The BBC published an interview with BigSoccer's own @Edgar about FIFA rankings, and the possibility of manipulating a team's ranking to affect the seeding process: http://www.bbc.co.uk/news/magazine-25134584 Particularly instructive is the top-ten FIFA rankings table as it would be if no teams had played any friendly matches.
Seeds: 1. Spain 2. Brazil 3. Germany 4. Argentina 5. Netherlands 6. Italy 7. Belgium 8. Chile No longer seeds: 9. Switzerland 10. Uruguay (out of top ten) Colombia
Thanks. I think most people would agree that Italy and the Netherlands are better choices as a seeded team compared to any of Switzerland, Colombia or Uruguay. However, a few might disagree with Chile being rated above Colombia or even Uruguay. I am agnostic when it comes to Uruguay, since in their last showing in Conmebol, in a league format, they finished 5th and that dents their Copa America title and makes their 4th place finish in WC 2010 look dated. But for the same reason, I am not entirely sure Colombia should be rated behind Chile? All the same, interesting to see how the rankings would look if friendly matches didn't count.
Chile got more points than Uruguay during 2011-13, so makes sense that Chile is above Uruguay. Also Chile beated them 2:0 this year, and in the last friendly (2010) we beated them too. Of course a ranking is a sort of "photograph", useless as a prediction system, so Brazil, Uruguay or any other NT can win the WC not being ranked as Nr.1 or Nr. 2. Finally, the only reason that I see to rank Chile over Colombia is because in this year Chile got more points than Colombia. But during the whole period 2011-13, they got more points than us.
The real point of the article, though, is that teams can manipulate their ranking by choosing how many friendlies to play and against whom: