Wednesday, 24 June 2015

Copa América: An Unfair Tournament by Design?

Much of Latin America is currently in the grips of the Copa América, a very exciting tournament indeed. Yet I take great issue with two aspects relating to its format. As I will explain, the tournament’s particular format makes that some countries are awarded an unfair advantage relative to others, just because of the group they were drawn into and/or the timing of their games.

Source: quepasa.cl

The first issue is a straightforward one. In all major tournaments (e.g. FIFA World Cup) as well as national leagues, the final games in the group stage of any particular group or league are scheduled at the same time. The purpose is that no team has an undeserved advantage of knowing what its competitors did and thus of being able to know what result it would need – whereas the competitors who had to play before didn’t. 

A real-life instance that made the involved unfairness all too obvious in practice and which is part of our collective football memory occurred at the 1982 FIFA World Cup. Both West Germany and Austria were to proceed to the next round if, in their final group game, West Germany were to beat Austria 1-0. After ten minutes, the West Germans scored and both teams ran down the clock for the remaining eighty. The Disgrace of Gijón incited FIFA, in future world cups, to have the final games of each group scheduled at the same time. At the Copa América, anno 2015, CONMEBOL still seems content awarding half of the teams the potential advantage of knowing what result they will need from their final group game.

The first issue can very easily be fixed. This is quite different for the second, which is more structural and relates to the two (out of three) “best thirds” qualifying. Naturally, this implies that what happens in other groups also affects what happens in any given group. Again, the teams that will play later will have the advantage of superior information about the result they need: at the time of the final games of group A, the final games in groups B and C are still left to be played, implying uncertainty of outcome, whereas at the time of the final games in group C, the final ranking of groups B and C is already known.

So the teams in later groups know better what result they need to progress. But does this actually affect how likely they are to get it? For those who prefer numbers over reason, let’s have a look at the nine editions of the Copa América where the current “best-thirds” rule was used (i.e. since 1993) and see if there’s any differences. In case better information does not affect who progresses, we would expect, on average, to see three times the best third of each group not to have made it through. The actual results, however, show the following:

Each bar indicates for the respective groups how many (on Y-axis) out of nine times the
third of that group did not progress to the knock-out stage, due to being the "worst third"

In six out of nine editions, it was the third of group A which happened to be the worst third and thus the one not to progress. Interestingly, also, the one time the third of group C did not manage to qualify, in 2007, this was a mathematical certainty already ahead of the last matchday of group C.

So now that we are clear on what the problem is, how do we solve it? By making sure that the same number of teams from each group is to progress, so that what happens in one group does not affect how many are to progress in another (alternatively, the final games of all the groups could be scheduled at the same time, but this can be considered too much of a drawback for TV audiences). Let’s assume that Copa América wants to hold on to twelve countries playing the group stage and eight progressing, to the quarters. We could have four groups of three, with the two best of each group making it through. Downside here would be that the number of teams per group is uneven, so that in each group one team would be free on the last match day and thus continues to have more complete information. The solution would be to make two groups of six, with the best four of each group qualifying for the next stage. The first of group A would then play the fourth of group B and vice versa and the second of group A against the third of group B and vice versa. The teams that would make it to the final will have competed in eight games, one more than in the case of the World Cup. If this is considered too many, one could immediately move from the group stage to the semis instead, with only the top two qualifying, which would make even more sense. 

Is the problem of an undesirable format confined to the Copa América and CONMEBOL? Interestingly, UEFA has decided to import a similar system for the Euro Cup, starting from next year’s edition, when the number of teams participating will increase with a staggering 50%. In light of the above, it is curious that a federation that is becoming adamant about creating a level playing field through Financial Fair Play, more or less at the same time decides to start assigning arbitrary (dis)advantages to some of the teams competing in its main tournament, i.e. with now “best thirds” qualifying; with some group winners meeting teams that became second, others that ended up third; some teams that were second meeting group winners and others other seconds. Again, the solution may be simple: four groups of six teams, with either only the winner or the best two of each group progressing to the knock-out stage. That’s what I would call fair play.