By John Considine

What is the best way of discovering the top three teams from eight?  In terms of the number of games, the least costly method is to run a competition with single game eliminations.  Eight games over three matchdays.  Four quarter-finals, two semi-finals, a 3rd/4th place playoff, and a final.  A round-robin competition where teams play each other once requires 28 games over seven match days.  Four games played on every matchday.  It would be double if one was to have a home & away round-robin where teams played each other twice.  The elimination tournament has much lower costs in terms of games, i.e. 8 versus 28.  It also saves time in terms of match days as it requires only 4 where as the round-robin requires 7.

However, there are other costs.  The most important of these are the cost of inaccuracy.  The costs of making a mistake when determining the better teams.  Where luck plays a role then the better team does not always win.  This can be fatal in a single game elimination competition.  The "one & done" format allows a greater role for luck.  Less so in the round-robin format.

These two costs have to be balanced.  One way of doing this could be to combine the competition formats.  Split the eight teams into two groups and use a round robin format to decide the teams to advance to the elimination tournament.  It reduces the number of round robin games from 28 to 12 and it reduces the number of matchdays from 7 to 3.  It then requires some way of combining the two groups, e.g. the top 2 from each group contest the semi-finals followed by a 3rd/4th place playoff and a final.

The proposed 2027 Leinster Hurling competition follows the two groups format above.  It has an added feature in that the two groups are originally decided by a provisional ranking with the four stronger teams placed in one group.  All four teams in this group will end up contesting the elimination stage with the outcome of the round-robin determining the stage at which they enter.  The top 2 teams from the group enter at the semi-final stage.  The other two teams go to the quarter-final stage where they play the top 2 teams from the second division.  It will involve 18 games over 6 matchdays.

These are the discovery costs.  The costs of discovering the best team.  But there are other costs and benefits to consider.  Games are the product of the competition organiser.  They also have to consider the format that will extract the maximium amount from spectators at the games and broadcast revenue.

By Robbie Butler

Three weeks and four games. The first international break of 2026-27 was a long one, and it left me wondering: is there any other job where your employer must lend you to someone else, you might get hurt working for them, and your employer picks up the bill? Imagine a firm's best engineer called away for three weeks, coming back with a broken ankle, and the firm paying her wages while they recover. Economists would call this an externality: the national team gets the benefit and passes part of the cost on.

Liverpool have handed us a live case study. Alexander Isak scored for Sweden against Romania, then went home with a "minor" thigh problem. Two days later Cody Gakpo lasted 17 minutes in Belgrade before a heavy tackle left him with an ankle injury. Both are doubts for Sunday's game against Manchester City at Anfield, the same City that nearly signed Gakpo in the summer.

​FIFA does pay something. Since 2012, its Club Protection Programme has covered a club's fixed salary costs, when a player is injured on international duty. But it only applies once a player has been out for more than 28 days. Most international knocks never get that far, and nothing covers the points lost, the bonuses, or the hit to a player's value.
So why does football's system survive? It's written into players' contracts. Tradition does a lot of heavy lifting. Still, nobody building a business around £125m assets would design it this way, and on Sunday Liverpool may find out what four Nations League games cost them.

By Robbie Butler

Most labour markets have many employers competing for workers. Football, baseball and rugby are all great examples of this, as are many other team sports. This competition pushes wages toward what those players are actually worth.

In Ireland, inter-county GAA doesn't look like that. There is, in effect, one "buyer" of inter-county playing services. Except for a limited number of Gaelic football players moving to play in Australia, players can't transfer their skills to a rival league or team paying more, the way a professional footballer could. Those playing hurling have no route to Australia at all. This scenario is close to a textbook example of a monopsony: a single dominant buyer (the GAA) facing many sellers (the players). And the standard result of monopsony power is that "compensation" gets pushed below what a competitive market would offer.

While there is no pay in GAA, there is a form of funding through which players are compensated: a grant scheme, run through Sport Ireland in partnership with the GAA. This is required because the GAA remains an amateur association. The commitment required of players, however, is enormous, and they are effectively "professional" in all but name. That's exactly where a union earns its keep. A monopsony buyer, left alone, sets "pay" low. A union, acting as the single seller of labour and bargaining collectively, can push wages up.

The Gaelic Players Association (GPA) is this union. Formed in 1999, it doesn't negotiate wages in the conventional sense, since inter-county players remain amateur, but it does seek to have government funding increased. This week it sought an increase of €700 in player allowances to combat the ongoing "cost of living crisis" in Ireland. This would bring the total allowance from €1,800 to €2,500 per player, per year.

Should they succeed, it will be a textbook case of an industrial union in the sports domain successfully engaging in collective bargaining for the benefit of its members.

By David Butler

Last week my paper (with Robbie Butler and Carl Singleton) saw the light of day in the Journal of Economic Psychology.  It can be read here. In it we ask whether facing partisan audiences affects the execution of a skilled task. Our setting is penalty shootouts in the high-stakes environment of elite professional football. Since 17/18 prior to a shootout the referee tosses the coin to pick which end the kicks will be taken. One set of fans erupts, and a pundit usually declares that the team shooting towards its own supporters now has the edge. Sometimes this is deemed 'unfair'. 

The ideas are intuitive. Shoot towards a wall of your own supporters and you feel their encouragement... or maybe you feel extra stress and could choke? Shoot towards the away end and you face jeering, waving arms, banners and a barrage designed to put you off but maybe less pressure from expectant fans? But is it actually true? Are effects running any way? The short answer is that we found no evidence for either. 

We watched the footage and built a dataset of 1,322 penalties from 134 shootouts in the later rounds of the two big English cup competitions (the FA Cup and the EFL (League) Cup) across every season from 1992/93 to 2025/26.  For each shootout we went through archive video, mostly on YouTube, and coded by hand whether the takers were facing their own supporters or the opposition's. Importantly, this is directly observed, not assumed. Almost without exception, fans behind the goal try to distract opposition takers and goalkeepers with noise (and movement for takers), then fall quiet or supportive when their own player steps up. The headline result is a null one however. We find no statistically significant evidence that having supportive or hostile fans behind the goal changes the probability of scoring or of missing.

None of this means the authorities should scrap the coin toss. Even if it does not affect takers, it keeps the process visibly fair and heads off accusations of bias, which is worth something. And is the null effect of fans too surprising? Arguably not  - these players are elite footballers who are accustomed to such taunts and distractions. Of course, it is difficult to generalise to other settings, and we do note some small effects, but it is optimistic to think that elite pros seem to be able to ignore the noise.

By Riona Williams

With 13 second of normal time left in the 2014 All Ireland Hurling Final the referee awarded a scorable free shot to Tipperary.  The situation was extremely important as were Tipperary to convert the score it would put them a point up. The pressure on the umpire was immense as his decision would decide the whole match. The fact it was a free shot created a pause between the awarding of the free and the taking of the shot. This gave the umpire time to consider what to do if there was any doubt. Over one minute the Tipperary player took the shot.  At this stage there was only a handful of seconds left in the game.

When the free was taken, the sliotar travelled towards the post at considerable speed. From the umpire's position, deciding whether it had passed inside or outside the post was extremely difficult. This was where Hawkeye became important. Rather than relying entirely on what they could see, the umpire decided to send the decision to Hawkeye. The break in play allowed the umpire to realise the significance of the free and instead of potentially making the wrong decision when he was not entirely certain he knew he had to send it to Hawkeye.

If the umpire had correctly signalled wide without using Hawkeye, the immediate consequence would have been that the scores remained level. The game would have then gone to a replay, as happened after Hawkeye confirmed that the shot was wide. In this situation the umpire would have been correct, and there would have been no need for further intervention. However, from the umpire's perspective, the difficulty was that they could not be completely certain that the sliotar had gone wide. The fact that the correct decision was ultimately wide does not mean that the umpire could have known this with confidence at the time.

The more demanding situation arises if the umpire had incorrectly signalled a point. Because the match was being played under a system in which Hawkeye could be used to review decisions, an incorrect point would have had the potential to trigger an intervention. This could have led to there being confusion and considerable controversy before the final outcome was established. The crucial point that is the possibility of such a situation was precisely why the umpire was correct to use Hawkeye when there was any doubt.

The significance of this incident is that Hawkeye didn't replace the umpire. Instead, it supported the umpire at precisely the point where human judgment was most difficult. The umpire recognised that the decision was potentially decisive and that there was uncertainty about what they had seen. By referring the decision to Hawk-Eye, they avoided simply guessing and allowed the technology to provide a more certain answer. In this case, the short pause was a small price to pay for ensuring that the final shot of the game was judged as accurately as possible.

Hawkeye works because the official recognises when a decision is important enough, and uncertain enough, to justify using it. The 2014 All-Ireland Final provides a particularly clear example: a tied game, a final free, an uncertain decision, and technology capable of resolving that uncertainty. The umpire's decision to use Hawk-Eye ensured that the match was decided by what happened to the sliotar rather than by an uncertain human judgement.

(Riona Williams is a second year student of economics in UCC.)

By Robbie Butler

Maybe one of the reasons I enjoy football and economics is because both are fundamentally about making decisions. Maybe this is why I have written so much about VAR, my dislike of the technology and how it has changed how the sport is consumed. In my opinion, things reached a new low on Sunday with the awarding of Erling Haaland's goal against Manchester United.

And a new low is reaching some depths. I didn't think Luis Diaz's disallowed goal against Spurs in 2023 could be surpassed but we may have a candidate now.

As Roy Keane alluded to in post-match commentary, anyone who has played the game knows Enzo Fernandez (below) is interfering with play, and as a result, offside. The match officials concluded the same in real-time. The assistant referee raised his flag immediately after Haaland scored and the goal was deemed offside.

I don't believe a single football fan would have objected to this decision had the game restarted with a free-kick, as has been the case for about a hundred years when such a scenario occurs. Enter VAR. The team in Stockley Park interpreted the rules to mean Fernandez was not offside and the goal stands. It is not how offside is meant to work, nor, in my opinion, the correct interpretation of the wording.

But to compound this mistake, Pro Ref, formerly known as the PGMOL or Professional Game Match Officials Limited, issued an apologetic statement, saying: "the VAR did not recognise the likely impact of his position and should have recommended an on-field review." This statement fails to state that there was an "on-field review". It was done in real time. By the match officials. Who made the right decision.

The VAR then got involved in something where it was not "clear" or "obvious" that a mistake had been made and ended up overturning the correct decision. Today's statement confirms that, and also confirms what a growing number of people believe: VAR makes the game worse, not better, and introduces mistakes that would not be there had the technology not been adopted. It effectively does the exact opposite of what it was intended to do.

By Robbie Butler

I recently co-authored a paper that has been published in Oxford Economic Papers. The sport under examination - horse racing. Using data from racing in Great Britain we investiagte instances were jockeys are banned for percived non-trying. The abstract is below. ​

A fundamental challenge in the principal–agent relationships is ensuring agents exert effort when their actions are not perfectly observable. We examine a natural experiment in effort enforcement in British horseracing, where jockeys are required not only to try to win but to be visibly seen to be trying. Jockeys who appear not to be trying face suspensions and significant penalties in terms of income forgone. Using a dataset of some 60,000 races involving more than 1,500 jockeys and 17,000 horses, we test whether these apparent non-trying events represent attempts to fix a race or influence future races. We find no evidence to support either hypothesis. Performance neither improves in the races before, nor deteriorates in the races after, a ban. Punishments for non-trying may nonetheless be rational as a deterrence mechanism–making an example of a small number of jockeys serves as a motivation for the rest: pour encourager les autres.

The full paper is available here.

By Robbie Butler

Many papers in sports economic focus on referee behaviour. This work rests on a load-bearing assumption: referees correctly apply the laws of the game. Whether one measures  home bias, added-time effects, card thresholds or anything else the referee does the core questions ask there is a deviation from that neutral baseline? It works because we assume referees are entirely professional and impartial.

This week gave the field an unusually pointed real-world test of that assumption. On the Jamie Vardy's Having a Party podcast, former Premier League referee Mike Dean, who officiated 553 top-flight matches in the English Premier League over more than two decades, described private "games" he played during live matches.

Dean spoke about how he would see how long he could avoid blowing his whistle (once reportedly 20–25 minutes) and how long he could avoid leaving the centre circle (around 15 minutes on one occasion).

While many former referees have rebutted these claims, the fact Dean himself is saying he did this is extraordinary. It also occurred to me that a past paper of my included Mike Dean in the sample. While his admission did not address what we tested, one has to wonder where these “games” ended?

By Shona Finn

Hawk-Eye is a ball-location and score-detection technology that is used to help referees make more accurate decisions during matches. It combines specialised cameras and sensors with computer software that calculates the exact path of the ball or sliotar. The software analyses information from multiple camera angles in real time before determining whether the ball has passed inside or outside the posts. Similar technologies are also used in sports such as tennis and soccer to make accurate line and goal decisions.
 
This post will ignore the financial cost of the system to the GAA.  Instead, it will focus on other costs involved in the decision-making system.  Hawk-Eye demonstrates cost minimisation and trade-offs by balancing two competing costs: delaying the game and making an incorrect decision. Reviews pause play, but avoiding reviews increases the risk of wrong decisions. The optimisation challenge is finding the balance that minimises the combined cost of these two outcomes.
 
At one extreme, imagine Hawk-Eye reviewed every shot, incorrect decisions would become rare, but constant interruptions would slow the game, disrupt momentum, frustrate spectators and increase operating costs.  In this situation, the cost of incorrect decisions is very low, but the cost of delay becomes extremely high.
 
At the opposite extreme, imagine Hawk-Eye was never used or not available. The game would continue without interruptions, making the cost of delay almost zero.  However, the likelihood of incorrect decisions would increase, particularly in close calls where it is difficult for umpires to judge accurately.  A wrongly awarded or disallowed point could affect the result of a match, leading to frustration for all. The cost of delay is minimised, but the cost of incorrect decisions becomes very high.
 
The benefits of using Hawk-Eye also depend on the score margin during the game. If one team leads by 20 points, one incorrect score is unlikely to affect the result, so the benefit of a review may be lower than the cost of delaying play. However, if the teams are level or separated by only one point, the value of making the correct decision is much greater. A single point could decide the winner, meaning the cost of an incorrect decision is much higher.
 
Timing also affects the trade-off. Early in a match, teams have more time to recover from an incorrect decision. Near the end, however, a single point may decide the result, making the benefit of a review greater than the cost of a short delay. This demonstrates the economic principle of trade-offs, where the best decision is the one that minimises the total decision-making costs rather than eliminating only one type of cost.

(Shona is a second year student of economics in UCC.)

By Robbie Butler

The annual gathering of the European Sport Economics Association was held in Zurich from the 19th to the 21st of August.

Hosted by the University of Zurich, the event was preceded by the PhD Workshop, which I had the pleasure of speaking at. More than a dozen students, registered on various postgraduate and PhD programmes across the world, enjoyed two and a half days of instruction from leading sports economists and other academics in the field, including Carl Singleton (University of Stirling), Alex Krumer (Molde University College), Dennis Coates (University of Maryland), Ulrich Brandes (ETH Zurich), and Brian Soebbing (University of Alberta), as well as local hosts Helmut Dietl and Raphael Flepp.

The conference began on Wednesday afternoon with the traditional panel discussion entitled "Innovation in Sports," followed by the first set of parallel sessions. The evening ended with a spectacular, panoramic view of Zurich from high up in the university. In total, almost 80 papers were presented, across a multitude of sports and topics in economics, with nearly 100 participants engaging in the conference in one form or another.

The annual general meeting was held in an iconic room that had once hosted Winston Churchill in 1946, where he spoke of a "United States of Europe" as the threat of Soviet expansion emerged on the European continent.

The highlight of the three days for me was a visit to the FIFA Museum in the heart of Zurich. Delegates were given a guided tour of the artefacts on show from each World Cup tournament to date (excluding 2026). Items and memorabilia included jerseys worn by Diego Maradona (World Cup 1986) and Lothar Matthäus (World Cup 1990), as well as match balls from finals, football boots from the 1954 final, and even the pipe of a World Cup-winning manager, now nearly 100 years old.
​
The 2027 conference — the 18th to date — will be held at the University of Poznań, Poland, in the last week of August.