Friday, September 4, 2015
FIFA Rankings - Both Flawed and Discriminatory
(Originally posted on this blog on Friday, January 9, 2015 as part of my longer Debate on Football in response to Jerome Champagne.)
The FIFA rankings are not well understood by fans, and even many coaches and players seem to be at a loss to comprehend them. Part of the problem is that people don't realise that they are intended to be reactive (ranking teams' past performances) rather than predictive (identifying which teams are likely to win future matches).
Many may feel that the FIFA rankings are not important; just a meaningless exercise to rank teams for the sake of it with no actual consequences. However, this is not the case.
The FIFA rankings are often used to seed teams in international competitions, including, crucially, the pots used for making the draw for the first round groupings at the World Cup. Teams drawn into a tough group are much less likely to progress than teams drawn into an easy group. International pride is at stake to be sure, but so are coaching careers, winning bonuses for players and prize money payments to national associations.
In addition, certain countries will not permit players to be given work permits if they play for a national team that is ranked too low. This means that excellent players from lower ranked nations are not given the chance to improve by playing in stronger leagues, thus further hindering the nation they play for.
Another factor to consider is that national teams ranked higher are more likely to be invited to play friendly matches than teams ranked lower. This means countries with lower rankings are given less opportunity to play against strong opponents.
Higher ranked countries can also demand a higher match appearance fee than lower ranked countries.
Some coaches may even have a rankings goal built into their contract, meaning their job can theoretically be lost (or retained) based on FIFA Rankings.
All of these points show that the calculations used to determine the rankings must be fair and should attempt to reflect reality as closely as possible.
I see four major problems with the rankings that need to be fixed.
a) For each nation, the total of their points accumulated throughout the year is divided by the number of matches they have played. However, where a country plays fewer than five matches, the total points are divided by five. This means that if a country only plays once in a year, the points gained would be divided by five, meaning that country loses 80% of the points it has gained. I understand the concept that national teams should be encouraged to play matches. However, there are certain regions in the world where nations find it particularly difficult to play five matches.
In particular, the eleven nations that belong to the Oceania Football Confederation rarely play five matches in a calendar year. By downloading the list of full internationals played from 2010-14 from the FIFA website it was easy to see the negative affect on the Oceania nations.
Analysis of Nations Playing Fewer than Five Full Internationals, by Confederation, 2010-14
A quick glance shows that almost all UEFA nations played at least five matches every year during the period 2010-14. The only exception was Faroe Islands, which played four matches in 2012 and 2014, thus losing 20% of any ranking points won.
Similarly, of the CONMEBOL nations, only Bolivia in 2010 failed to play at least five matches.
By contrast, 2011 was the only year that more than half of the Oceania nations played at least five matches. Mostly these matches were played in the South Pacific Games in New Caledonia, and are officially classified as friendlies, the lowest ranking of the match categories. Even New Zealand, which was the only Oceania nation that played at least five matches in 2010, as a result of qualifying for the World Cup Finals, and also in 2014, only played three full internationals in 2011.
Oceania faces the major problem of a very small land area located in a huge area of ocean. Even distances between neighbouring nations can be large. There are no buses, trains or ferries. Every away match requires an expensive flight, for nations already lacking funding.
While nations belonging to other Confederations play an extensive series of World Cup and Nations Cup qualifying matches, often the Oceania teams are limited to a few qualifying matches for the Oceania Nations Cup, and because this tournament evolves into World Cup Qualifying, having failed to finish in the top four, they are excluded from the World Cup and don't play any more matches.
Because of its location, it is a difficult and expensive exercise for New Zealand to arrange friendly matches, especially at home. Because New Zealand's players are spread throughout the world, in Europe, North America, Japan, Australia, New Zealand and now South Africa too, whatever location they play their matches requires a huge logistical effort and financial cost.
As a result, while the national teams of Vanuatu, Fiji or Papua New Guinea are by no means world beaters, they are probably actually much better than their respective FIFA rankings suggest.
I would like to see a way that the minimum number of matches calculation can be adjusted for nations that find it difficult to play matches a s a result of their isolation. Alternatively some development funding should be made available for these nations to play more matches.
b) In a league system three points are awarded for a win, one for a draw and none for a loss. This is a fair and equitable system, because each participant in the league plays every other participant an equal number of times. Where teams are level on points, goal difference is generally used to rank the teams.
However, the FIFA Rankings are not a league system. It is impossible for every nation to play every other nation. If countries could somehow play one match a week, it would still take four years for each team to play every other team once.
Therefore, the ranking system needs to be capable of comparing the results of teams that play completely different opponents. It is for this reason I think that, for losing teams, the strength of the opponents and the losing margin need to be taken into account.
Consider these two hypothetical results in World Cup matches:
Germany 4 Panama 3 after extra time.
Tahiti 6 Panama 0.
In the current FIFA rankings calculations, Panama would receive the same number of ranking points (zero) for their narrow extra time loss to Germany as they would for their thrashing at the hands of Tahiti. It is immediately apparent that this cannot be right. A team that loses 4-3 to Germany is clearly much stronger than a team loses 6-0 to Tahiti. In a league system it is fair enough that both losses earn zero points, but in a ranking system where every team plays different opponents, the strength of those opponents and the margin of victory should be taken into consideration. I do not know what calculation SHOULD be used, but I am sure there is someone out there who can do for the FIFA Rankings what Messrs Duckworth and Lewis did for rain-affected one-day cricket matches.
c) The location of matches is not taken into account.
An away win gained by Venezuela against Bolivia at altitude in La Paz is surely worth more than a home win against the same opponents in Caracas.
An away win earned by Canada against Mexico in the heat, humidity, noise and altitude of the Azteca Stadium is surely worth more than a home win on a cold February evening in Toronto.
d) And now to the most egregious injustice of all: the Confederation Coefficient. This is an artificial number based on the number of victories achieved at recent World Cup Finals tournaments by countries from each Confederation. The theory is that Confederations that achieve more victories at World Cup Finals tournaments are stronger than Confederations that achieve fewer wins, and therefore wins against countries from stronger Confederations should be worth more ranking points.
It is absolutely unjust. It is discriminatory. It smacks of colonialism. It is a concept favoured by those who imagine that UEFA and CONMEBOL nations are much stronger per se than nations from the other Confederations. These people appear to base their opinions of the four 'weaker' Confederations on the performances of El Salvador 1982, Zaire 1974 and Haiti 1974, tournaments played thirty of forty years ago.
Let's imagine some scenarios.
i) Two hypothetical World Cup Finals matches:
Chile 2 Uruguay 1.
Costa Rica 3 Uruguay 1.
Uruguay are currently ranked 10 in the FIFA Rankings.
World Cup Finals matches are given an Importance Value of 4.
The Confederation Coefficient is 1.0 for CONMEBOL and 0.85 for CONCACAF.
Using the current formula, Uruguay would receive zero points for both games, because they lost both.
For winning teams, the current formula is Match X Importance X Opposing Team X Confederation Coefficient.
Chile would receive 3 (Match) X 4 (Importance) X 190 (Opposing Team, based on 200 minus the rank of the opponents) X 1 Confederation (because both teams are from CONMEBOL) = 2280 points.
For achieving the same outcome (actually slightly better because they have a larger winning margin), you might expect Costa Rica would receive the same amount of points. But no.
Costa Rica would receive 3 (Match) X 4 (Importance) X 190 (Opposing Team, based on 200 minus the rank of the opponents) X 0.925 Confederation (the average of 1.0 for CONMEBOL and 0.85 for CONCACAF) = 2109 points.
So Costa Rica would receive 171 fewer points than Chile for beating the same team. This is obviously unjustifiable.
ii) Two hypothetical friendly matches:
Romania 1 Austria 0
Algeria 1 Austria 0
Current rankings are 15 (Romania), 18 (Algeria) and 23 (Austria).
Romania would receive 3 (Match) X 1 (Importance) X 177 (Opposing Team) X 0.99 (Confederation) = 525.69
Algeria would receive 3 (Match) X 1 (Importance) X 177 (Opposing Team) X 0.92 (Confederation) = 488.52.
Algeria are in effect punished because UEFA nations won more World Cup Finals matches than CAF nations did, even though Algeria actually won one match in the 2014 Finals, tied another, progressed to the second round and only lost in extra time to eventual champions Germany, whereas Romania didn't even qualify for the 2014 World Cup Finals, much less win a game.
iii) Two hypothetical friendly matches:
Tajikistan 2 San Marino 0
Liechtenstein 2 San Marino 0
Current rankings are: 135 (Tajikistan), 132 (Liechtenstein), 179 (San Marino)
Tajikistan would receive 3 (Match) X 1 (Importance) X 50 (Opposing Team Minimum) X 0.92 (Confederation) = 138.0
Liechtenstein would receive 3 (Match) X 1 (Importance) X 50 (Opposing Team Minimum) X 0.99 (Confederation) = 148.5.
Why should Tajikistan receive fewer points than Liechtenstein for achieving the same outcome? How are wins in World Cup Finals matches relevant to lower ranked countries that will probably never qualify for the World Cup?
If you still think the Confederation Coefficient is fair, consider the fact that sometimes nations change which Confederation they are affiliated to.
In 2002, Kazakhstan left the AFC and joined UEFA.
In 2006, Australia left the OFC and joined the AFC.
Imagine if Guyana, Suriname, Panama or Mexico decided to leave CONCACAF and joined CONMEBOL. Each win or draw they achieved would suddenly be worth 8.85% more, not because they are suddenly a better team, but simply by way of being affiliated to a Confederation with a better Coefficient.
This makes absolutely no sense.
Why do countries from the CAF, AFC, OFC and CONCACAF tolerate this obvious injustice? It is based on a form of colonial thinking that believes that somehow achievements by Europeans are worth more than the same achievement by Africans, Asians, Pacific Islanders or Central Americans.
What if FIFA World Cup Finals match points were awarded in a similar way. If Italy received 0.995 points for drawing with Paraguay, but New Zealand only received 0.925 points for the same result?
If FIFA wants to stamp out discrimination, eliminating the Confederation Coefficient might be an excellent place to start.