rank rankings

[rone]

If you think that FIFA's team rankings are a joke, compare them to these much more realistic rankings (scroll about 40% of the way down).

Comments

[sweh.livejournal.com]

Both are really a joke, IMHO. ELO rating for England is based on 839 games, considering games back as far as 1872. WTF? I think there is a large assumption of consistency that isn't at all in evidence. You can see this at the national level; a team that was hot 30 years ago might be nowhere today (example: Nottingham Forest were big big big in the late 70s and in the 80s (league champions, multiple cup winners, multiple european cup winners). Where are they today? A 3rd tier that almost went bankrupt 5 years ago. A team that sold the first million-pound player (Trevor Francis) aren't even wannabes.

National teams may not show as much variety, since the selection of the best of the best should hopefully keep a level of consistency, but teams do change over time, especially over 130+ years. One good player can cause a team to excel for a period of time then as he fades the team drops back again. IMHO a "decaying" points system which emphasis the results of the most recent games will provide a truer picture of how teams are to be ranked; it decreases the emphasis on historical data (but doesn't eliminate it entirely).

[ronebofh.livejournal.com]

Sure, even among national teams we can point to Hungary as an example of a has-been. The trick would be to figure out where the cutoff for decaying value is.

[eejitalmuppet.livejournal.com]

Apparently FIFA are rejigging their rankings system shortly after the world cup. Among other things they are going to reduce the period over which rankings are calculated, from 8 years to 4, and juggle the regional values. It sounds like it will enable teams to move up and down the ratings a bit faster, but it will be interesting to see whether the results are any more accurate...

[marknau.livejournal.com]

Oh please.

Palau could SO take Bhutan.