This post follows on from the last post about the relationship between attendance figures at league and non-league clubs. In that post I attempted to answer the question; would falling attendances at a league club translate into more spectators at a nearby non-league side and would the opposite be true – would a league club attracting greater numbers of spectators result in crowds then falling. The short answer being that there was only a fairly weak relationship between the two.
Behind the question was a basic assumption. This is that football crowds are a zero-sum game. In other words there are only so many people who want to go to football matches so more spectators at one ground mean less to go round at others – resulting in lower average attendances. From the previous analysis it seems that for league and non-league clubs this is simply not the case – in fact it seems they are two separate worlds and that stay-away spectators of league clubs do not substitute their spectatorship at non-league clubs, just as non-league spectators do not desert their favoured sides to join the crowds attracted by success at league sides.
I thought it would be interesting to see whether or not this would be the case for league clubs. Could attendances at one league club be affected by attendances at another nearby league club?
Conventional wisdom would say that the results would be similar to the non-league analysis – after all supporting a football club ,unlike any other choice, is regarded as a choice for life unlike any other however, Simon Kuper and Stefan Szymanski in their book ‘Why England lose & other curious football phenomena explained’ contrast this with an alternative view which casts the football fan as something of a floating-voter. Carrying out their own statistical analysis of 61 years worth of English league attendances they conclude that
it seems it has always been the case that the majority of people who go to English football matches go only once in a while and are often quite fluid about who they choose to watch
First I need two clubs, preferably in the same city. There are a number of cities with more than one league club; London has many Arsenal, Spurs, West Ham, Millwall, QPR, Fulham, Brentford and Leyton Orient to name just some. Sheffield have United and Wednesday. Liverpool, Manchester, Birmingham, all have at least two league clubs.
In the event I have chosen Nottingham. Home to both Notts Forest and Notts County. The main reason for this choice is that the grounds are I was once told are the closest together of any league club in England. This is also confirmed by Wikipedia, and the evidence of my own eyes as I once sat watching Mickey Evans grab a brace for Southampton in a 3-1 win over Forest. From the stand I was in the corner of Countys ground, Meadow Lane, was just visible.
I obtained the data for average league attendances from European football statistics.co.uk and decided on 1970 as the year to begin the analysis.
The result is much the same as with the league and non-league comparisons. A fairly weak relationship between the two sets of figures with our correlation coefficient R = -0.29 (with 0 being the weakest possible and -1 or +1 being the strongest) leading to the conclusion that there is not much swapping of grounds going on among the football fans of Nottingham.
This is perhaps unsurprising. Like siblings Forest and County may come from the same place, but they possess unique personalities. County on one hand are credited with being the oldest football league club in England, but Forest, the younger upstart have been the more successful of the two in terms of league position and silverware. Forests dominant position relative to County can be seen in the following chart which compares the level played at by each club on a season-by-season basis. In the chart level 1 represents the highest level – Currently the Premier league.
What I then decided to do was use the chart to compare the correlation coefficients for average attendances at two points 1.) 1974-1984 – The period where the two clubs were closest together in terms of the level at which they competed and 2.) 1985-2005 – The period where the clubs were furthest apart.
Closest together 1974-1984
This period saw the clubs competing at the same level in 7 years out of the 10, 3 of these being at the top level. For the remaining 4 County played only 1 level below Forest. According to the Notts County website this period was something of a golden age:
Arguably the most successful period in the Club’s history, under arguably the most successful manager, came in the 1970’s when Jimmy Sirrel put so much pride back into the Club including promotion to Division One in 1981 after a historic win at Chelsea.
Meanwhile Forest were similarly enduring their own golden age under Brian Clough winning the league championship in 1978 followed by European Cup wins in 1979 and 1980. But what did attendances look like at the two clubs during this time?
Just looking at the graph we can see an association between the two. This period in fact gives us our strongest correlation yet R = -0.76 firmly in the strong end of the range. This also means R Square is 0.58, so 58% of the difference in one attendance figure at one club is explained by the difference in the attendance figure at the other club. The relationship is also negative so as attendance rises at one club it falls at the other. This suggests the possibility that during this period of time, one where both clubs were experiencing success, a significant number of spectators were indeed floating between the two clubs.
Furthest apart 1985-2005
During this next period Forest were overwhelmingly the more successful. For only 2 of these years did the clubs compete at the same level and for 9 of the years County were 2 levels below Forest
During this period the correlation coefficient is much lower, in the weak end of the range R = 0.22. The relationship also becomes positive. This seems to suggest that as the clubs are playing at different levels the affect seen during the period 1974-1984 has completely evaporated.
First we need to be wary of over-stating the findings. This analysis is based on just one case study of two clubs over a somewhat limited period of time. There is therefore no way we can say this applies to other league clubs without first analysing their data. What the analysis does show is that whilst two nearby league clubs, County and Forest, are playing at the same level, or are both enjoying periods of success the relationship between average attendances at either club is more pronounced than when the clubs are competing at different levels. Therefore it seems that one explanation is that for the ‘floating spectator’ affect to work clubs need to be in close proximity to each other – not just physically, but in terms of league position. This is something which re-enforces the non-league analysis which found little evidence to suggest that attendances at league clubs have any real impact on non-league clubs several divisions below them.