Our blog will now take its customary Christmas break and return on Monday the 27th of January 2025.

We would like to wish all our readers a very Happy Christmas and best wishes for 2025.

By Robbie Butler

We are delighted to announce we have recently been published in Kyklos. Our paper - "The Baby Club: Paternity and Performance in a High-Pressure Setting" - attempts to understand worker performance before and after the birth of a child. Specifically, we look at male footballers and try to assess how child birth effects their performance on the pitch. 

Our paper offers new insights into fatherhood by asking if the onset of paternity changes workplace productivity. We do this in the well-monitored and high-pressure setting of professional football using a novel dataset that matches 115 birth disclosures to the performance of 96 players.

Our empirical approach involves specifying a performance equation for a suite of match-level performance statistics and estimating OLS and Poisson fixed-effect panel regressions. We find a negative correlation between fatherhood and collaborative performance as measured by expected assists—a player's ability to create goalscoring opportunities. We also report negative effects for the perinatal period for expected assists and passing measures.

There is no evidence of performance changes resulting from expectancy news. As negative performance effects are observed in a context of ‘superstar wages’, this raises concerns for high-pressure labour markets where workers are remunerated less but have low uptake of leave entitlements or where paternity leave is culturally taboo.

​The full paper, available in open access, can be found here.

By David Butler

This week a new paper of ours was published in Frontiers in Behavioral Economics as part of a Sports Economics special issue co-edited by Raphael Flepp, Romain Gauriot and Carl Singleton. 

With colleagues Alex Farnell and Rob Simmons, we ask if elite football clubs use advanced performance metrics to pay players? Given the data revolution that has occurred in football over the past years, one is left wondering if, or to what extent, metrics such as expected goals or expected assists etc. are used to determine contract awards. 

Using data from Capology and Fbref, our take-away message is that clubs are using very few of these advanced  statistics in salary negotiations. Based on the results, it seems plausible that executives use individual inputs to team output as a key determining factor in pay  - individual aspects of performance may treated as a black box so to speak and adding to team points may be the guiding metric. 

The full paper can be abstract here and the abstract is below

Abstract
"​Labor economists aspire to understand how workers' productivity impacts pay. While professional football is a well-established domain to explore this relationship, so far, research has relied on basic productivity measures. Football is now awash with advanced and granular performance metrics that can allow a deeper understanding of the pay-performance relationship. We specify a salary model considering the newly available data and use sophisticated performance measures to explain contracted salaries in the English Premier League and Italian Serie A. We make a methodological breakthrough by identifying a sample of players who are in the first year of a new contract only. This results in a much tighter relationship between pay and performance. We estimate different salary equations using both basic and advanced performance statistics. Our main findings are, first, that few of our advanced performance metrics help to explain player salary and, second, that there is misalignment between individual performance determinants of team points and player salaries."

By Robbie Butler

It has been interesting to read recently that IFAB, football’s custodians of the laws of the sport, are considering moves “to deal more effectively with goalkeepers who unfairly hold the ball for too long”. This statement appears in the 2024/25 Laws of the Game and seeks to target one of the most unapplied/broken rules, in my opinion, in the game.

What is more interesting is that goalkeepers holding the ball is currently covered by Law 12 – Fouls and Misconduct. The rule states that “An indirect free kick is awarded if a goalkeeper, inside their penalty area, commits any of the following offences: controls the ball with the hand/arm for more than six seconds before releasing it”.

Anyone that watches football will know this “six second rule”, introduced around 20 years ago (I think) is almost never applied. Prior to the introduction of the rule, goalkeepers would simply hold the ball for long periods of time. The final minutes of the 1992 UEFA European Championship Final between Denmark and Germany are a must-watch for those that want to witness how dreadful this, and the back-pass to the goalkeeper, were prior to important rule changes.

Goalkeepers were permitted to take 4 steps with the ball in hand. They would often then drop the ball, before picking it up again, to continue. This loophole was closed, but the steps rule remained. Timing was simply at the referee’s discretion.

This changed with the 6-second rule. But does this rule actually matter?

I recently watched an English Premier League game and counted to 34 seconds before the goalkeeper, who had caught a cross, decided to release the ball from his hands.

The question, for me at least, is why do referee’s ignore this formal rule? Will trialing a new punishment e.g. awarding a corner instead of an indirect free kick, increase application of this rule? Maybe, but why is the current punishment not applied?

Another interesting dynamic of this rule is that players rarely seek it’s application. For example, during the 34 seconds that the goalkeeper had the ball in his hands, not a single player visibly appealed for an indirect free kick to be awarded. They were instead happy to wait for the ball to return to play.

By Robbie Butler

Our latest effort has just been published in European Sport Management Quarterly.

The inspiration for this work has it's origins in a television report I watched back in 2021 - during lockdown - when I learned Welsh football fans had to travel to Baku to watch their countries first two group games, and then travel onto Rome for game 3, before moving to Amsterdam in the Round of 16. A total travel distance of more than 8,000km.

The new research explores the environmental impact of air travel during two international football tournaments – the Mens FIFA World Cup and UEFA European Championships – by estimating CO2 emissions from the movements of competing teams. Specifically, we consider how travel associated with major international football tournaments has evolved over time.

We access data for 895 international football matches (June 1990–July 2024). Our data focuses on the travel associated with players and management travelling to, and within, host countries and includes the movement of more than 16,000 people. Estimates of CO2 emissions per passenger (per kilometre) are used to examine the environmental impact of travelling to host countries, and competing, during the tournament.

Our findings are the first attempt to demonstrate the CO2 emissions from team travel to and within major football tournaments. Despite improvements in aeroplane fuel efficiency, the carbon footprint from team travel has generally not been reduced. The decision to expand these tournaments, locate matches across large geographic areas, and use co-host countries has exacerbated this issue.

Our findings can support sustainability initiatives and could be adopted by both federations to curb CO2 emissions which appear to undermine genuine attempts to become more environmentally friendly. The full paper can be found here. 

By Robbie Butler

This website has regualrly addressed competion in sport, and by extension competitve balance and uncertainty of outcome. These are central planks of sports economics. Racing this weekend at Punchestown gives us a great example of an extreme form of this - or lack there of.

It's safe to assume - barring an extreme weather event or falling - owner Michael O'Leary and trainer Gordon Elliott - will have at least one winner. The entries for the 1pm race - the ​3m (3m51y) Boodles Florida Pearl Novice Chase (Grade 2) (5yo+) - are presented below.

The race is on free-to-air television and showcases three up-and-cmoing elite novice chasers. The issue of course is that all three are owned by Ryanair boss Michael O'Leary and trained by Gordon Eillott. The "competition" in the race is clearly amongst the jockeys: Jody McGarvey, Sam Ewing and Jack Kennedy.

Some have questioned whether the race should be a betting event. No odds are available yet - that is not unusal this far in advance. Betting will occur and punters can be safe in the knowledge this will be a true run race. It's happened before with the same owner. It's just not very exciting for the sport. Especially at the elite level. 

By David Butler

Previously, I have written about a recurrent issue in Football – players withdrawing from international duty.  This is a hot topic in the media again given the recent high-profile withdrawal for the English squad for upcoming Nations League matches. There have been various other withdrawals outside of the England team.

The dynamic of footballers jeopardising their ability to fulfil contracted duties (due to injury risk for their national team) is interesting to think about from a labour market and legal perspective. It highlights the uniqueness again of sports labour markets - is there any other labour market where employers holding contractual rights knowingly release their workers to perform the same duties for a third-party?

As these athletes can decline an international call up, themselves and the clubs are essentially choosing to take on risk. Given that most professional contracts have performance bonuses built in – and that these contracts have significantly increased in value over the years – I wonder are players now evaluating the personal financial cost to a greater degree, especially for repackaged friendlies.

Previously, I have documented the frequency of injury on international duty – it is not trivial. These injuries  appear random on the surface and carry a cost both in terms of missed performances and rehabilitation.  

​For example, to quote my piece in  September 2022

“Due to injuries sustained during international performances last week the following players were substituted: Nathan Patterson, Memphis Depay, Teun Koopmeiners, Richard Almeyda, Andreas Weimann, Jules Koundé, Ivan Perisic, Kiefer Moore, Jere Uronen, Miralem Pjanic, Giorgi Aburjania, Aleksandar Mitrovic, Kieran Tierney, Aaron Hickey, Miguel Vítor, Jakub Brabec, Anastasios Bakasetas, Zan Karnicnik, Hrayr Mkoyan, Eden Hazard, Steven Berghuis, Dominykas Barauskas, Marcelo Brozovic, van Bakhar and Eduards Emsis.”

I wonder who will be on the list next week?

For detailed insights on this general topic see Jeffrey Cisyk and Pascal Courty recent paper in the Journal of Sports Economics - An Economic Approach to Sports Injury Policies

By John Considine

It is not unusual to hear those opposed to statistical decision making in sport arguing that there is a greater need for situational analysis.  One of the more influential minds of the twentieth century, Karl Popper, made a similar suggestion about the benefits of situational analysis for social sciences like economics.  Popper later reviewed his position and admitted a greater role for statistics in economics.  There is an interesting paper in the latest issue of the European Economic Review that combines the two approaches in the context of professional road cycling.
 
Julian Mattes and David Piazola outline a range of scenarios with useful illustrations and associated payoff matrices.  An example is presented below.  The energy levels of each cyclist is printed above them.  There is a 2-unit energy gap to the leader from the black team.  It will cost 2 units to attack (A) and close the gap.  It will cost 1-unit to mark (M), or slipstream, a cyclist making an attack.  In situations 1.a and 1.c the cyclists will have an equal likelihood of victory.  Situation 1.b is the well known free rider problem applied to public goods.  The equilibrium for the pink and blue riders is [0,0].

These scenarios are used to examine if cyclists compete or cooperate.  The combination of the two strategies is known as coopetition.
 
Mattes and Piazola make a brave effort to test their predicted outcomes with the data.  Using data on the time gaps at the finish line, they infer what must have happened during the race.  This is not ideal but it is the best that can be achieved with the data.  The data set captures 40 years of the three Grand Tours and other races.
 
An alternative, or complementary, approach is to gather data on what actually happened during each stage of a race.  This issue has been examined here previously.  Jane O'Sullivan examined coopetition in the 2021 Giro D'Italia.

The specifics of what happens in a given situation can make it difficult to conduct statistical analysis.  Consider the situation involving three cyclist in Stage 15 of 2023 La Vuelta.  The changing dynamics of the last 10km of the stage are sketched here in a YouTube video.  This three person contest is very different to other three person contests.  Why?  Because it involves Rui Costa.  Those who follow proffessional road cycling know that Rui Costa is probably the best illustration of a "free-rider".  His approach is descripted by a fellow rider as "looking at other riders to do the work and saving as much energy as possible".  This description is included in a brilliant book by Peter Cossins called Full Gas! How the race was won: Tactics from inside the peloton.  The book is a superb account of various situational situations in cycling.  It would be worth reading it as a companion piece to the more statistical European Economic Review article by Mattes and Piazola.

By David Butler

Yesterday Ignacio Palacios-Huerta wrote an interesting article in the Guardian on the decision to  raise the number of fixtures in European competitions. In a nutshell, the argument is that limiting supply can lead to long-run benefits. UEFA need to carefully consider satisfying demand. 

After 3 game weeks (54 matches) under the new format what do the live attendance numbers looks like?

- For the new Champions League the total estimate of attendance is 2,179,399, with an average of 40,359 attendees per match (median 41,577).

- For the new Europa League I can only access data for the first two rounds (36 matches). The total estimate of attendance is 975,084 with an average of 27,085 per match (median 22,604).

It will be interesting to see how the attendance numbers and viewer figures at the end of the tournament compare to past ones.  

By Robbie Butler

Our latest paper, will soon be available online in European Sport Management Quarterly. It seeks to understand the movement of teams at major international football tournaments through time and assess whether these tournaments are aligned with the environmental goals of their organisers. 

Using data from 1990-2024 we attempt to provide a baseline measure of the impact these movements have had on carbon dioxide emissions. All data is publicly available and uses distance calculator and historical football archives to track teams.

This longitudinal study allows us to consider both the expansion of these competitions overtime and improvements in aeroplane technology, with long-haul aircraft now substantially better at reducing CO2 emissions than they were in the past.

Once the paper is available online, we will provide a further update and link.