Author: pembeci
Date: 21-09-2010, 14:35
Edited by: pembeci
at: 21-09-2010, 15:21 | The last discussion about the access list made me think of a better way to determine the team numbers and their initial stage. One way could be instead of using the rankings of the countries, we can employ the country coefficients to determine these. This way there wouldn't be complaints about the cut points (i.e the difference between #3 and #4 country or between #12 and #13). Based on the spread of country coefficients each year different number of teams and placements can occur for a certain rank.
For instance let's take the country ranking 2009 which determined this season's access list. In the following calculations, as a rule:
a) Each country's champion will be included in the CL tournament
b) No country can send more than 4 teams
There are 76 teams involved in the current setting. 53 of those are fixed as 53 countries' champions. We have to find a way to select 23 additional teams and at what stage they will start the tournament. We can do this by giving points to each potential country team, sorting these teams and then let's say take the first 22 to determine the teams in the group stage, take the next 5 teams to find direct qualifiers for Q4 and so on.
I'd use the the current number of teams for each stage. My first suggestion is to use the ratio of country's coefficient to the total coeffiecient as a way to determine how many teams a country deserves. For instance England has 79.499 points, the total coefficients of 53 countries is 1002.29. This gives England 6.028 (76*79.499/1002.29) teams among 76 teams. Next we determine the points of each English team. For instance:
England_1 : 6.028
England_2 : 5.028 (6.028 - 1)
England_3 : 4.028 (6.028 - 2)
England_4 : 3.028 (6.028 - 3)
We make this for each country, sort the teams and then determine the berths. Here are the results:
Group Stage (22 teams direct, 10 from Q4)
Title_Holder (7.000), England_1 (6.028), Spain_1 (5.631), England_2 (5.028), Italy_1 (4.770), Spain_2 (4.631), Germany_1 (4.299), England_3 (4.028), France_1 (3.804), Italy_2 (3.770), Spain_3 (3.631), Russia_1 (3.611), Germany_2 (3.299), Ukraine_1 (3.173), England_4 (3.028), Netherlands_1 (2.967), Romania_1 (2.950), France_2 (2.804), Italy_3 (2.770), Portugal_1 (2.765), Spain_4 (2.631), Russia_2 (2.611)
Q4 (5 teams direct, 15 from Q3)
Turkey_1 (2.444), Germany_3 (2.299), Ukraine_2 (2.173), Greece_1 (2.136), Scotland_1 (2.114)
Q3 (13 teams direct, 17 from Q2)
Netherlands_2 (1.967), Romania_2 (1.950), Belgium_1 (1.920), Switzerland_1 (1.915), Denmark_1 (1.854), France_3 (1.804), Italy_4 (1.770), Portugal_2 (1.765), Bulgaria_1 (1.611), Russia_3 (1.611), Czech Republic_1 (1.573), Turkey_2 (1.444), Norway_1 (1.426)
Q2 (32 teams direct, 2 from Q1)
Austria_1 (1.352), Germany_4 (1.299), Ukraine_3 (1.173), Israel_1 (1.156), Serbia_1 (1.156), Cyprus_1 (1.144), Greece_2 (1.136), Sweden_1 (1.114), Slovakia_1 (1.112), Poland_1 (0.979), Croatia_1 (0.935), Finland_1 (0.742), Lithuania_1 (0.733), Ireland_1 (0.720), Latvia_1 (0.695), Slovenia_1 (0.689), Belarus_1 (0.657), Bosnia-Herzegovina_1 (0.657), Hungary_1 (0.619), Iceland_1 (0.505), Moldova_1 (0.505), Georgia_1 (0.505), Liechtenstein_1 (0.417), Macedonia_1 (0.392), Azerbaijan_1 (0.341), Estonia_1 (0.328), Albania_1 (0.303), Kazakhstan_1 (0.246), Armenia_1 (0.227), Wales_1 (0.177), Northern Ireland_1 (0.164), Faroe Islands_1 (0.164), Luxembourg_1 (0.101)
Q1 (4 teams)
Montenegro_1 (0.076), Andorra_1 (0.038), Malta_1 (0.038), San Marino_1 (0.019)
As it can be seen this formula produced better results for stronger countries and make the list more unbalanced. If we made a minor tweak to our team points for the fourth team, the results are more satisfactory:
England_4 : 2.028 (6.028 - 4) (deduction is increased from 3 to 4).
The Access List:
Group Stage (22 teams direct, 10 from Q4)
Title_Holder (7.000), England_1 (6.028), Spain_1 (5.631), England_2 (5.028), Italy_1 (4.770), Spain_2 (4.631), Germany_1 (4.299), England_3 (4.028), France_1 (3.804), Italy_2 (3.770), Spain_3 (3.631), Russia_1 (3.611), Germany_2 (3.299), Ukraine_1 (3.173), Netherlands_1 (2.967), Romania_1 (2.950), France_2 (2.804), Italy_3 (2.770), Portugal_1 (2.765), Russia_2 (2.611), Turkey_1 (2.444), Germany_3 (2.299)
Q4 (5 teams direct, 15 from Q3)
Ukraine_2 (2.173), Greece_1 (2.136), Scotland_1 (2.114), England_4 (2.028), Netherlands_2 (1.967)
Q3 (13 teams direct, 17 from Q2)
Romania_2 (1.950), Belgium_1 (1.920), Switzerland_1 (1.915), Denmark_1 (1.854), France_3 (1.804), Portugal_2 (1.765), Spain_4 (1.631), Bulgaria_1 (1.611), Russia_3 (1.611), Czech Republic_1 (1.573), Turkey_2 (1.444), Norway_1 (1.426), Austria_1 (1.352)
Q2 (32 teams direct, 2 from Q1)
Ukraine_3 (1.173), Serbia_1 (1.156), Israel_1 (1.156), Cyprus_1 (1.144), Greece_2 (1.136), Sweden_1 (1.114), Scotland_2 (1.114), Slovakia_1 (1.112), Poland_1 (0.979), Netherlands_3 (0.967), Croatia_1 (0.935), Finland_1 (0.742), Lithuania_1 (0.733), Ireland_1 (0.720), Latvia_1 (0.695), Slovenia_1 (0.689), Belarus_1 (0.657), Bosnia-Herzegovina_1 (0.657), Hungary_1 (0.619), Iceland_1 (0.505), Moldova_1 (0.505), Georgia_1 (0.505), Liechtenstein_1 (0.417), Macedonia_1 (0.392), Azerbaijan_1 (0.341), Estonia_1 (0.328), Albania_1 (0.303), Kazakhstan_1 (0.246), Armenia_1 (0.227), Wales_1 (0.177), Northern Ireland_1 (0.164), Faroe Islands_1 (0.164), Luxembourg_1 (0.101)
Q1 (4 teams)
Montenegro_1 (0.076), Andorra_1 (0.038), Malta_1 (0.038), San Marino_1 (0.019)
Another method may be using this to determine team points:
England_1 : 79.499 (country coefficient)
England_2 : 39.749 (country coefficient / 2)
England_3 : 26.500 (country coefficient / 3)
England_4 : 19.875 (country coefficient / 4)
This is used in our country to determine the number of deputies in a certain election region at parliament elections.
Here is the Access List produced by this method:
Group Stage (22 teams, 10 from Q4)
Title_Holder (100.000), England_1 (79.499), Spain_1 (74.266), Italy_1 (62.910), Germany_1 (56.695), France_1 (50.168), Russia_1 (47.625), Ukraine_1 (41.850), England_2 (39.749), Netherlands_1 (39.130), Romania_1 (38.908), Spain_2 (37.133), Portugal_1 (36.462), Turkey_1 (32.225), Italy_2 (31.455), Germany_2 (28.348), Greece_1 (28.165), Scotland_1 (27.875), England_3 (26.500), Belgium_1 (25.325), Switzerland_1 (25.250), France_2 (25.084)
Q4 (5 teams direct, 15 from Q3)
Spain_3 (24.755), Denmark_1 (24.450), Russia_2 (23.812), Bulgaria_1 (21.250), Italy_3 (20.970)
Q3 (13 teams direct, 17 from Q2)
Ukraine_2 (20.925), Czech Republic_1 (20.750), England_4 (19.875), Netherlands_2 (19.565), Romania_2 (19.454), Germany_3 (18.898), Norway_1 (18.800), Spain_4 (18.567), Portugal_2 (18.231), Austria_1 (17.825), France_3 (16.723), Turkey_2 (16.113), Russia_3 (15.875)
Q2 (32 teams direct, 2 from Q1)
Italy_4 (15.727), Serbia_1 (15.250), Israel_1 (15.250), Cyprus_1 (15.082), Sweden_1 (14.691), Slovakia_1 (14.665), Germany_4 (14.174), Greece_2 (14.082), Ukraine_3 (13.950), Poland_1 (12.916), Croatia_1 (12.332), Finland_1 (9.790), Lithuania_1 (9.666), Ireland_1 (9.499), Latvia_1 (9.164), Slovenia_1 (9.082), Belarus_1 (8.666), Bosnia-Herzegovina_1 (8.665), Hungary_1 (8.166), Iceland_1 (6.665), Moldova_1 (6.665), Georgia_1 (6.664), Liechtenstein_1 (5.500), Macedonia_1 (5.165), Azerbaijan_1 (4.498), Estonia_1 (4.332), Albania_1 (3.999), Kazakhstan_1 (3.249), Armenia_1 (2.999), Wales_1 (2.331), Northern Ireland_1 (2.165), Faroe Islands_1 (2.165), Luxembourg_1 (1.332)
Q1 (4 teams)
Montenegro_1 (1.000), Andorra_1 (0.500), Malta_1 (0.499), San Marino_1 (0.250)
I think, the results produced by this last method is the most balanced one.
I also liked Platini changes. The difference between being the champion and runner-up is huge. This is called the champions league so I'd prefer to see a champion of a middle ranked country instead of a strong league's #4. I'd suggest adapting the current champion's path and non-champions's path to group stage to this method. So for instance at Q4 and Q3 NC teams will play each other and C teams will be drawn together. The number of these teams can vary each year. If the number of NC teams is even then there is no problem. If it is odd then we have to move one NC team to the other side. How to select that team? If it is the highest ranked team, it'd be unfair to the C team who draw that team. If it is the lowest ranked team than it'd be unfair to the rest of the NC teams. So I think the only solution will be to select that NC team by chance.
I wanted to find the differences (which teams moved up or down) between the current Access List and my lists produced by the last two methods but for today I run out of time. May be tomorrow or another poster can do this for a better analysis. |