Prediction

Prediction

How will your team do in the first qualifying matches for Euro 2016?

Dr Nicolas Scelles

Posted: September 2, 2014

During the World Cup, I suggested two models to explain scores of national men’s football team games based on matches from August 2012 to December 2013. The first model was based on 13 variables: population, GDP per capita, climate, experience, percentage of players, player quality, foreign managers, technology transfer through managers, home advantage, prize, prize difference in favour of the favourite, prize difference in favour of the underdog, no prize. The second model was based on home advantage and dummies for every team (1 when a team plays at home, -1 when a team plays away).

So as to propose predictions for the UEFA Euro 2016 qualifiers, Wladimir Andreff and I applied the same models in taking into account only European national teams. For the first model, we also compared real scores and those provided by the model. Thus, we calculated the average gap per game for every team that we included then so as to correct the model. From August 2012 to December 2013, the 10 over performing teams were Bosnia and Herzegovina (+1.58), Ukraine (+0.79), Liechtenstein (+0.75), Finland (+0.73), Belgium (+0.64), Iceland (+0.625), Netherlands (+0.54), Armenia and Israel (+0.5), and Ireland (+0.47). The 11 underperforming teams were San Marino (-1.2), Turkey (-0.76), Slovakia (-0.75), Georgia (-0.58), Latvia (-0.54), Croatia (-0.533), Norway (-0.529), Wales (-0.5), and Denmark and Italy (-0.44). 6 teams performed in average as predicted by the model: Czech Republic, England, Kazakhstan, Russia, Slovenia and Spain.

Here are our predictions for match day 1 (7, 8 and 9 September):

Home Away Model 1 Model 1 corrected Model 2
Denmark Armenia 1.51 0.57 0.59
Georgia Ireland -0.19 -1.25 -0.81
Hungary Northern Ireland 1.15 1.15 0.54
Faroe Finland -1.93 -2.74 -2.01
Greece Romania 0.40 0.92 0.95
Germany Scotland 1.96 2.10 2.54
Gibraltar Poland -4.48 -4.08 -4.381
Portugal Albania 0.86 1.37 1.61
Serbia France -0.57 -0.496 -0.37
Russia Liechtenstein 4.11 3.36 3.495
Austria Sweden 0.12 0.49 0.31
Montenegro Moldova 0.63 0.97 1.16
Luxembourg Belarus -1.01 -1.33 -1.29
Spain Macedonia 2.14 2.02 2.19
Ukraine Slovakia 0.68 2.21 1.37
Estonia Slovenia -0.31 -0.37 -0.96
San Marino Lithuania -1.85 -2.98 -3.93
Switzerland England -0.85 -0.68 -0.20
Kazakhstan Latvia -0.08 0.46 0.59
Azerbaijan Bulgaria -0.22 -0.29 -0.15
Croatia Malta 1.71 1.58 2.86
Norway Italy -1.10 -1.19 -0.69
Czech Republic Netherlands 0.04 -0.496 -1.14
Iceland Turkey -1.57 -0.18 0.26
Andorra Wales -2.44 -1.63 -2.33
Bosnia & Herzegovina Cyprus 1.42 3.23 3.44
Israel Belgium2 -0.58 -0.72 -1.07

1Given that Gibraltar did not play in the past, we arbitrarily chose to allocate San Marino’s coefficient to Gibraltar.

2 Israel-Belgium has been postponed and will be played on March 31, 2015.

 

Surprising outcomes are obviously not excluded: in June 2013, Armenia won 4-0 in Denmark whereas our models predict a Danish success. Consequently, as for the World Cup, take care before betting on the basis of our predictions!

About Dr Nicolas Scelles

Nicolas Scelles is a Lecturer at the School of Sport, Stirling University, Scotland. He holds a PhD in sports economics from the University of Caen Basse-Normandie, France. He has articles in international journals including Applied Economics, Economics Bulletin, International Journal of Sport Finance and International Journal of Sport Management and Marketing.