Rating National Teams

by Dave Clark on Nov 11, 2009 12:00 PM PST

Certainly there are the FIFA rankings, with all of their positives and negatives. They now have gotten to the point where they are a decent attempt at backwards rankings of teams.

They aren't predictive though. I like the predictive. BUT, it is so difficult to do. Many nations just don't play much, especially outside of their home region.

Nate Silver, of Baseball Prospectus and 538.com, has entered the fray though for ESPN. When a soccer guy asked me about Silver and statistics I responded "he's the Messi of stats, but with a decade long history of success." That's a really long way of saying "He's Damn good."

Here's his rankings

Brasil
Spain
England
Netherlands
Argentina

I have emailed him to see when we can expect him to enter the club and player rating fray as well. His methodology is here.

OMG, Nate Silver is doing soccermetrics and sharing the methodology!

FanPosts only represent the opinions of the poster, not of Sounder at Heart.

Tweet Comment 10 comments | 0 recs |

Do you like this story?

Comments

With the USA up against Slovakia and Denmark shortly

We can see that the USA at Slovakia has an edge of +0.29 and at Denmark of -0.26. Guess that means we should expect a win and loss or two draws.

I am not a Supporter
I am not a Fan
I am a Sounder
Sounder At Heart

by Dave Clark on Nov 11, 2009 12:21 PM PST reply actions

Holy crap Nate Silver is analyzing soccer.

Seriously, that’s awesome.

by BrianL on Nov 11, 2009 3:53 PM PST reply actions 1 recs

Yeah

Bigger than Beane in my mind.

I am not a Supporter
I am not a Fan
I am a Sounder
Sounder At Heart

by Dave Clark on Nov 11, 2009 5:22 PM PST up reply actions

I have high hopes

I’ve enjoyed Nate’s work on 538. Having a statistician that can write well and is willing to be accessible in his methodology is an all too rare treat.

From the rather narrow perspective of watching the predictive power of the USA position in the FIFA rankings, it has always seemed to me that it was more of a marketing tool than a useful statistical test.

I’m thrilled that November 14th is going to give us a chance to see how well this SPI rating works.

by -jmc on Nov 11, 2009 7:15 PM PST reply actions

what for?

Sorry, but I think such rating, ranking or whatever is quite useless ;)

I know you Americans are crazy on stats, but that +0.29 on Slovakia and -0.26 on Denmark. What’s the use of it?

Argentina 5th? But not in the year 2009 ;)

PS – great to see a real football team in the USA. Impressive that Seattle people are into it :) Watching your team from Europe a bit :)

http://www.flickr.com/photos/loeffle/

by loeffleitroms on Nov 14, 2009 10:07 AM PST reply actions

You see no use in comparing teams based on actual performance?

I’m confused by soccer fans that discount the importance of statistics in analysis.

As for Argentina 5th, I quite agree. I guess we will be able to test this based on this weekends matches. There should be some movement, but if there are only tiny moves as most results went the expected way Nate was basically correct.

I am not a Supporter
I am not a Fan
I am a Sounder
Sounder At Heart

by Dave Clark on Nov 14, 2009 10:19 AM PST up reply actions

No, I don't

… if you see the whole list, you will discover many strange things.

Norway which did lousy in a lousy group, is ranked way better than Slovakia which did great in a hard group.

I think you can’t rank or compare teams by the numbers. Germany usually doesn’t care at all for friendly games (refer to its friendlies before the 2006 WC), while it runs hot “when it counts”.

Or take the Confed Cup – a tournament some teams take serious, while others don’t.

Football is not baseball, where you can predict a guy having problems against lefties, or hitting better at day or night.

Of course it’s easy to predict that Russia at home will more or less likely win against Slovenia ;)

If that guy can predict the results of the 8th division Black Forest, I admit I’m wrong … but until then?

http://www.flickr.com/photos/loeffle/

by loeffleitroms on Nov 14, 2009 10:24 AM PST up reply actions

How many of today's matches would he need to hit?

Or how many getting out of the Group Stage in South Africa?

Is there ever a measure of success that a statitiscian could hit that would make it worthy in your mind?

I am not a Supporter
I am not a Fan
I am a Sounder
Sounder At Heart

by Dave Clark on Nov 14, 2009 10:34 AM PST up reply actions

Could you give me the link to his predictions for todays games?

Africa’s games today weren’t quite hard, as most of them aren’t very important ones. He should have 75 % of them right …

The only upset was the defeat of Tunisia (did he predict that one?).

What I mean is that todays Africa games are much easier to predict than let’s say Norwegian Premier League or 2nd Division Germany, where “everybody beats everybody”.

IMHO most “statistical rankings” in football suck, FIFA ranking included. I think UEFA’s system for clubs is quite ok.

http://www.flickr.com/photos/loeffle/

by loeffleitroms on Nov 14, 2009 10:41 AM PST up reply actions