# World Cup Predictor Mathematical Model

by

Berry Berry Jenius of Berry Berry Easy would like to showcase a fun (meaning: not to be taken too seriously) and non-scientific (meaning: it comes from his head) manner of predicting World Cup results (overall and individual matches). It attempts to use pseudo-scientific manner to look at the art of prediction. Better to use a proper prediction method than gut feelings in the long term. Gut feelings allows good prediction when inspired, prediction model allows more objective review for long term. The best part is you can fine-tune this model yourself to suit your own beliefs.

Berry Berry Jenius’ World Cup “Second Round Matches” Predictor Mathematical Model for World Cup 2010 in South Africa (Prediction Model Release to Public Domain before Start of Second Round Match involving Uruguay and S. Korea)

This model predicts the scores from the knock-out stage onwards using few assumptions, which are:

1. Tayar pancit” syndrome where top performance of a team cannot be sustained for too long in a month long tournament, so teams that performed well early will be ‘penalised’ in this mathematical model. (Argentina and holland are great examples in the 2006 World Cup where they thrashed their initial opponents)
2. Fresh legs” syndrome where teams that had extended matches will invariably feel tired in the next match. So teams that played extended matches (penalty/extra time) will be ‘penalised’ in the next round’s prediction. (South Korea in 2002 played a lot of extended matches until they become toothless in their match against Germany in the semifinal match)
3. Fear factor. Face it, European and South American teams tend to have the fear factor. So they will have a slight ‘advantage’ given to them for the entire duration of the tournament. (So far, proven with South Americans all qualifying for the next round)
4. Momentum“. Winners tend to have winning mentality. Once you start winning, you tend to be more confident. (It took South Korea many World Cups to win their first game, and when they do, it is off to the semis) So the last game a team plays have great bearing on the form.

So, for the second round match 5 factors are selected to predict form of each team,and the 5 factors are:

1) Points obtained in group stages (Pts):

• If a team obtained : 9 points , they get a weighting factor of = 0.7
• If a team obtained : 5-7 points , they get a weighting factor of = 1.0
• If a team obtained : 3-4 points , they get a weighting factor of = 0.4
• Why do teams that get 9 points (maximum) get lower weighting factor? This is due to the “tayar pancit” syndrome. But when teams struggled to enter the knock-out stages, it probably mean that they are too weak.

2) Goal Difference (GD):

• If a team has a goal difference of > +5, they get a weighting factor of = 0.8
• If a team has a goal difference of 3-5, they get a weighting factor of = 1.0
• If a team has a goal difference of 1-2, they get a weighting factor of = 0.6
• If a team has a goal difference of <1, they get a weighting factor of = 0.4
• Once again, teams that peaked too early is slightly penalised with a slightly lower weighting factor. Nonetheless, weak teams with poor goal difference is still weighed down.

3) Goal Against (GA):

• If teams conceded no goals in the group stages, they will get a weighting factor of = 0.8
• If teams conceded 1-2 goals in the group stages, they will get a weighting factor of = 1.0
• If teams conceded 3-5 goals in the group stages, they will get a weighting factor of = 0.7
• If teams conceded > 6 goals in the group stages, they will get a weighting factor of = 0.5
• The numbers are chosen based on a multiple of numbers played, which is n=3. Teams that are strong defensively are given high weighting factors, but teams which have yet to let in a goal will be slightly penalised as statistically speaking, it is a matter of time before they do.

4) Continent (C):

• A team from Europe will get a weighting factor of = 1.0
• A team from South America will get a weighting factor of = 1.0
• The rest will get a weighting factor of = 0.8
• This unfair advantage given to European and South American teams is due to their ability to ‘strike fear’ into teams before they even play. Teams won’t admit it but it is true to a certain extent.
• (It is tempting to pull Europe’s weighting factor down to 0.9, considering that they are performing poorly as a continent)

5) Last group game results (LG):

• A team that won their last game is given a weighting factor of = 1.0
• A team that drew their last game is given a weighting factor of = 0.8
• A team that lost their last game is given a weighting factor of = 0.5
• Momentum is everything in a short tournament like this, provided you don’t peak too early. Slowly building momentum is better.

So, the FORMULA:

For second round predictor, the % weight for each factor are:

1. Point scored in group games (Pts): 20%
2. Goal difference (GD): 15%
3. Goal against (GA): 25%
4. Continent (C): 10%
5. Last group game results (LG): 30%

Final formula:

Form Factor: [(Pts) x 20] + [(GD) x 15] + [(GA) x 25] + [(C) x 10] + [(LG) x 30]

The form factor for each calculated teams are as follow (you may contact me for the raw data if required)

 Teams Form Factor Argentina 91 Holland 94 Uruguay 95 Brazil 94 Germany 100 Japan 92.5 Spain 94 Chile 79 Portugal 86 Paraguay 88 USA 85 England 94 Mexico 65.5 Ghana 62.5 S. Korea 59 Slovakia 66.5

How to interprete the form factor?

When two teams meet in the second round, the team with the higher form factor will win the match. At second round, the interpretation of winning will be as follow:

• If the difference of the form factor between the two teams are:
• If difference of form factor less than 5, then win by penalty
• If difference of form factor is 5<x<10 – win in extra time
• If difference of form factor is >10, win in regular
• Every 10 predictor point advantage = 1 goal win

Second Round Prediction (Form factor in parenthesis) :

• Uruguay (95) vs S. Korea (59) – Uruguay to win by 3 goals
• USA (85) vs Ghana (62.5) – USA to win by 2 goals
• Netherlands (94) vs Slovakia (66.5) – Netherlands win by 2 goals
• Brazil (94) vs Chile (79) – Brazil to win by 1 goal
• Germany (100) vs England (94) – Germany to win in extra time
• Argentina (91) vs Mexico (65.5) – Argentina to win by 2 goals
• Paraguay (88) vs Japan (92.5) – Japan to win via penalty
• Spain (94) vs Portugal (86) – Spain to win in extra time

===================================================

Berry Berry Jenius’ World Cup Winner Predictor Mathematical Model for World Cup 2010 in South Africa

The earlier model is only valid for the second round, subsequent knock out round prediction would require a change in the model. So, the extra factors to consider are:

6) Winning Margin (WM) in regular time:

• If a team win by >1 goal , they get a weighting factor of = 0.6 (QF), 0.8 (SF), 0.9 (Final)
• If a team win by 1 goal , they get a weighting factor of = 1.0 (For all stages)
• If a team win by 0 goal , they get a weighting factor of = 0.8 (QF), 0.9 (SF), 0.95 (Final)
• where QF = quarterfinal, SF = semifinal (weighting factor changes according to stage)
• Why do weighting factor change by stage? This is due to form becoming less important as the tournament progress. (They always say, it is anybody’s game in the final, so our progressive weighting factor models this)

7) Previous Form Factor points (PFF) from the previous stage:

• The form factor of a team should be carried over but at a little over the half the level of the previous stage for the next stage. So we suggest a carry over weightage of 60%

So, the new  FORMULA for quarterfinal until the end of the tournament:

For second round predictor, the % weight for each factor are:

1. Point scored in group games (Pts): 15%
2. Goal difference (GD): 15% (Hard to predict so remove as variable)
3. Goal against (GA): 25% (Hard to predict so remove as variable)
4. Continent (C): 15%
5. Last group game results (LG): 5%
6. Winning margin (WM): 5%
7. Previous Form Factor points (PFF): 60%

Final formula:

Form Factor: [(Pts) x 15] + [(C) x 15] + [(LG) x 5] + [(WM) x 5] + [(PFF) x 60/100]

The form factor for teams in the quarterfinals as predicted by the previous model are:

 Teams Form Factor Argentina 84.1 Holland 85.9 Uruguay 91 Brazil 96.4 Germany 96 Japan 89.75 Spain 92.4 USA 84.25

How to interprete the form factor in the Quarterfinals?

When two teams meet in the quarterfinal, the team with the higher form factor will win the match. At quarterfinal, the interpretation of winning will be as follow (value halved from second round):

• If the difference of the form factor between the two teams are:
• If difference of form factor less than 2.5, then win by penalty
• If difference of form factor is 2.5<x<5 – win in extra time
• If difference of form factor is >5, win in regular
• Every 5 predictor point advantage = 1 goal win

Quarterfinal Prediction (Form factor in parenthesis) :

• Uruguay (91) vs USA (84.25) – Uruguay to win by 1 goals
• Netherlands (85.9) vs Brazil (96.4) – Brazil to win by 1 goals
• Germany (96) vs Argentina (84.1) – Germany to win by 2 goals
• Japan (89.75) vs Spain (94) – Spain to win in extra time

===================================================

As for the Semifinals prediction:

Using the same formula, the form factors are:

 Teams Form Factor Uruguay 94.6 Brazil 94.84 Germany 94.6 Spain 92.94

How to interprete the form factor in the Semifinals?

When two teams meet in the semifinal, the team with the higher form factor will win the match. At semifinal, the interpretation of winning will be as follow (value halved from second round):

• If the difference of the form factor between the two teams are:
• If difference of form factor less than 1.25, then win by penalty
• If difference of form factor is 1.25<x<2.5 – win in extra time
• If difference of form factor is >2.5, win in regular
• Every 2.5 predictor point advantage = 1 goal win

Semifinal Prediction (Form factor in parenthesis) :

• Uruguay (94.6) vs Brazil (94.84) – Brazil to win via penalty
• Germany (94.6) vs Spain (92.94) – Germany to win in extra time

===================================================

Finally, for the World Cup 2010 final match prediction:

Using the same formula, the form factors are:

 Teams Form Factor Brazil 95.154 Germany 96.01

How to interprete the form factor in the final?

When two teams meet in the final, the team with the higher form factor will win the match. In the final, the interpretation of winning will be as follow (value halved from second round):

• If the difference of the form factor between the two teams are:
• If difference of form factor less than 0.625, then win by penalty
• If difference of form factor is 0.625<x<1.25 – win in extra time
• If difference of form factor is >1.25, win in regular
• Every 1.25 predictor point advantage = 1 goal win

Semifinal Prediction (Form factor in parenthesis) :

• Brazil (95.154) vs Germany (96.01) – Germany to win in extra time

===================================================

So based in this model, the winner would be Germany. However, this model is best used when it is updated at every round rather than an early prediction like this. Let’s see if this prediction hold? After every round, I will also update the new data to see if the model works, because it seemed unlikely that Germany can win as its path is littered with England, Argentina, Spain and possibly Brazil in the final. A safer bet based on the draw without this model prediction is Brazil and Spain, their paths is easier.

Phenomena not looked upon (maybe in the future):

1. Average age of squad, preferably at overall age of 27-29 will be given higher weighting. (Even so, preferably strikers to be younger than midfielders, which in turn be younger than defenders)
2. Strength of previous opponents as stronger opponents invariably drain away more energy than weaker opponents.
3. Number of rest day before a match. Too long a rest will make players lethargic, while too short will hinder their recovery time.
4. Tradition. It seemed that like most sports, a certain country tend to win frequently while others do not.