Comparing FIFA and ELO Football Rankings

First of all, as a former boxing fan, I can tell you that nothing is more infuriating that "well, it depends on whose ranking system you go by." I think NCAA football fans deal with the same thing, or has that been solved with the BCS? Anyway, as annoying as it is, the comparison can be kind of instructive. Read on and see if you agree.

First, let's go over the parameters: FIFA's ranking system weighs the world's national sides (I believe there are currently 209), with #1 being the best. These rankings are updated about once every month, with any results taking place in the mean time forced to wait until the next ranking. Like the UEFA country coefficients we looked at earlier in the month, they use a rolling period of results, with older results "dropping off" with the passage of time. But unlike those UEFA coefficients, the FIFA rankings also weigh newer results more heavily than older ones, in addition to eventually dropping off older results entirely. FIFA's official rankings also weigh the "match status" and the strength of the opposition, which UEFA coefficients do not (although, in a sense, I guess you could say that UEFA's added points for Champions League group stage wins are an approximation). The result is that teams climb the ranking by winning games (duh), but especially by (1) winning big matches, (2) winning the biggest tournaments, and (3) beating the best opponents.

The Elo rankings attempt to answer the same question, but do so through the Elo ranking system. This is a series of mathematical formulae, originally developed for ranking chess players, that tabulates results and gives weight to different underlying variables (to the extent that they can be mathematically calculated). If Wall Street joins us in the comment section, he can probably lend some more expertise here, but my understanding is that Elo rankings are centered around a mathematical constant, and that results are weighted mostly based upon a side's opponent's strength (based, in turn, on how they've performed recently, calculated by their opponent's strength). If you're into college basketball, I think a good (but not exact) comparison would be the RPI. If you have a stomach for it, you can read about the FIFA formula here and the Elo football ranking formula here, but the long and short of it is this: both are ranking systems that attempt to compare national football sides based on wins, losses, and draws, and based upon strength of opposition, importance of the match, and recency of results - "who have they beaten, and when?" Also, it's important to remember that the current Elo rankings contemplate this past Friday's results, while the FIFA rankings do not.

Anyway, here are the current top 10 under each system:

And some scattered "also of interest" sides under each system:

So, some things that jump right out when comparing the lists (keep in mind that FIFA's official rankings don't include the most recent round of qualifiers):

1) Huge discrepancy as far as Brazil. And Brazil's recent matches have been friendlies (and draws or losses), so you can't even chalk it up to FIFA's lack of very recent updates. For whatever reason, Elo's algorithm like's Brazil's recent results. Interesting, because Brazil kind of stunk it up (by their own high standards) in both the 2010 World Cup and the 2011 Copa America. And because they're hosting the upcoming World Cup, they don't have any meaningful matches until the Confed Cup this summer. I'm curious as to how Elo ranks them so highly. At the same time, if Elo's 18 (Brazil) played FIFA's 18 (Japan), better believe I'd be betting heavily on Elo's 18. Does that weigh in favor of Elo?

2) Belgium, in my opinion, get somewhat short-changed on both ranking systems. They're atop the table in one of the most competitive World Cup qualifying groups, undefeated after 5 matches played, including a decisive ass-whomping of Serbia in Belgrade, and have a good pool of young talent. Yet they're ranked behind Switzerland, Ecuador, and a few other (in my opinion) middling sides in both ranking systems. And, again, you can't attribute it just to recency of events, because the Elo rankings supposedly include this past weekend's result - another clean-sheet win for Belgium on the road - and they're currently ranked lower under the Elo formula.

3) Portugal is ranked much better under the FIFA formula, though is is one instance where FIFA's lack of currency probably DOES make a difference. Presumably that draw to Israel will catch up with Portugal when the next FIFA rankings come out in April.

4) Man, Colombia is riding their one good player about as far as you could possibly hope: top 8 rankings under both systems?

5) Other than Brazil, both systems have mostly the same teams at or near the top. But they have a discrepancy as far as #10: Russia under the FIFA system, Mexico under Elo. It's an interesting comparison ... both traditionally good, dangerous sides, without either being in the truly elite tier ... both with a couple good players currently suited up for big teams in the top leagues, but most of the roster playing in their respective domestic league (Mexico with a slight edge due to Chicharito, though he isn't seeing much of the pitch in Man U's biggest matches) ... Russia looking like a good bet to qualify from a relatively easy group in Europe, while Mexico in an unexpected fight-for-life in a somewhat difficult final CONCACAF qualifying stage.

Who do you think's a better team, Mexico or Russia? And does the comparison of each system's #10-ranked teams (with each ranked lower in the opposite system) serve as a useful yard-stick when analyzing the systems themselves? And which system is better at accurately ranking the top teams - FIFA or Elo? For that last question, at least, you can weigh in with our poll. Thanks for reading, we'll have a Match Day Thread tomorrow for the Germany game.