Football has changed into a ‘killer instinct’ game: the statistical evidence

EURO 2012: Who would have been the quarter finalists based on statistical performance?

By Darsei Canhasi

Chelsea got their 1st corner of the game on the 88th minute and scored. Meanwhile Bayern wasted 20 corners without scoring from them. The game ended with Bayern having 8 shots on target, 18 shots off target, whereas Chelsea had 4 respectively 3 and only 45% of possession. Chelsea were crowded the Champions of Europe..

In this article I would like to show whether statistics are influential in the final score. Through my model I intend to offer you a way to transfer the statistical performance in a result comparable with the actual one. In the end, I will provide you with the statistical result of every game during the group stage of the tournament, and this will be preceded with the corresponding table of each group. This way the 4 quarter finals of the EURO 2012 will be formed.

The Model

I will use a simple and easy to understand model in order to come up with a ‘fair’ result based on the data of the four below-mentioned statistical components.

– number of shots on target……. 0.12
– number of shots off target…… 0.08
– possession percentage………… (50,1%-53%)-0.2; (53,1%-56%)-0.4; (56,1-59%)-0.6; >59.1%-0.7
– number of corners…………… 0.03

Each component has a quantitative coefficient that will get multiplied to the appropriate number, or in the case of the possession will be added.

Statistics are taken from the UEFA official website. Every shot on target brings additional 0.12 points, every shot out of target gets 0.08 points, and every corner gets 0.03 points. As for the possession, I decided to reward the team that has a possession from 50% to 53% with 0.2 points, from 53% to 56% with 0.4, from 56.1% to 59% with 0.6 and the team that has a possession greater than 59.1% with 0.7 points. In the end, the points will be added and the statistical result of the match will be provided by rounding the points. The threshold used for rounding is as follows:

0 to 0.75……..0
0.76 to 1.49…..1
1.50 to 2.49…..2
2.50 to 3.49…..3
3.50 to 4.49…..4

Apart from the 1^st goal, all use the 0.5 criteria. I decided to reward with one goal the teams that score at least 0.76. If it would have been 0.5, a team with only 2 shots on goal, 3 shots out of target and 2 corners would get one goal. This way the number of teams with zero goals would have been way too low (almost non-existent)

Other statistical ingredients are excluded from the model for two reasons: First, the four above mentioned components pretty much explain all the things that determine a result during a game. For example, red cards are excluded because they directly affect the four other factors: a team with 10 players will probably have fewer shots on and off the target, a lower possession and fewer corners. By including the red cards we would have multicollinearity, a statistical term that refers to the case when two variables are highly correlated. In this case, the red cards are directly related with the four other factors and it is highly recommended to remove that variable from the model. Second, with four factors the model is easy to understand and you can even try the results by yourselves.

We can argue about my choice of the model, of course. My intention is not to initiate a debate about the model. I believe that I designed a fair assessment of the statistical indicators for the only purpose of creating a final score solely based on statistics. Various models were tested and this happened to be the best. Not that I favour any particular country, and the country I come from is not in the tournament at all (as things are going on it won’t be for a while). Hence, I will try to prove my points that the tactical approach of football has changed in a recent years.

Illustration

To illustrate the procedure used, I will use the opening match of the tournament between Poland and Greece. Below you can find the analysis.
POLAND

Factor	Number	Coefficient	Points
Shots on target	4	0.12	0.48
Shots off target	9	0.08	0.72
Corners	4	0.03	0.12
Possession	53%		0.20
			Total:       1.52

 

GREECE

Factor	Number	Coefficient	Points
Shots on target	2	0.12	0.24
Shots off target	5	0.08	0.40
Corners	3	0.03	0.09
Possession	47%		0
			Total:       0.73

From the calculations, Poland got 1.53 points and Greece 0.73, which means the statistics score of this game is 2:0 for Poland (with respect to the rounding threshold that I introduced earlier).

The results

Group A

Poland – Greece                       1.53 : 0.73 = 2-0
Russia – Czech Republic           1.73:  2.03 = 2-2
Poland – Russia                        1.68 : 1.86 = 2-2
Greece – Czech Republic         1.19 : 0.88 = 1-1
Greece – Russia                        0.63 : 3.46 = 0-3
Czech Republic – Poland          1:86 : 1.74 = 2-2

Teams	W	D	L	Goal Difference	Points
1.Russia	1	2	0	7:4	5
2.Poland	1	2	0	6:4	5
3.Czech Republic	0	3	0	5:5	3
4.Greece	0	1	2	1:5	1

 

Group B

Netherlands – Denmark            3.09 : 1.08 = 3-1
Germany – Portugal               1.58 : 1.49 = 2-1
Netherlands– Germany            1.66 : 1.47 = 2-1
Denmark – Portugal               1.77 : 2.02 = 2-2
Portugal – Netherlands        2.33 : 1.99 = 2-2
Denmark – Germany             1.19 : 1.89 = 1-2

Teams	W	D	L	Goal Difference	Points
Netherlands	2	1	0	7:4	7
Germany	2	0	1	5:4	6
Portugal	0	2	1	5:6	2
Denmark	0	1	2	1:5	1

 

Group C

Spain – Italy                       2.71 : 1.10 = 3-1
Ireland – Croatia                1.27 : 2.27 = 1-2
Spain – Ireland                   3.82 : 0.70 = 4-0
Italy – Croatia                    1.86 : 1.21 = 2-1
Croatia – Spain                  0.64 : 2.55 = 0-3
Italy – Ireland                    3.90 : 0.71 = 4-0

Teams	W	D	L	Goal Difference	Points
Spain	3	0	0	10:1	9
Italy	2	0	1	7:4	6
Croatia	1	1	2	3:6	3
Ireland	0	0	3	1:10	0

 

Group D

Ukraine – Sweden          1.53 : 1.26 = 2-1
France – England           3.15 : 0.40 = 3-0
Sweden – England          1.29 : 1.85 = 1-2
Ukraine – France            0.95 : 2.10 = 1-2
Sweden – France            1.42 : 3.16 = 1-3
England – Ukraine         1.10 : 2.38 = 1-2

Teams	W	D	L	Goal Difference	Points
France	3	0	0	9:2	9
Ukraine	2	0	1	5:4	6
England	1	0	2	3:7	3
Sweden	0	0	3	1:10	0

 

According to my model, the Quarter finalists would be:

Russia           vs        Germany
Netherlands    vs         Poland
Spain            vs         Ukraine
France          vs         Italy

Is this what you expected?

I believe that the most recent tactical trend in football is the defensive approach. Depending on the coaches, one might use counter-attacks or passive attacks towards the opponents. By passive attacks I refer to the situation where the team attacks during specific periods through the match, but the defensive play is prioritized.

Football has changed into a ‘killer instinct’ game. The number of shots on target is no longer important, and what the coaches are seeking is a decisive goal and this leads them to apply defensive tactics.

Is luck important in football? It is up to you to decide; hopefully this article has shaped your thoughts.

Follow me on Twitter @darseic

—————————————-

Postscript: here are the actual tables

	Team	W	D	L	GD	Pts
1	Czech Republic	2	0	1	-1	6
2	Greece	1	1	1	0	4
3	Russia	1	1	1	2	4
4	Poland	0	2	1	-1	2

Group D

	Team	W	D	L	GD	Pts
1	England	2	1	0	2	7
2	France	1	1	1	0	4
3	Ukraine	1	0	2	-2	3
4	Sweden	1	0	2	0	3