Would it surprise anyone to know that if size and money were removed as factors driving success that the world’s top-ranked soccer power would be Cameroon? They didn’t even qualify for the World Cup. How can any ranking system declare them to be number one? Read on for the short answer. (A longer, more academic discussion has been demoted to the footnotes since eyes tend to glaze over with the prospect of statistical blather.)
This ranking is not saying that Cameroon is better than Brazil. What it says is that once certain factors influencing success (income and population) are taken into account, what’s left over is a different kind of measure of the country’s commitment to the sport.
Approach:
The way to achieve this modification is with a statistical technique that attempts to explain as much variation in FIFA points as possible using measures of other variables, namely, per capita Gross Domestic Product and population.* Note that this is not meant to be the most accurate model of FIFA point levels – it’s meant instead to be a way of explaining that part of the ranking that is due to good fortune. By looking at the rankings after the effects of these advantages are removed, we’ve essentially leveled the playing field. Any other variables that could help explain FIFA standings (e.g., how long ago soccer was introduced in that country, whether it has a well-supported professional league, percentages of kids who play in the youth system) are factors we would not want to exclude since they’re part of what it is we’re trying to measure.
Before listing the results of this new ranking format, I should mention that I’m no great fan of FIFA rankings even as just a starting point for modifications. Plenty of people have commented on the biases inherent in the system. Now if the Czech Republic reaches the finals and the US makes it to semis, I’ll happily eat my words, but most people agree there is too much emphasis on older results. That being said, I needed some ranking to start with and FIFA’s is the one cited most often.**
Results:
So, now that I’ve explained the intent as well as the caveats, let’s take a look at the results. Starting with the top 100 FIFA-ranked teams as of last month and applying our filter, we see the table below. Our new index, the Adjusted Ranks of Soccer Eminence (AROSE, for short***) shows a very different picture compared to the FIFA table.
AROSE Country (FIFA Rank)
1 Cameroon (15)
2 Czech Republic (2)
3 Senegal (29)
4 Portugal (8)
5 Denmark (11)
6 Uruguay (22)
7 Costa Rica (26)
8 Netherlands (3)
9 Brazil (1)
10 Croatia (24)
11 Nigeria (12)
12 Cote d’Ivoire (32)
13 Tunisia (21)
14 Paraguay (33)
15 Argentina (8)
16 Sweden (16)
17 Jamaica (44)
18 Trinidad & Tobago (47)
19 Spain (5)
20 Republic of Ireland (30)
21 Greece (19)
22 Mexico (6)
23 Turkey (13)
24 Romania (25)
25 Honduras (41)
26 Bahrain (54)
27 Egypt (18)
28 Bulgaria (38)
29 Ecuador (39)
30 France (7)
31 Serbia & Montenegro (46)
32 England (10)
33 Zambia (56)
34 Guinea (52)
35 Colombia (27)
36 Togo (59)
37 Morocco (36)
38 Italy (14)
39 Iran (22)
40 Poland (28)
41 Zimbabwe (55)
42 Switzerland (35)
43 Slovakia (43)
44 Norway (40)
45 Saudi Arabia (34)
46 Ghana (50)
47 Mali (65)
48 Iraq (52)
49 Angola (58)
50 Korea Republic (30)
51 Bosnia-Herzegovina (63)
52 Israel (47)
53 Finland (49)
54 Germany (19)
55 USA (4)
56 Qatar (76)
57 Uzbekistan (59)
58 Japan (17)
59 Latvia (69)
60 Guatemala (61)
61 Estonia (79)
62 Ukraine (41)
63 Slovenia (71)
64 Belarus (64)
65 Panama (81)
66 Albania (86)
67 Russia (37)
68 Australia (44)
69 Congo DR (70)
70 Kuwait (73)
71 Cuba (79)
72 Scotland (62)
73 Oman (82)
74 Jordan (83)
75 Wales (74)
It’s interesting to see which teams rose and fell the most compared to their FIFA positions. Cameroon is a fairly small country (population 16 million) and very poor (GDP per capita of $1,900) with a FIFA ranking of 15. When their disadvantages are statistically isolated and purged, though, they move to the top of the list. Maybe Roger Milla and the Indomitable Lions that made it to the quarterfinals of the 1990 World Cup in Italy created a legacy of overachievement. Several other African nations climb to higher ranks, too. Senegal looks impressive in the number 3 slot, up from a FIFA rank of 29. Cote d’Ivoire went from 32 to 12. Smaller nations like Uruguay and Costa Rica also look like bigger successes by this criterion, as do Jamaica, T&T, and Bahrain.
On the flip side, football superpower Brazil falls from #1 by FIFA’s accounting to #9 by the size and wealth-adjusted measure. Their fall-off is attributable to their big population (fifth largest in the world). Spain, Mexico, France, and Japan slide, too, due to their size and/or wealth. The most precipitous drop, though, goes to the US. We had a sneaking suspicion that it would. Despite growing strength and support, US soccer can’t really brag about its success relative to the nation’s advantages. (Bruce Arena et al. can and should feel good about their improvement, though.)
Before anyone starts assuming a political agenda here, like world-wide socialism to even things out, I should emphasize the goal of this re-ranking system. It’s not a statement of what should be. It’s more a statement of what might be if size and money were not driving forces. We applaud the smaller, less wealthy soccer nations for devoting so much of themselves to competitions on the world stage and hope to give them their due with a risen AROSE status.
Footnotes:
* For those who are not familiar with the approach I used – regression analysis – the best way to understand it is to visualize a scatter plot of points where a country’s FIFA point total is measured in the vertical dimension and the per capita GDP is measured in the horizontal dimension. Let’s keep it simple for the time being and not consider the additional impact of population. Since FIFA points and per capita GDP are positively related, the cloud of points on this plot will have a general upward slope. Regression analysis provides the best fitting straight line going through these points. To extend the analysis further, you can think of a scatter of points in 3 dimensions where population is the third variable. The regression in this case provides the best fitting plane through the 3-D scatter. The new ranking index is based on how far above the plane (good) or below the plane (bad) the actual FIFA point total is. Natural logs were used to transform the data entering the ordinary least squares regression to improve explanatory power. More technical details are available upon request, as is the raw data that went into the analysis.
** I may come back to this modified ranking later with a more suitable starting point. The Elo system is a candidate as is one I’m developing myself. FIFA itself plans a system after the World Cup that counts only the last 4 years of results.
*** Some may argue that the O from “of” should be excluded from the acronym, especially if they don’t like what it says about their teams.
[Submitted by Dr. Statto]