Author Archives: Ian Mitchell

Time to shine a light on the 3pm kick-off blackout

Rather than helping football fans and lower-league clubs, the ban on televising games that kick off at 3pm on a Saturday is actually making life harder for fans, and needlessly pushing up prices.

Here we look at why the ban on televised games at 3pm on a Saturday is no longer needed.

Half of Premier League to be televised from 2018 

Today the Premier League announced that from the 2018-19 season, it would televise a minimum of 190 Premier league games. Half of all games in the season. However, it comes with an implicit catch – since these games cannot be televised at 3pm on a Saturday, it means that more top-flight games will need to be show at other times, which many fans see as a major inconvenience.

Remarkably, rather than welcoming the fact they can watch more of their teams’ games than ever, many fans see this as a further erosion of the game they knew and love. The Football Supporters Federation have expressed their concern and opposition to further increases.

The origins of the TV Blackout

The ban on televising games between the hours of 2.45pm and 5.15pm on a Saturday dates back to 1960 when it was adopted by Football League Chairman, to protect attendances at smaller clubs.

Three reasons the rationale is weak if not dead

The original argument was that fans would not attend lower league games if they could watch top-flight games on TV. But there are at least three reasons why this rationale is weak.

First, no other industry would be allowed to club together and protect each other from competition. Imagine if cinemas agreed that no-one could rent a film at home on a Wednesday night because it would damage local movie theaters.

Second, lower and non-league clubs would not be less viable. Many other countries do without the ban – this includes France, Germany, Italy and Spain. (the source is from 2011). Many lower league football clubs now generate millions in income. Some still survive by the skin of their teeth – but this rather reinforces the point that the viability of clubs in the lower divisions has not been cured by higher income alone.

Third, assuming that lower league football and televised premier league games are in direct competition, then the ultimate result for match-attending fans would be that ticket prices would fall as owners are forced to compete. This might reverse some of their rises over the past twenty years. A “National League South”  club near me charges £15 a game. Perhaps it needs a bit more competition.

Tradition can obscures fans’ interests

Like so many things in football, the economic truth is obscured and the Premier League can take advantage of certain traditions (like the 3pm blackout) while ignoring others (like moving 3pm kick-offs).

The 3pm black-out rule is archaic and works against fans. Allowing games on television on a Saturday at 3pm would increase the number of games fans could see, lower prices for everyone, and give more people access to the game.  See the light, it’s time to end the blackout.

Quantifying loyalty as Tottenham move to 40,000 season tickets

Tottenham Hotspur have just sold out 18,500 new season tickets to members on the waiting list.  But with 40,000 season ticket-holders in total, how will the club decide who gets to high demand games like cup semis and finals?

We look at the problem the club faces and propose an approach that would resolve it by combining the currently separate loyalty systems for the club’s ‘members’ and season ticket holders.

You wait 10 years for a season ticket then 18,500 turn up at once

Some members have waited over a decade to get a season ticket. Tottenham’s move to Wembley for the 17/18 season, and their subsequent move to a new 61,559 seater stadium, have enabled the club to release the extra season tickets. To get one ‘bronze’ members joined the list and paid an extra £15 per year more than other Spurs members. 

During the last few weeks, the club have made season tickets available to tranches on the waiting list – starting with those at the very top; moving on to everyone up to 25,000; then for those up to 40,000; then 55,000 and beyond. In end-year reporting, the club announced that the waiting list had grown to 62,300. It’s striking that only around 1 in 3 took up the opportunity and while some have waited over ten years, others have waited less than two.

The process of giving out season tickets has been rushed. Fans waited years but then had to decide on a £1k per year commitment in literally a matter of days. A key issue for the club now is how it treats its season ticket holders and members.

Who will go to be able to go cup finals?

Loyalty points are key for the big games. For many away games there are only 3,000 tickets – even season ticket holders need several hundred loyalty points and they are also used to allocate members tickets for cup semis or finals.

A Tottenham season ticket will no longer guarantee a ticket to a Cup Final, nor even for the 33,000 semi-final tickets. A major challenge that now faces Tottenham and its new season ticket holders is how to decide who gets tickets for high-demand games like cups, or games away from home.

The default position may be that new season ticket holders are bottom of the list (see chart below). That may be fair (though some have been to more games than season ticket holders), but then who should get priority within new season ticket holders with the same (very low) number of loyalty points?

The club has already caused controversy on its approach to season tickets. In 2012, with no consultation, it introduced a 5 year cut-off for loyalty points so that season ticket holders who attended games 25, 15, or 6 years ago saw no recognition in their loyalty points. So, further wholesale changes seem unlikely.   

What are the options?

The default position is that new season ticket holders start again with no loyalty points – so, 18,500 new season ticket holders, some who have been to a hundred games in the last five years, and some who have only been to a handful, will be equalised. The club appear to be suggesting ballots to allocate tickets. This is clearly unfair.

Carrying forward members’ loyalty points would also be unfair – as a currency, these are more liberally awarded, and less valuable than season ticket loyalty points. It would be ridiculous for a new season ticket holder who has been to half of games to jump ahead of one that has been to all of them.

Another option is to ‘freeze’ new season ticket holders existing (member) loyalty points, and use these to prioritise until the season ticket points are more differentiated. A similar option is to use new season ticket holders (last) position on the waiting list (which would represent how long they had waited). Both of these have the feature of being immediately available and the Supporters’ Trust have supported a weighted version.

Our proposed option is to create a single loyalty point system by boosting season ticket holder loyalty points. To achieve this, season ticket holders should be awarded the loyalty points they would have received for attending games as members. This could be done based on the games they actually attended (which corrects for when season ticket holders sold their ticket) – or it could be a simpler per-year award. Based on the mix of games, this would boost season ticket-holders’ loyalty points by some 60 points for each year they held the ticket. If all season ticket holders got 300 extra points (a few wouldn’t as they’ve had one for less than 5 years), and if we assume the new season ticket holders have the same distribution of loyalty points as all members, the resulting distribution would look as follows:

There is some overlap for the small proportion of members (1.5%) that have been to more games than season ticket holders (ie earned over 300 loyalty points) but otherwise, existing season ticket holders have a clear advantage.

The way forward

Overall then, this proposal would:  

– solve the problem of how to assess the loyalty of new season ticket holders fairly

– treat season ticket holders and members consistently;

– reward the most loyal supporters; and

– create in incentives to attend less popular games.

With only one month to go till the new season, and no word from the club on its plans, it seems unlikely there will be any consultation on the approach. The Supporters Trust have rightly criticized the club for mis-information, a lack of transparency and no consultation on dramatic changes to both season ticket holders and members.

This proposal would require some effort on the part of the club to go back and calculate loyalty points – but the benefits would be a single clear system of loyalty points that avoids further major reforms.


Ian Mitchell – Director of Football Economics (and, full disclosure, Spurs ‘bronze’ member)

With excellent advice from Paul Lewis (Spurs season ticket holder of 19 years with a son on the waiting list for the last 5 years)

Premier League Managers – Trends and Club Pedigree

We’ve produced a new report looking back at the mix of managers in the Premier League since it began in 1992, and whether those managers had played for the club.  The report provides the full analysis but below we give the main conclusions.

The mix of nationalities of Premier League managers has transformed beyond recognition in the past 20 years, and is still changing:

  • UK managers were a minority in the Premier League for the first time in 2015.
  • For English managers, numbers declined substantially over the first 20 years of the Premier League (from 17 in 1992 to 4 in 2012) – see chart below
  • For the combined number of Northern Irish, Scottish and Welsh managers the major falls have happened in the last four years (from 8 in 2012 to 3 in 2016) – see chart below

Blog chart Prem managers 92-16

  • The era when most top managers had played for their club appears to be long gone. However, the number of Premier League managers who played for the club has risen from zero in 2013 to 3 in 2016

Blog - pedigree chart Aug 16

As an indicator of progression from playing to managing at the top level of club football, these results suggest that for the vast majority of clubs, former players have not been developed sufficiently to take over the managerial reigns.

We think manager development is key for clubs and countries, and will be looking more at the topic in future analysis.



Premier League Managers – nationality and club playing history since 1992

In this analysis, we look over the life of the Premier League to consider the extent to which top-flight clubs still look to former players to manage their club and at the shifting mix of nationalities of top-flight managers with a focus on English managers.

See our blog post on the conclusions and below is the full report:

Premier League Managers Nationality and Club Background


England’s Euro 2016 TV audience a record low

In the past few weeks, the Broadcasters’ Audience Research Board (BARB) have released the viewing figures for England’s Euro 2016 campaign and in this blog we compare them to the figures from previous tournaments which we compiled for our comprehensive research into audience drivers.

In short, the news is not good for the Football Association or England’s sponsors – viewers over the four games averaged just 8.9m – lower than any of the previous seven tournaments they played in (back to and including Euro 2000). Indeed, this is down over a third on the average 14.5m households tuning in to those tournaments.  These figures are taken from BARB’s weekly top 30 figures which measure the average audience over the course of a programme, and can be updated up to six weeks after their initial release (quoted ITV figures use ITVHDSD Total).

Major Tournament Viewers - Aug 16

Perhaps viewers were turned off by England’s opposition, awkward kick-off times or even by England’s poor result in their first game against Russia? We don’t think these are plausible explanations – first, our econometric model of all of England’s games since the year 2000 suggests the opposition’s ranking makes only a limited contribution to viewing figures. Second, all but the Wales game were in a prime viewing slot, and even the figure for Wales was below all of the awkwardly-timed 2002 South Korea and Japan games (see chart above). Third, the audience for the first game against Russia game was also poor – at just 9.74m. ITV chose this game precisely because weekend evening games attract the most viewers. Based on similar games, our modelling suggested 15m was a realistic expectation for the Russia game. Finally, all of the games mattered – in previous tournaments, audiences have dropped when England were effectively out (spot the 2014 Costa Rica game in the above chart), or safely through, but that wasn’t the case in any of the three group games.

There are some other potential explanations – perhaps the number of viewers in public places was higher than previously, or BARB’s methodology missed some internet viewing, or maybe the Brexit vote diverted viewers’ attention. It could be that fewer players in the starting eleven were from well-supported English clubs.  These might be part of the explanation, and further analysis might shed some light on these issues, but they are unlikely to explain all of the difference. It would be interesting to explore viewer age break-downs, and whether club now comes before country.

It’s also worth keeping all this in context – more UK households watch England games than any other football event – the Premier League draws up 2m on Sky and the FA Cup final attracted 6.7m (inc BBC 1 and BT Sport). We also think the UEFA Nations League will offer a big improvement on the pointless friendly games it replaces from September 2018 (although it looks like the will only be broadcast on Sky). Still, overall, it’s hard to escape the conclusion that watching England games is just not as appealing to viewers as it was.



Are you watching England? An economic analysis of TV audiences for England football matches

Our research report looking at the trends and drivers of the England football team’s TV audiences over the last 16 years:

England TV audience analysis

This report uses an econometric model built by Football Economics to quantify the drivers of audiences and estimate household viewing figures for future games. By understanding what drives TV audiences, we can develop a better understanding of football’s appeal and design competitions and schedules that work for fans.

Friday night Premier League football – how many tuned in?

Friday 14th August saw an unusual event – a televised Premier League football game.  We look at how many people tuned in, and whether Friday night could become the new Sunday afternoon.

The FA and Premier League have been at pains to point out this was due to “unique circumstances” but, from 16/17, Friday night games are part of the TV package so this serves as an interesting experiment.

The headline figure from the Broadcast Audience Research Service released figures this week suggested 861k viewers tuned in to Sky Sports 1 for the Friday evening game of Aston Villa vs Manchester United. This was higher than both the 612k who watched Manchester United defeat Tottenham in the season opener on BT Sport the previous Saturday, and the 663k who watched West Brom vs Man City on Sky Sports on the previous Monday. However, Sunday afternoon games tend to attract over a million viewers (see below) and that was the case here as 1.3m saw West Ham upset at Arsenal.  Of course, the opening weekend figures may have been buoyed by the two-month summer break and these figures don’t account for those watching in a pub. Still, the margin here is big enough to suggest Friday night is less attractive to home viewers than Sunday afternoon.

Arguably the most interesting comparison is with the Saturday night football the Friday game replaced.  Again using the opening weekend as a reference (Sat 8th August) 857k watched Chelsea vs Swansea on Saturday night – so, arguably a similar game and very similar to Friday’s 861k.  Looking back to spring, other similar Saturday night games also attracted under a million – for example 828k for United vs West Bromich Albion; 973k for Crystal Palace vs United.

So, the Friday night figure appears in line with similar Saturday games. If the ‘novelty factor’ of a Friday game boosted the figures this could point to Friday being less attractive than Saturday – though the opposite could be true if watching on Friday becomes a ‘habit’. A Friday game could also shift some demand away from the other weekend viewing slots.

The first weekend’s figures may disappoint BT Sport. For the big kick-off last year,  they achieved 818k viewers for Man United vs Swansea – so this year’s figure of 612k is down by more than 25%. Still, both figures were well-up on the 451k who watched the opening encounter in BT’s first season in 2013 (Liverpool vs Stoke). It’s probably too early to say how BT’s prices changes and the addition of Champions and Europa League will affect viewing overall.

Looking more broadly at viewing figures, last year Sky Sports achieved up to 2.1m for a key game in winter and several intra-top-four games attracted over 1.5m. Outside of the top 6, a good example might be 1.1m for Sunderland vs Newcastle in early April. England internationals attract much bigger audiences – it’s unusual they attract under 4m viewers, and in the knock-out stages of major tournaments, viewers can exceed 20m.

In terms of key points then, Friday night’s game appears similar to Saturday night in terms of attractiveness to viewers and BT Sport got 25% fewer viewers for the opening game than last year. This is a relatively simple analysis – we plan to undertake more detailed analysis and modelling of what drives Premier League and England TV audiences in the coming weeks. Get in touch if you’d like to sponsor or report on that work .





U21 success a key yardstick

With England under-21s in tournament action, we look at the statistical relationship between under-21 performance and the subsequent success of their senior National sides in major tournaments.

Germany won last year’s World Cup, having won the under-21 tournament in 2009, and with several players progressing to the senior side. But is this representative? We originally looked at this in a full research report a couple of years, ago, and have updated some of the analysis here.

First, we look at the European winners of the major tournaments (ie the World Cup or Euro Championships), and see whether they won any of the preceding under-21 tournaments – see the left-half of the below chart.

FE u21 Chart Final

Whilst only one in four (25%) of European major tournament winners had won the under-21 in the past 8 years, the striking element here is that almost two thirds of winners (ie 64%) won an under-21 tournament in the previous 12 years.   Put the other way around, only 1 in 3 European winners of major tournaments did so without winning an under-21 tournament in the preceding 12 years.

If we look at whether senior winners reached the semi-final of an under-21 tournament (right half of the chart), less than 1 in 5 European winners (18%) of a World Cup or Euro Championship did so without reaching the semis of preceding under-21 tournaments.  Similarly, 6 out of 10 (58%) reached the semis in the previous 8 years.

England’s tournament performance and prospects

The below chart records a 5 for winning a tournament, 4 for reaching the final, 3 for the semi, 2 for the second round or last sixteen, 1 for reaching the group stage or 0 if the team did not qualify.

England tournament performance

*Note that after 2006, the under-21 tournament was put back a year to odd years. In the chart 
 these appear against the previous year (ie the 2009 under-21 tournament is charted as 2008)

England’s 2009 under-21 final appearance stands out as the only recent foray into the last four of that tournament since the 1980s. The 2009 side contained a number of current England internationals including Joe Hart, James Milner and Theo Walcott. The under-21 wins in 82′ and 84′ were followed by two of England’s stronger performances in Italia 90′ and Euro 96′ – although as our full report shows, those senior performances only featured 3 players from the winning under-21 teams.

So, in summary then, it’s rare for a national team to win a major tournament without having reached at least the semi-finals in one of the previous six under-21 tournaments.  History suggests England’s 2009 under-21 final appearance may still has some relevance to the performance of the senior team – and the progression of players in the 2009 group to the senior team appears higher than in the 1980s.

Premier League – how many points to stay up?

Having emerged the other side of the hectic winter fixtures list, the premier league table now gives a good indication of who’s likely to finish where. Whilst much of the discussion in the press focuses on races for the title and European spots, the battle at the bottom of the league rumbles on.

In terms of avoiding relegation, it’s often said that 40 is the ‘magic’ number of points. Since the Premier League reduced in size to 20 teams in 1995/6, 40 points has only been insufficient to keep a team in the league in 3 out of 19 seasons: Sunderland and Bolton were relegated with 40 points in 1996/97 and 1997/98 respectively whilst West Ham famously succumbed to the drop with 42 points to their name in 2002/3.

Given changes in the number of teams in the Premier League, simply looking at points attained is inappropriate, if you want a picture of how ‘survival’ has evolved since the Premier League began back in 1992. So, for the clubs in the bottom places, we calculate the % of the total points available they won. This is shown in the chart below, for every Premier League season.


A few things jump out:

(1) over time, we find a small downward trend in the points won by the club which finished above the bottom 3. In other words, in general, clubs need marginally less points to stay up than before. Based on the trend line, a club now needs around 2 points less to stay up compared to 10 years ago.

(2) but there is considerable variation in points won by relegated clubs from season to season; particularly for the bottom club.

(3) apart from a period in the late 1990s, it has usually been the case that the 3rd bottom club and the first safe club are very close in terms of points won i.e. at least one relegation place is decided on the last day of the season.

So what might explain the downward trend in the points required to stay in the league? Perhaps, there’s has been a decline in the performance of teams in the bottom half of the premier league, relative to those in the rest of the league. Such an explanation fits with the ‘stylised fact’ that the premier league has become less competitive over time and/or the premier league has split into performance ‘tiers’. See Football Economics’ previous work on league competitiveness here

But if you want  bite-size pieces of information to go away with, then since the Premier League adjusted to 20 clubs:

  • the 40 point mark isn’t a bad guide – achieve that and based on history you have an 85% chance of staying in the league.
  • drawing every game and reaching 38 points gives you a 68% chance of survival.

We’ve a number of ideas on how to extend the analysis, perhaps by looking over a longer-time horizon, comparing with Europe’s top leagues or making a comparison between the old Division 1 and the Premier League – but that’s for another post. If you have any queries or questions, or would like to sponsor some of these further analyses, please do get in contact with us.

Grant Davies – Football Economics

Themes from the League Managers Association Conference

An excellent day today at the annual LMA conference which included a range of inspirational talks on leadership with fascinating insights from a range of speakers including Sir Alex Ferguson. Here we draw out our themes and conclusions from the day:

Manager development and support – Richard Beven (Chief Exec of the LMA) highlighted the ongoing challenge of massive turnover in club managers. He suggested that 50% of dismissals are of first-time managers, and 50% of those fail to ever get another job in management – a world away from the Wengers and Mourinhos. In the quarter to end October, LMA’s new stats show 17 managers had been dismissed – a 10-year high.  Manager development remains an issue for clubs, as our research on managers’ club experience also shows.

Diversity – Lord Herman Ouseley (chairman of Kick it Out) set out the diversity management challenge, looking at experience and skill, rather than expectations. Steve Bradbury’s recent research on BME representation has brought the issue into sharp relief – just 3.4% of senior coaching positions in the 92 football league in 2012 were filled by BME coaches, compared to 6.4% of those with a UEFA Pro Coaching License, 14% of the wider population, and 25% of professional players.

Evaluating impact – the Premier League has spent £350m from 2012-2015 on the Elite Player Performance Plan (EPPP) – apparently similar to expenditure by team GB – to improve the 38% of the Premier League’s talent which is home-grown, versus 50% in the Bundesliga and 72% in La Liga.   Credit to the Premier League for establishing measurable targets, which should enable this programme to be independently evaluated, assessed and focussed.

From data science to insight – we heard from Rasmus Ankerson on interpreting big data in football with his company 21st Club and from Steve McClaren and Alec Stewart on how player data has transformed football and cricket. Still, it’s clear that there’s further to go in actually making sense of data for decision-makers, and for new feedback to be fully accepted alongside ‘gut instinct’.

To conclude then, an inspirational day with leadership insights for all, and challenges ahead for managers and football in general. Of course, at Football Economics, we see that research and analysis have a big role to play in each of these areas – turning data into information, and then practical insight to help football develop its approach to meet the challenges it faces.