Maguire to Manchester United for £78.30m.
Comparing the two rankings demonstrates that their has been a distinctive shift in investment strategies from PL clubs, as they are increasingly looking to spend on foreign talent rather than players already playing in the English leagues. This could be due to the increased premium on English players as well as teams like Southampton becoming more reluctant to selling players to their domestic rivals, choosing instead to retain talent in order to improve their league position.
Predictions
We have therefore been able to demonstrate key shifts in PL spending over the course of the last decade. Any article attempting exploratory analysis of such data would be incomplete without providing predictions based on this insight, so the object of this conclusion will be to provide several predictions based on the data I observed. Let’s recap the insights that we have so far:
- Defenders are rapidly gaining in value, as teams are looking to buy elite defenders that help them play with high-pressing tactics
- Despite our original instincts, inflation in player prices has significantly risen for players aged 27 or 28 then any other age group, providing a more nuanced image of PL clubs’ transfer strategies.
- The average age of a transfer has steadily increased over the years, indicating that clubs are becoming more averse to risk. This indicates that clubs prefer to spend both on quality youth in addition to buying “prime” players who will have a direct impact on the first team
- Whilst PL clubs preferred to invest in PL-proven players, they now prefer to invest abroad in order to obtain lower transfer fees and avoid the English “premium” tax and reluctance of clubs selling to domestic rivals
- French clubs have superseded Spanish clubs in providing talent to the PL, as Ligue 1 continues to develop youth with great potential at lower costs
Using these insights, I will formulate the following predictions:
- Elite defenders’ prices are going to continue inflating, as teams are facing increased competition over elite defenders allowing them to develop modern tactics
- Fullbacks will be a particular area of unprecedented investment, as the combination of attacking and defensive output has become extremely valuable to clubs looking to play attacking football
- The current COVID-19 pandemic and it’s associated economic uncertainty will mean that clubs will be more averse to risk when gambling on youth which they cannot quickly integrate into the first team, and will prefer to buy “prime” players that can quickly improve their league standing even if not a economic investment
- Despite clubs increasing their investments in foreign players, PL clubs may have extra hurdles when signing European talents as a result of Brexit- this may cause a reversion to 2014/2015 levels of spending on domestic players
Thank you for reading! If you want to see more football analysis articles, don’t hesitate to subscribe on Medium, and don’t forget to check out my code on Github.
