Share – It’s something I’m always telling my toddler, but are there times when sharing is a bad thing? This article concerns the merits of groundsharing beginning with a look at AFC Wimbledon and their ground-sharing tenants Kingstonian. This is a slightly unusual arrangement in that Kingstonian were the former owners of Kingsmeadow however financial problems saw the ground put up for sale. It’s purchase by AFC Wimbledon, who happened at the time to be homeless and in search of a ground, ensured that football not food shopping continued to be the main weekly activity at Kingsmeadow….
AFC Wimbledon: Cuckoos in the Kingstonian nest?
..and they lived happily ever after. All is not however, not rosy according to Kingstonian Fan Jamie Cutteridge. Writing on the blog The Real F.A. Cup he argues that though the ground-sharing deal with AFC Wimbledon has in large part ensured Kingstonian’s survival (as rent-free tenants) it is nevertheless a cause of concern. Issues he points to include continued uncertainty over the long-term future of the ground, Kingstonian’s inability to raise revenue from the ground , ‘removal’ of kingstonians identity from the ground and issues over arranging fixtures. The main gripe however, caused me to sit up:
As long as AFC exist in Kingston, the Ks crowds will suffer. Necessarily high prices amongst all the teams in the Ryman means the disparity between Ks and AFC ticket costs are not large enough to ensure new fans come to Ks. Your new or neutral fan in Kingston or the surrounding area will be drawn to AFC through a combination of a higher standard of football, and the chance to see the media darlings in the flesh.
This arguement shares some similarities in logic with the suggestion that non-league sides attendances may be affected by gates at nearby league clubs (benefiting when they fall, but declining when they rise), but when I recently put this assumption under the microscope it seemed that the intuition was wrong and there is at best a weak relationship between attendances at league clubs and their non-league neighbours. Also when looking at Notts County and Notts Forest attendances it seemed that the relationship between the two clubs became stronger the closer together the clubs were in terms of the league they played in. As Wimbledon are playing at a higher level than Kingstonian, and are now a league club I suspected that the concerns expressed in the article may well be unfounded. In any event repeating the same correlation test (a test of how much one set of numbers is related, or ‘correlates’ with another set of numbers) using data for Kingstonian and AFC Wimbledon should, in theory, be relatively simple.
One issue, as always with non-league clubs, is the availability of data. Whilst some sources exist for league clubs non-league history is often quickly shrouded in the mists of time (which is strangely what I like so much about non-league – all the hidden histories which exist). I managed to find this data for Wimbledon, for Kingstonian however, I was forced to rely on Wikipedia which is less reliable, and only goes back to 2005-2006. The figure for 2011-2012 was obtained from the Kingstonian FC website. This has the effect of unfortunately eliminating the seasons 02/03 03/04 and 04/05 from the analysis leaving us with seven years worth of data in which to compare. Looking at the raw data it seems that whilst AFC Wimbledon’s average has grown considerably over Kingstonian’s has remained relatively stable over the time-period in question ranging from 307 to a high-point of 412.
If we turn the raw-data into a line graph then we can see more clearly the overall trends in the respective clubs attendances:
Incomplete though this may be it does seem to suggest that Kingstonian’s gates haven’t necessarily been affected by the growing stature of AFC Wimbledon. Had this been the case we may have expected to see Kingstonian’s average decline as Wimbledon’s average rose. In fact it seems that until last season they had followed a similar trajectory. In 2011-2012 Kingstonian’s average gate did decline whilst Wimbledon’s increased though this in itself does not reveal anything much – it may simply be the result of the K’s finishing in the mid-table obscurity of 11th place compared to a more respectable 7th the season before rather than any kind of trend.
the statistical test I am using will produce a correlation coefficient; this is a number between 0 and on one side -1 and on the other +1. 0 means the two variables are unrelated whilst moving towards either extreme +1 or -1 suggest a strong correlation between the two variables. A positive (+) correlation means that as one figure increases so does the other whilst a negative (-) means that as one figure rises the other falls. If it were the case that a number of people were making a choice between watching either AFC Wimbledon or Kingstonian then we could reasonably expect to witness a negative relationship – that is high average gate at one club means a lower average at the other – as people switch between the two.
The graph however shows the opposite – a positive relationship with higher gates at one club being associated with higher gates at the other. The correlation coefficient itself comes in at 0.42 suggesting that the relationship between attendances at the two clubs is towards the weak end of the scale. Whilst there may be many possible explanations for the positive correlation (including coincidence) what we can say from this analysis is that it does not appear that increased average gates at AFC Wimbledon necessarily result in reduced gates at Kingstonian; that is to say it does not seem as if spectators are in fact switching from Kingstonian to AFC Wimbledon.
The Selhurst Park Story:
Getting AFC Wimbledon off the hook I was left wondering more about the impact of ground sharing and looking for an example of a ground-sharing arrangement which had lasted a number of years. Strangely Wimbledon themselves provide another case having played at Selhurst Park, home to Crystal Palace, between 1991 and 2003. Prior to this Palace shared Selhurst with Charlton Athletic from 1985. So in total Palace shared their ground almost continuously for 19 years.I hopped over to what is fast becoming my favourite website www.european-football-statistics.co.uk to see what data I could find.
Palace and Charlton:
This graph shows that during the period of their ground share (1985-86 to 1989-1990) both clubs experienced rising attendances. The correlation coefficient is in fact 0.81 for this period. There may be many possible reasons for a positive correlation like this, even coincidence though, one likely explanation is that having reached a low point in 1984/85 football attendances were beginning to recover. As two clubs playing in a similar area at around the same level both clubs would have benefitted similarly from this increase hence the strong correlation which can be observed in their attendance figures. Certainly it seems as if neither club appears to have been adversely affected by the ground-share
The Dons move in:
With Charlton off to ground-share with West Ham for a season before finally returning home to The Valley Selhurst park became home to Wimbledon who vacated their their home at Plough Lane in May 1991 due to issues around bringing the ground up to the standards required by the Taylor report. This graph shows the respective average attendances of Palace and Wimbledon during this period – the period 91/92 to 01/02 is used as though the ground-share arrangement included both the 1990/91 and 2002/03 seasons neither was for a complete season. The graph shows the dramatic rise, and equally dramatic fall, in Wimbledon’s average attendances. Two key events lie behind the collapse; 1999/00 saw the club relegated from the Premiership and 2002 saw the confirmation of the move to Milton Keynes. Unsurprisingly in this turbulent period their attendances evaporated – It’s interesting to note that between 99/00 and 00/01 whilst Wimbledon’s average attendance dropped by a staggering 9260, but Palace’s increased by a more modest 1399, and this in the season before AFC Wimbledon’s inaugural season in 2002/3 which drew an average crowd of 3003. If those 9260 had been fair-weather fans who floated between Palace and Wimbledon then surely we would expect to see far more switching back to Palace.
Overall the graph seems to show that despite peaks in 1997/98 when Palace returned to the premiership, and one just off the graph in 1990/91 where they had finished 3rd in the top flight having been FA cup runners up the previous season, Palace’s attendances have over the long-term been on a gradual upward slope which seems unaffected by Wimbledon’s rising attendances and increasing prominence throughout the 1990s.
If we plot the attendances on a scatter graph we can see that there appears to be very little correlation at all between the attendance figures at the two clubs for the period. In fact the correlation coefficient is an extremely weak 0.15. This suggests that changes in the average crowds at one club had very little impact on attendance figures at the other club.
There may be many valid concerns about the politics behind ground-sharing and the affects of ground-sharing, but taken together the evidence from the cases of AFC Wimbledon at Kingsmeadow and Charlton and Wimbledon FC at Selhurst Park suggest that losing crowds to the other team who are sharing the ground shouldn’t keep fans awake at night – at least not in the short and medium term.
This would perhaps seem counter-intuitive, but there are a number of possible reasons why ground-sharing may have a much more limited impact on attendances than may have been expected:
1.) A rising tide lifts all boats
If we combine all the data for Charlton and Wimbledon ground-sharing at Selhurst from 1985/86 to 2001/02, also including the partial season of 1990/91, we get this graph: What seems interesting is that despite Palace’s short term peaks, the final collapse of Wimbledon’s attendances, and the slight dis-juncture between Charlton’s and Wimbledon’s attendances the attendance figures seem to be following the same long-term upwards trajectory. If we add the figure for the average attendance for the football league we get this rather interesting chart: It seems that for both Palace and Charlton/Wimbledon the biggest impact on attendances has been the steady rise in football attendances since their 1985 low-point.
Looking at the scatter graph using data from 85/86 to 03/04 there appears to be a reasonably strong correlation between Crystal Palace’s average and the combined average for all the professional leagues. The coefficient is a moderate-to-strong 0.64 which suggests that the variations in overall league attendances explain 41% of the variation in Palace’s attendances over the period shown.
In economic terms the period in question is a growth market for football. With more spectators coming to football matches year-on-year then clubs could reasonably expect attendances to rise; even taking into account short-term fluctuations in form then the trend would still be for increases over the course of several seasons.This overall rise in new spectators may well mitigate any loss of spectators arising from a ground-sharing arrangement allowing both clubs to flourish – though one thing we cannot be sure of is whether Palace would have attracted more new spectators had Wimbledon not been at Selhurst Park.
2.) Local geography
The argument that ground-sharing results in a potential impact on crowds for an incumbent rests on the assumption that fair-weather fans will switch to spectating at the club which seems to be more attractive in terms of the standard of football on offer, as well as other considerations like price, seat comfort, and quality of pies. Whilst we can debate whether this theoretical model of behaviour is true of football supporters what we can say is that even if it does reflect reality then in a metropolitain area such as London ground-sharing is less likely to have an impact. Good transport links mean there is a wealth of choice available to the floating-fan within just a short travelling distance anyway, and even more if they wish to go a little further afield. If anything a ground-share means that there is less competition as when a club is at home, the other side must necessarily be away. As most supporters only attend home games this means that a floating fan deciding where to go on any particular Saturday has their choice reduced – as in the case of Wimbledon and Palace their respective grounds before the ground-share were only 6 miles away so had someone been inclined to switch to the team which was performing best at the time they would have been able to do so easily.
The rise of AFC Wimbledon and the continued controversy over MK ‘Dons’ – in particular their badge design shows how important identity is in football. Between 99/00 and 01/02 Wimbledon’s average attendance dropped by 10176 whilst over the same period Palace’s increased by only 2439. If we were expecting free floating-fans who readily switch between either club then surely this figure would be higher? Where did the other 7737 go? Well the answer for 3003 of them in 2002/03 was watching AFC Wimbledon in the Combined Counties League Premier division taking on the likes of Farnham Town, North Greenford United and Bedfont in a division where attendances rarely get out of double figures. Fair-weather fans these were most definitely not- the Wimbledon name, and identity meant something to them – and this is true for other clubs named after, or with an otherwise relationship with a locale stretching back often for more than a century. Quite simply supporters don’t select a football club in a straightforwardly clinical rational way; it is a decision involving feeling, collective memory and community ties.