Serie A

Competitive inequality in Europe’s top four football leagues

This post offers an overall analysis of the findings of ITMH’s competitive inequality project, tracking the impact of financial inequality upon football. The introductory post explains the measures that are used in the analysis. Individual commentary posts follow, discussing top flight football in Germany, Italy, Spain and England.

Each of the individual league commentaries has noted trends across the four measures of competition I am tracking. But what can be said about the overall picture? Are there any general messages within the data? For each measure, I’ll include a chart showing the rolling averages for each of the leagues plotted alongside each other (they always look a bit messy, but can give some indications of cross league performance) then an overall average of the measure across all four leagues.

Fig 1: correlation comparison

(Lines here display five-year rolling averages for the correlation value between each season’s finishing positions and those in the preceding season for each of the four leagues discussed. See the introductory post to this series for an explanation of the choice of measurements.)

As noted in the individual commentaries, the Italian league has typically displayed slightly higher levels of year-to-year consistency, while the German and English leagues have been more prone to periods of real flux. All four leagues currently display correlation levels well above historical averages.

Fig 2: correlation overall average

(The line here displays a five-year rolling average of an average correlation value between each season’s finishing positions and those in the preceding season for each of the four leagues discussed. The dotted line indicates the overall trend in this measure. See the introductory post to this series for an explanation of the choice of measurements.)

An average across the four countries smooths out some of the variation to show a general rising trend across all leagues, indicating less fluidity of finishing positions. Furthermore, at present, the five year rolling average sits well above the trendline, which indicates that this is a trend that may currently be intensifying.

Fig 3: top four turnover comparison

(Lines here represent five-year rolling averages for each of the four leagues of the top four turnover measure, assessing levels of variation among clubs finishing in the top four of each league. See the introductory post to this series for an explanation of the choice of measurements.)

The most notable collective trend across the individual results is that each league has seen periods of sharp decline since the mid-1990s. In England, this is followed by something of a rebound (marking the emergence of a financial Big Six, which created greater competition for top four places), in Spain the decline is steady and continuous, in Italy there remains greater scope for variation.

Fig 4: top four turnover overall average

(The line here displays a five-year rolling average of an average of the top four turnover measure, assessing levels of variation among clubs finishing in the top four of each league. The dotted line indicates the overall trend in this measure. See the introductory post to this series for an explanation of the choice of measurements.)

The overall average confirms the above trend: the overall direction of travel is downwards, but this is more marked from around 2000 onwards. Given that this was the point at which the Champions League was opened up to more teams from the biggest leagues, we can point to a clear effect of format changes at the continental level: the wealth attained through big club participation in Europe’s most prestigious competition allows these clubs to maintain a more stable grasp upon the top positions in their domestic leagues. As with correlation measure, the rolling average here sits below the overall trend, indicating this impact is intensifying. With further reform to the Champions League format, potentially enabling greater rewards for participants, even further decline in this measure is possible.

Fig 5: top six turnover comparison

(Lines here represent five-year rolling averages for each of the four leagues of the top six turnover measure, assessing levels of variation among clubs finishing in the top six of each league. See the introductory post to this series for an explanation of the choice of measurements.)

Fig 6: top six turnover overall average

(The line here displays a five-year rolling average of an average of the top four turnover measure, assessing levels of variation among clubs finishing in the top six of each league. The dotted line indicates the overall trend in this measure. See the introductory post to this series for an explanation of the choice of measurements.)

A similar effect is noticeable in the top six turnover measure. Prior to around 1995, there is no real collective trend in the results. Since then, to different degrees, all four leagues have seen decline in this measure. As a result, the decline in the overall average is notable and appears to currently be intensifying.

Fig 7: champions’ points per game comparison

(Lines here represent five-year rolling averages for each league of the points won per game by the team that finished champions. See the introductory post to this series for an explanation of the choice of measurements.)

In the individual commentaries, I highlighted the way that measures of dominance for championship winners enable some individual teams to be picked out. We can again, for instance, pick out the pinnacles of Real Madrid in the late 1950s/early 1960s and again in the late 1980s. But, beyond this, what is more apparent is that the recent steep rises in performance are consistent across the leagues, rather than being obviously attributable to individual teams. All four leagues have seen peaks of performance in the past fifteen years that surpass anything seen before.

Fig 8: champions’ points per game overall average

(The line here displays a five-year rolling average of an average across the four leagues of points won per game by the team that finished champions. The dotted line indicates the overall trend in this measure. See the introductory post to this series for an explanation of the choice of measurements.)

When I was young, I remember my Dad trying to instruct me of the value of a hard-earned away draw by saying that a team who won all of their home games and drew all of their away games would win the league. Now, the averages here show that, while a title victory would not be entirely certain across these leagues, a two points per game average would certainly have made a team a contender up until about twenty years ago. Since then, it would leave you trailing some way back. So, alongside winning all home games, a team today would probably also need to win about one-third of those away and draw the rest in order to accumulate enough points to pick up a title.

Now, it is worth noting that the past few seasons display some regression in this measure. Is this a sign that we had reached unsustainable levels of dominance, or is it perhaps due to the financial hit taken by many of the big clubs through the pandemic? This is something to monitor in future years.

Fig 9: champions’ goal difference per game comparison

(Lines here represent five-year rolling averages for each league of the average goal difference per game of the team that finished champions. See the introductory post to this series for an explanation of the choice of measurements.)

While the overall shape of this chart is similar to that of points per game, with the performance of great teams visible a various different points, followed by a general recent rise, it is notable that the recent performance of Spanish and German champions jumped well clear of the other two leagues before both regressing. It can also be seen that, despite this season being poor by their recent standards, results here would point to Bayern still being the most dominant of the champions of the respective leagues.

Fig 10: champions’ goal difference per game overall average

(The line here displays a five-year rolling average of an average across the four leagues of the average goal difference per game of the team that finished champions. The dotted line indicates the overall trend in this measure. See the introductory post to this series for an explanation of the choice of measurements.)

As I discussed in the Spanish and Italian commentaries, the period prior to the mid-1960s was one marked by some flamboyant champions. This is something that registers in the overall average here in values that would not be surpassed for another half a century. Recent years have seen champions of the four leagues consistently recording average goal difference of over 1.5 goals per game. The same kind of recent regression seen in respect to average points won is also visible here, so champions are not exerting quite the same level of dominance they managed across the 2010s.

Summary

Across all measures here, support can be found for the conclusion that competition within domestic leagues is profoundly influenced by the format of European competitions. The opening up of the Champions League to non-champions in 1997, then the shift to up to four teams qualifying from top leagues in 1999 corresponds to noticeable upswings in the average correlation of finishing positions, lower turnover in top four and top six finishers, alongside more dominant performances by champions.

In other words, dominance over the past 20 years is different to the dominance of Real Madrid in the late 1950s or Bayern Munich in the 1970s: there is a deeper structural foundation to it in terms of the way the Champions League allows for a massive concentration of financial power in the hands of the continent’s elite clubs, which makes it easier for them to defend their status from domestic challengers. And, with the coming expansion of the Champions League format, the structural unfairness visible within these figures has potential to deepen even further.

Competitive inequality in Italian football

This post analyses Serie A, as part of ITMH’s competitive inequality project, tracking the impact of financial inequality upon football. The introductory post explains the measures that are used in the analysis. See the other posts in this series for commentary on top flight football in Germany, Spain and England, or the overall discussion of trends within Europe’s top leagues.

Serie A has formed the single top national league in Italy since the 1929-30 season. The analysis below tracks performance across the period since its resumption after WWII. Over this period, three teams have consistently claimed most success: Juventus (36 championships in total, 29 of which are post-war); Internazionale (19 championships – 14 post-war); Milan (19 championships – 16 post-war). Other major contending teams include the two Rome clubs (Roma and Lazio), Napoli and Fiorentina.

The three dominant Italian clubs have also won 12 European Cup/Champions League trophies between them (Milan 7, Inter 3, Juventus 2). All three participated in the European Super League proposal, with the Juventus chairman, Andrea Agnelli, being thought to be one of the prime protagonists behind the plans. Indeed, Juventus only signalled their intention to pull out of the enterprise in recent days (6th June 2023).

Methodological mithering: in keeping with the national talent for Machiavellian subterfuge, Italian football has frequently descended into scandal, resulting in much greater use of points deductions than in other major leagues. This leaves a choice to be made when assessing league finishing positions between what has been determined by performance on the field and that subsequent to the application of penalties. E.g. in 2005-6, Juventus finished at the top of the Serie A table, but were subsequently penalised for their involvement in the Calciopoli scandal by being officially placed bottom of the table and relegated to Serie B. For the sake of this exercise, should their finishing position be recorded as 1 or 20; should it be their record, or that of the team subsequently awarded the title (Inter) that counts to points per game and GD per game calculations? My choice has been to, throughout the Italian results, use a raw league table based upon finishing positions as if there were no deductions, as this feels truer to the purpose of the assessment by reducing any artificial volatility.

Fig 1: correlation

(Bars here represent the individual correlation value between each season’s finishing positions and those in the preceding season. The red line is a five-year rolling average to pick up on trends in the medium term. The dotted line is an overall trendline representing the overall direction of travel for this measure across the results. See the introductory post to this series for an explanation of the choice of measurements.)

Of the four leagues I’m looking at as part of this exercise, Serie A generally exhibits the highest average correlation between finishing positions from year to year. While the individual bars show great variation, the rolling average is typically operating at a slightly higher level than other leagues. This suggests that Italian football may possibly have developed an established order of clubs at an earlier stage, and have maintained this more thoroughly than other countries. It certainly is pronounced that Italian football is dominated by teams from its largest (and in the case of its three dominating clubs, its most affluent) cities, with fewer provincial upstarts than elsewhere. This lends credence to the idea that finance has mattered more, for longer, in the Italian league than with other major European leagues. Further support for this idea would come from the fact that, while the overall trend for this measure is upwards, the increase is less pronounced than that seen in some other leagues, suggesting that the effects of money are already baked into Italian football to a greater degree.

Also of note is the fact that recent years (since 2011) have been characterised by sustained higher levels of correlation, which have driven the averages up steadily.

Fig 2: turnover

(Lines here represent five-year rolling averages for the turnover measure, assessing levels of variation among clubs finishing in the top four, top six and top ten (in seasons with a 20+ team league) of the league. NB – since the calculation of this measurement relies upon future finishing positions, the current season’s figure can only be known in three year’s time. The latest figure, therefore, is for the 2019-20 season. See the introductory post to this series for an explanation of the choice of measurements.)

There is little in the way of overall trend for the turnover measurements for both top four and top six finishing positions in Serie A. With each, there is fluctuation between periods of greater or lesser turnover, yet it tends to consistently regress towards a mean of around 70% for top four turnover and 55% for top six turnover. The fact that top six turnover is in this range in a league with three dominant clubs suggests that, in most years, it is highly likely that these three finish in the top six, although they have not been able to colonise the top four, as evidenced by higher turnover at this level.

The measure for top ten turnover is only calculated following the expansion of the league to 20 teams in 2004-5 – it does not seem particularly worthwhile to calculate earlier than this for a group that would comprise more than half of the league. Yet, the short period in which this has been tracked sees a rapid decline in turnover levels, suggesting some degree of consolidation among more successful clubs that will be worth watching in future.

Fig 3: points per game

(Bars here represent the average points won per game by the team that finished champions that season. The red line is a five-year rolling average to pick up on trends in the medium term. The dotted line is an overall trendline representing the overall direction of travel for this measure across the results. See the introductory post to this series for an explanation of the choice of measurements.)

The rolling average here traces a loose upward sloping u-shape, being higher in the immediate post-war, lower through the 1970s, 80s and 90s (with odd exceptional higher performance such as from Juventus in 1976-77 and Inter in 1988-89), then generally higher levels of performance in the 21st century, capped by Juventus’s romp to 102 points in 2013-14.

Fig 4: goal difference per game

(Bars here represent the average goal difference per game recorded by each season’s title winners. The red line is a five-year rolling average to pick up on trends in the medium term. The dotted line is an overall trendline representing the overall direction of travel for this measure across the results. See the introductory post to this series for an explanation of the choice of measurements.)

Italian football is all about tight, stifling defensive tactics, right? Well, yes, but it wasn’t always like that. You can clearly see the period in which Catenaccio fully emerged and became the defining Italian football mindset from the chart’s results – Helenio Herrara’s Inter team, who deployed it effectively the dominant force in both Italian and European football in the mid-1960s, became a model for much of what was to follow.

Prior to that, however, the immediate post-war period featured some incredible free-scoring performances from Italian champions. In 1950-51, Juventus, Inter and the eventual champions – a Milan side led by their Gre-No-Li Swedish attacking trio – each scored more than 100 goals in a 38-game season. But, the absolute pinnacle of post-war Serie A were the exploits of the dominant but ill-fated Grande Torino team who won five titles over the period of the war and its immediate aftermath. Their finest performance came in 1947-48, a season in which they scored 125 goals in 40 games, recording five or more strikes in half of their home matches, which included a 10-0 victory over Alessandria.

Unfortunately, tragedy would strike that team. In May 1949, in poor visibility from adverse weather conditions, the plane carrying them home from a friendly with Benfica in Lisbon crashed into one of the walls of the Basilica of Superga, which stands on a hill over Turin. None of the 31 on board, including team players and coaches survived the crash. Unlike Manchester United, who were able to return to prominence after a similarly tragic crash in the following decade, Torino have never again been a force in Italian football – they managed a solitary title in 1975-76, but have also had spells in Serie B and currently are a resolutely mid-table Serie A club.

In recent years, Italian title winners have been recording higher goal difference levels than the 1960s-1990s period. They have not, however, been able to get close to the incredible performance of that legendary Torino side.

Summary

This project is primarily interested in the impact of recent flows of money into football, and how this might be making the game more unequal and less competitive. The Serie A results show some evidence for this, with slight upward trends in correlation, reduced turnover in higher league positions and stronger performance from champions. However, it is also clear that Italian football has always displayed signs of stratification of clubs, making these trends less pronounced than they are elsewhere.