The “triumphant” return of The Knack of the baseball world

vorosmccracken.com header image 2

World Cup 2010: The Semifinals

July 6th, 2010 · 12 Comments

We’re down to the last four teams, and the favorites, Brazil, have been knocked out. The good news for them, I suppose, is that when they got knocked out, they also became the first team to qualify for the 2014 World Cup. Silver linings and such.

The National Team Rankings have been recalculated (this is the most I’ve ever recalculated these things in such a short time frame) and because of Germany’s impressive back to back wins over teams ranked just ahead of them, they’ve surged into fourth place. Now three of the final four teams are ranked in the top four (Spain and the Netherlands remain 2nd and 3rd respectively, same as pre-tournament). The rankings can be found here.

And here is the table for the remaining four teams and the chances of each to finish in each of the remaining four spots:

Country4th Place3rd PlaceSecond Place Winners

The numbers assume that the third place game will be contested as tightly as any other World Cup game, and we all are pretty sure that’s not going to be the case. So obviously take the differences between 3rd and 4th with a grain of salt.

I wasn’t sure which team would be the favorites now. Though Spain is the highest ranked team remaining, they have a more difficult opponent than the Netherlands do (though Uruguay I suspect will give the Dutch all they can handle). The numbers however suggest Spain is the favorite followed by the Dutch, Germans and Uruguayans in that order.

Here are the semi-final match numbers:

Team          Goals    Advance%
Netherlands     1.4      64.97%
Uruguay         0.9      35.03%

Spain           1.4      60.66%
Germany         1.0      39.34%

A request: if anyone has a link to where I can find the pre-cup National Team ratings for ELO, Nate’s SPI and for Castro, could you post it in the comments. I’ve come up empty thus far. After all the games are done, I’m going to do a quick comparison to see how various systems did. This will help me identify where my rankings could use some improvement and what to tweak. The ELO site has been down for weeks. Thanks.

Tags: Soccer!! · South Africa 2010 · Uncategorized

12 responses so far ↓

  • 1 Stevan // Jul 6, 2010 at 7:21 am

    Hi Voros,

    since Elo is down since July 21st
    is keeping the elo-rankings up to date, see:

    What is the internet-adress for “Castro Ranking”, I don’t know it yet?


  • 2 Edwin // Jul 6, 2010 at 7:46 am


    I do have the Soccer Power Index Offensive and Defensive ratings from the last week of May (this is, pre World Cup) for the 32 teams that qualified. What I did not saved was the overall rating. Something else: It is an access database, and it is in Spanish, although I don’t think that last part would be a problem (hint: Alemania = Germany). Anyways, you can grab it the from here:


    I hope that helps…

  • 3 Amir // Jul 6, 2010 at 9:34 am

    I have actually made a comparation between your ranking(s), ELO’s and FIFA’s for the group stage:
    Unfortunately you came only second after ELO.

    Of course I saved the ELO ratings before the tournament:

  • 4 Voros // Jul 6, 2010 at 10:13 am

    Thanks Amir. What I’m going to do is a standard confidence pool for the 64 games. Each system assigns a score of 1 to 64 for each game (1 being the game the system is least sure about, 64 being the most) and then see how many points each system gets. This fixes the problem that FIFA’s rankings can’t easily be translated into any sort of meaningful goals or win%.

    This little contest on rec.sport.soccer had me ahead of ELO after the round of 16, though this one rates the systems on the ability to predict advancement probabilities.

    In any event, what you’re likely to see, other than for maybe FIFA, are a bunch of systems that don’t differ very much from one another to the extent that this sort of sample size isn’t likely conclusive anyway.

  • 5 Voros // Jul 6, 2010 at 10:14 am

    Thanks Edwin. I can handle the access database and will figure out from Nate’s explanations how to use the two ratings.

  • 6 Matt // Jul 7, 2010 at 10:49 am

    If you do figure out from Nate’s explanations, can you post something here too? :-) Because I’ve tried to read all of his explanations but the final formula is missing as far as I can tell. He says it’s a non-linear regression to combine the offensive and defensive numbers to get a prediction, but that’s as far as he goes.

    One comparison I’d like to see is a simple one like this: sum up all the percent chances of winning for the favorites in each match. Did the favorites win that often over the whole tournament?

    This would be interesting because it would give a decent sample size, and would help calibrate the system overall.

  • 7 california viola // Jul 7, 2010 at 1:37 pm

    World Cup 1982-2010:

    Champions: 5 from Europe, 3 from South America, 0 from Asia, Africa, Concacaf, and Oceania

    Finalists: 11 from Europe, 5 from South America, 0 from Asia, Africa, Concacaf, and Oceania

    Semifinalists: 25 from Europe (Belgium, Bulgaria, Croatia, England, France, Germany, Italy, Netherlands, Poland, Portugal, Spain, Sweden, Turkey), 6 from South America (Argentina, Brazil, Uruguay), 1 from Asia (South Korea), 0 from Africa, Concacaf, and Oceania

    By all means, less European teams in the World Cup and more Venezuelas, North Koreas, South Africas, El Salvadors, and Fijis.

  • 8 dorian // Jul 7, 2010 at 3:57 pm

    Looking at the last four World Cups, where 16 teams in each WC failed to get out of the group stage, here is the percentage of teams by confederation that did not advance:

    S. America 26%
    Europe 39%
    Oceania 50% (includes AUS ’06)
    N. America 54%
    Asia 75%
    Africa 81%

    Only one (S. America) had less than one-third, while two had more than two-thirds.

    Source: en.wikipedia.org/wiki/National_team_appearances_in_the_FIFA_World_Cup#Results_by_confederation

    In the spirit of relegating the worst performing teams as Edgar touches upon (www.football-rankings.info), focusing on the worst 16 teams (rather that the best 4 or 8) seems more important.

  • 9 dorian // Jul 7, 2010 at 4:16 pm

    Looking at just this WC, here is the decomposition by confederation by qualifying position (admittedly a bit apples to oranges) and failing to advance out of the WC group stage:

    S. America:
    – All positions from qualifying ( #1 through #5 ) advance: 0% fail

    – #1 in qualifiers, 5 of 9 advance: 44% fail
    – #2 in qualifiers, 1 of 4 advance: 75% fail (only Portugal advances)

    – #1: 100% fail

    N. America:
    – #1: 0% fail
    – #2: 0% fail
    – #3: 100% fail

    – #1: 50% fail
    – #2: 50% fail

    – #1: 80% fail

    So, worst than two-thirds:
    – Second-place teams from Europe
    – Third-place team from N. America
    – First-place teams from Africa

  • 10 dorian // Jul 7, 2010 at 4:17 pm

    … also worst than two-thirds:
    – #1 from Oceania


  • 11 california viola // Jul 10, 2010 at 12:34 pm

    2010 World Cup: Europe gets first, second, and third place (as in 2006)

    Since the expansion of the WC (1982-2010):

    First places: 5 Europe, 3 South America, 0 Rest of the World
    Second places: 6 Europe, 2 South America, O Rest of the World
    Third places: 8 Europe, 0 Rest of the World

    Total: 19 Europe (nine different teams), 5 South America (two teams), 0 Rest of the World

  • 12 california viola // Jul 10, 2010 at 12:38 pm

    A quick summary of the 2010 World Cup:

    Europe: gold, silver, and bronze
    South America: all teams eliminated by European opponents (except Chile eliminated by Brazil)
    Africa: all out in first round except one
    Concacaf: no team advanced to quarterfinals
    Asia/Oceania: no team advanced to quarterfinals

Leave a Comment