vorosmccracken.com

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

vorosmccracken.com header image 2

The 2010 World Cup Group Stages, Match by Match

June 10th, 2010 · 3 Comments

Alright, tomorrow is the big opening between Mexico and the host country. Here is how the National Team rankings handicap each match in the group stage. The way this is done is by separating each game into 33 equal scoring chances for each team, and their chances for each would obviously be the predicted goals scored divided by 33. From there it’s fairly easy to use the Binomial Distribution to predict who wins and loses.

This used to be different. The system used to use the Poisson distribution to predict these games. Toward the end of qualifying, I changed the setup. What Poission essentially is, is the binomial distribution divided into infinitely small segments. Research I did suggested that as those segments got larger, the better the system did at predicting results, including giving longshots smaller chances of getting results. Truthfully, going even smaller than 33 might produce better results, but not much and 33 covers the famous 31-0 message Australia delivered to the OFC against American Samoa.

So without further adieu and unfortunately without Freddy Adu, here are the matches:

Group A
June 11

Team           Goals     Win    Draw
South Africa    0.82   22.2%   27.8%
Mexico          1.37   50.0%

Uruguay         1.02   31.7%   29.3%
France          1.16   39.0%

June 16

South Africa    0.83   22.1%   27.5%
Uruguay         1.39   50.4%	

June 17

Mexico          1.00   31.5%   29.6%
France          1.14   38.9%	

June 22

South Africa    0.61   19.0%   31.1%
France          1.17   50.0%	

Mexico          1.35   36.9%   26.0%
Uruguay         1.36   37.1%	

Group B
June 12

Argentina       1.58   57.1%   25.2%
Nigeria         0.76   17.7%	

Korea Republic  0.95   33.0%   31.5%
Greece          1.00   35.6%	

June 17

Argentina       1.57   56.5%   25.3%
Korea Republic  0.78   18.2%	

Nigeria         0.92   32.2%   31.6%
Greece          1.00   36.2%	

June 22

Argentina       1.70   56.8%   23.9%
Greece          0.89   19.2%	

Nigeria         0.85   32.4%   33.8%
Korea Republic  0.88   33.8%	

Group C
June 12

England         1.91   59.7%   21.9%
USA             0.98   18.3%	

June 13

Algeria         0.77   22.6%   29.3%
Slovenia        1.26   48.2%	

June 18

England         2.35   79.7%   14.4%
Algeria         0.51    5.9%	

USA             1.48   48.0%   26.3%
Slovenia        1.02   25.7%	

June 23

England         1.87   67.2%   21.0%
Slovenia        0.65   11.9%	

USA             1.87   63.4%   21.8%
Algeria         0.79   14.8%	

Group D
June 13

Germany         1.56   52.6%   25.6%
Australia       0.92   21.9%	

Serbia          1.42   48.7%   27.1%
Ghana           0.92   24.3%	

June 18

Germany         1.68   49.8%   24.3%
Serbia          1.16   25.9%	

June 19

Australia       1.12   41.8%   30.8%
Ghana           0.85   27.4%	

June 23

Germany         1.92   62.7%   21.5%
Ghana           0.86   15.8%	

Australia       0.98   30.9%   29.7%
Serbia          1.14   39.4%	

Group E
June 14

Netherlands     1.46   52.6%   26.7%
Denmark         0.82   20.7%	

Japan           1.30   39.4%   27.3%
Cameroon        1.17   33.2%	

June 19

Netherlands     1.65   61.6%   23.9%
Japan           0.67   14.5%	

Denmark         1.57   50.1%   25.4%
Cameroon        1.03   24.6%	

June 24

Netherlands     1.81   65.5%   21.8%
Cameroon        0.66   12.7%	

Denmark         1.43   46.2%   26.7%
Japan           1.04   27.2%	

Group F
June 14

Italy           1.14   43.0%   30.7%
Paraguay        0.83   26.3%	

June 15

New Zealand     0.95   13.2%   17.8%
Slovakia        2.33   69.0%	

June 20

Italy           2.44   81.1%   13.5%
New Zealand     0.50    5.4%	

Paraguay        1.56   48.3%   25.5%
Slovakia        1.09   26.3%	

June 24

Italy           1.86   59.2%   22.4%
Slovakia        0.95   18.4%	

Paraguay        2.04   72.8%   18.4%
New Zealand     0.57    8.9%	

Group G
June 15

Brazil          2.50   85.8%   11.2%
Korea DPR       0.33    3.1%	

Côte d'Ivoire   1.04   28.2%   27.2%
Portugal        1.38   44.6%	

June 20

Brazil          2.17   68.6%   18.7%
Côte d'Ivoire   0.84   12.6%	

June 21

Korea DPR       0.41    8.9%   23.4%
Portugal        1.60   67.7%	

June 25

Brazil          1.57   56.3%   25.3%
Portugal        0.79   18.5%	

Korea DPR       0.58   12.0%   23.1%
Côte d'Ivoire   1.69   64.9%	

Group H
June 16

Spain           1.86   67.0%   21.1%
Switzerland     0.65   11.9%	

Honduras        1.04   27.4%   26.9%
Chile           1.41   45.7%	

June 21

Spain           2.06   71.4%   18.7%
Honduras        0.64    9.9%	

Switzerland     1.06   30.7%   28.0%
Chile           1.28   41.3%	

June 25

Spain           1.73   62.0%   23.1%
Chile           0.73   14.9%	

Switzerland     1.27   39.8%   27.8%
Honduras        1.12   32.3%

Check back as the games are played and I’ll try and give running updates of each team’s chances as they change.

Tags: South Africa 2010 · Soccer!! · Uncategorized

3 responses so far ↓

  • 1 Amir // Jun 12, 2010 at 1:30 pm

    I’ve decided to make a “Battle of the Rankings” Excel sheet.
    The rankings are FIFA official rankings, ELO rankings and Voros McCracken’s 2 rankings.
    For FIFA and Voros’ win I used May’s rankings.
    For ELO and Voros’ GD I used the latest rankings before the WC.
    For each rankings, all the results probabilities are ordered and get negative points accordingly.
    The probabilities for Voros’ GD rankings are calculated using the Binomial distribution (the way he calculates it).
    The Expectancy for Voros’ win rankings are calculated using the ratio between them. Using the formula P(draw)=1.31*E(pointsA)*E(pointsB) where win is given 1 point and draw is given 0.5 points, we get the probabilities (it was in Voros’ site too.
    For ELO the Expacnacy is calculated the in the regular way and the probabilities are calculated Like Voros’ win Rankings.
    For FIFA it was more tricky, since the points are basically pretty random… At first I decided to use the principal that if each team wins by its expectancy, the difference between both team won’t change. It gave some really close numbers even with high difference between the teams (it shows the rankings aren’t very good BTW), so I decided to use the ratio as in Voros’ Win rankings. Then I used the formula P(draw)=C*E(pointsA)*E(pointsB) where C was derived from the number of draw in previous world cups (22%).

    The Excel sheet can be found here:
    http://www.2shared.com/document/vWMsUpN7/Battle_of_Rankings.html

    I Hope some of you will find it interesting…

    At the end of each round I will update how many negative points each rankings got.

  • 2 dorian // Jun 14, 2010 at 10:14 am

    @Amir,
    Any chance of including the Castrol and ESPN SPI ranking systems?

  • 3 Amir // Jun 15, 2010 at 12:14 pm

    Unfortunately not in this Format, since I didn’t take the probabilities of Castrol before the tournement, and SPI doesn’t say how to calculate theirs.
    I did, however, took the chances they gave for each team to advance and win their group so I might make another model for that…

Leave a Comment