Wednesday, June 6, 2012

FIFA Ronkings

So, the latest FIFA rankings are out, to take account of the latest round of pre-Euro etc friendlies (am presuming the table includes all calculations, the team-by-team breakdown seems to be a missing a few things. Like matches. And points awarded. Don't quote me).

England are 6th. Roar.

The FIFA Ranking calculation is really very simple:
  P = M x I x T x C
"M" is the result of the match
"I" is the status of the match
"T" is the value of the opposing team
"C" is the value of the confederation they play in
(for further details, follow the link above)

The points are averaged per game for each twelve-month period up to the rankings release, depreciating over time, so 100% of the average is taken for the period to the current date, 50% to June 2011, 30% to June 2010 and 20% to June 2009.  These annual averages are added up to get to the total points.  I have some problems with this mixture of addition and averaging, but overall it does include important variables to prevent complete rigging of the points by putting 29 past Andorra while your neighbours are off at a tricky tournament.

The depreciation issue also makes sense due to the changing personnel in national squads due to retirement, injury, form, charges pending etc, although there is a vicious circle effect; if a team does badly and does not qualify for a tournament, for example, they will then not have the same level of higher-rated games available to them as a result so subsequent matches will be 'capped' in value.  A comparison of the average unweighted points per year gives a slightly different picture from the overall rankings. 

Comparing weighted and unweighted points

A brief look at the current situation going into the Euros can be seen in this dashboard.  Pick a team to see their results over time, or CTRL/CMD to multi-select for comparisons.  We can also see clearly that the presence of the hosts, with their lower rankings, skews the situation (compare the average - mean - versus median situation in the grid view).

As the rankings often come in for criticism (obviously, England are better than Brazil), in an attempt to address that, I would suggest that there are at least two additional elements would be helpful in properly assessing each team's chances going into the Euros.  I propose adding an "F" and an "E".

F is for Fatigue

Many commentators have pointed out, fatigue could be an issue at the Euros, and have a listen to the recent Red White and Blue podcast where the esteemed Jonathan Wilson explains how that isn't just a problem for the players.  

A variable is needed to reflect this, which would best be made up of two initial factors: the number of games played for both club and country by each member of the squad, and the number of miles travelled in the course of that (thus taking care of the additional fatigue factor on the Spanish squad being dragged all over the world to play friendlies in their cash-cow capacity as world champions).  

Obviously there will still be some drawbacks to this, as each 23-man squad will include some players who will mostly be doing crosswords in dugouts, but bringing in weighting for 'expected starting eleven', when players in the England and French camps in particular are dropping like flies, seems a bit too tricky to factor in.

The Fatigue Factor will be a reducer on the usual calculation, I would suggest:

1 + ((10,000 - ((Average games per player) x (Average miles per game per player))) / 10,000) = %

10,000 representing an arbitrary standard of 40 games a year with average travel of 250 miles as a starting point. Games here being FTEs with a benchmark of 70 minutes, and miles being all travel, whether or not minutes are played, calculated from the training ground to the stadium and back to the usual drop-off point, and with no allowance for frequent flyer miles.  Thus, if one team has average games per player of 40 but average travel of 300 miles, that would be:

10,000 - (40 x 300 = 12,000) = -2,000 / 10,000 = -.02 so the applicable factor is 80%

See? Easy as.  

E is for Expectation

Again, this would be a reducing factor - expectations being high means an inevitable crash out in the group stages, expectations being low means England could just bloody win the thing. Or something.  As each country may have different expectations of both its own and competing teams, this could get very complicated, so these are simply some suggestions for how the English Expectation Element ("EEE") could be calculated (to then be averaged with other national elements eg FREE, GEE, SEE etc).
  • FE, the Iron-y element - number of times 'ROAR' has been used unironically on Guardian football threads, less number of times used ironically (posters to confirm at time of posting);
  • ASDAQ - the number of player in a team who Alan Shearer could recognise in a supermarket;
  • PREMium - the number of players playing in the Premier League;
  • PODium ranking - has the team been written off by two or more contributors to Football Weekly?  This is usually a good sign for the team in question;
  • Under-under-estimation estimation - can this team traditionally never be underestimated? Again, a good sign;
  • Horse-darkness quotient - a straight reducing factor;
  • Levelling factor - how does Zonal Marking rate their chances, he's usually right.
FIFA Rankings: the Fairly Informative Form Assessment.