[Ed: This is a repeat of a post from three weeks ago. With the tournament starting today, it seemed like an appropriate time to re-post it.]
Here at World Club Rankings, we focus on club teams, as opposed to national teams. We do weekly rankings, annual rankings, but we don’t do international rankings. It’s not that we don’t appreciate a good international match and we are definitely looking forward to Euro 2008. It’s just that international football just isn’t our thing.
Nevertheless, it’s a relatively slow Saturday on the world club football scene (outside of Germany and France) and we haven’t created a new spreadsheet in some time, and we do love ourselves some math, so we decided to apply some old fashioned college basketball formulas to the Euro 2008 teams because, well, why not?
Here’s what we decided to do. We applied the RPI formulas to the sixteen teams participating in Euro 2008. We decided to use the pre-2004 RPI formula because (a) it is much simpler and time is always limited and (b) it makes it easier to account for draws. How is the old school RPI determined? I’m glad you asked. Basically, it takes team records and breaks them into three components, does a little math, and then adds them together.
What are the three components? You ask a lot of questions. The first component is team winning percentage. For the purposes of this experiment, a draw counts as half of a half of a loss. Winning percentage accounts for 25% of the RPI. The second component is opponent’s winning percentage. This is pretty straight forward – it’s the combined record of your opponents. If a nation plays another nation three times, the opponent’s record is counted three times. Opponent’s winning percentage accounts for 50% of the RPI. The third component is opponent’s opponent’s winning percentage. Same process, taken a step further. Opponent’s opponent’s winning percentage accounts for the final 25% of the RPI.
For the game pool, I went back five years. Why five years? It gives me enough data to give meaning to the math. There’s no magic to the number – it’s just what I decided to use. Also, for the purposes of the team records, I decided to apply all games each nation played against any other nation in Euro 2008. I didn’t count games against non-Euro 2008 nations because, to do that, I’d have to expand the RPI to over 200 nations instead of just 16, which would take forever and, frankly, I have a little bit of a life. Also, again, I believe I was able to get meaningful data from using these sixteen teams. Also, I included all games I could find, including friendlies and games that did not necessarily mean anything at the time they were played. Does this skew the numbers a little bit? Probably. It really doesn’t matter, because this is basically just an experiment and should give us interesting, if not exactly accurate results.
Before I get to the results, let me just say that this is merely an experiment to see how a commonly used formula in one sport applies to Euro 2008. It doesn’t take into account home/away games (like the post-2003 RPI) and is not meant to be definitive. The data I found included 138 games. If I missed one or two (though I tried to be careful), it’s not the end of the world. If you’re looking for any national bias, I am not European, so I don’t really have a dog in this hunt. I’ll be rooting for the home of my grandfather (Germany), but no bias that I am aware of enters the formulas. Finally, this was just done to kill some time on a boring Saturday afternoon. I hope you find it at least half as interesting as I found it to do.
THE RESULTS
Component One – Winning Percentage (25%)
This component is pretty straight forward. Again, assuming most people didn’t bother to read the admittedly verbose prose above, the records are based on all internationals games played between two teams that are in Euro 2008 over the past five years, with a draw counting as half a win and half a loss.
Winning Percentages:
1. Italy (.6500)
2. France (.6429)
3. Poland (.6071)
4. Germany (.5833)
5. Greece (.5750)
6. Netherlands (.5682)
7. Spain (.5667)
8. Romania (.5417)
9. Croatia (.4643)
10. Turkey (.4615)
11. Czech Republic (.4444)
12. Sweden (.4412)
13. Portugal (.4211)
14. Russia (.3750)
15. Switzerland (.3333)
16. Austria (.1538)
Component Two – Opponent’s Winning Percentage (50%)
This component is the combined record of each of the opponents for each national team.
Opponent’s Winning Percentage:
1. Sweden (.5518)
2. Portugal (.5442)
3. Romania (.5318)
4. Czech Republic (.5208)
5. Spain (.5194)
6. Turkey (.5180)
7. Italy (.5168)
8. Switzerland (.5113)
9. Austria (.5042)
10. Greece (.4865)
11. France (.4849)
12. Russia (.4794)
13. Croatia (.4729)
14. Germany (.4712)
15. Netherlands (.4629)
16. Poland (.4435)
A few observations: Teams that not only played above .500, but also did it against teams above .500 were Italy, Spain and Romania. Poland had a great winning percentage, but did it against weak competition. The opponents of both Switzerland and Austria played over .500 – primarily one assumes because those teams got to play Austria and Switzerland.
Opponent’s Opponent’s Winning Percentage (25%)
I don’t think I can explain it better than that. It is what it sounds like. It may be a tad convoluted, but I didn’t invent the formula. I just applied it to something to which it was never intended to be applied.
Opponent’s Opponent’s Winning Percentage:
1. Greece (.5131)
2. Poland (.5127)
3. Netherlands (.5118)
4. Spain (.5074)
5. Germany (.5065)
6. Russia (.5034)
7. France (.5018)
8. Croatia (.5018)
9. Turkey (.4996)
10. Italy (.4983)
11. Czech Republic (.4880)
12. Portugal (.4869)
13. Romania (.4857)
14. Sweden (.4856)
15. Switzerland (.4856)
16. Austria (.4816)
I won’t even pretend to know how to analyze these numbers. Now, we take the three components, put them together, add a pinch of salt and stir so nothing gets stuck to the bottom, and see what we get for the final RPI ranking.
EURO 2008 TEAM RPI RANKING:
1. Italy (.5454)
2. France (.5286)
3. Spain (.5282)
4. Romania (.5227)
5. Greece (.5153)
6. Germany (.5080)
7. Sweden (.5076)
8. Poland (.5017)
9. Netherlands (.5014)
10. Turkey (.4993)
11. Portugal (.4991)
12. Czech Republic (.4935)
13. Croatia (.4780)
14. Switzerland (.4603)
15. Russia (.4593)
16. Austria (.4110)
I think that’s pretty interesting. Now, let us apply it to the groups in Euro 2008:
EURO 2008 PREDICTIONS – THE EXPERIMENT
Group A:
(10) Turkey
(11) Portugal
(12) Czech Republic
(14) Switzerland
If Group C is the group of death, Group A has to be considered the group of the dead. I have to say that I was surprised to see any team with Cristiano Ronaldo ranked this low. The fact that all of the teams rank between tenth and fourteenth indicate that this group is wide open and any one of the four teams could advance. With home field advantage, I would not be at all surprised to see Switzerland go through. For the purposes of this experiment, however, let’s say Turkey wins the group and Portugal also advances.
Group B:
(6) Germany
(8) Poland
(13) Croatia
(16) Austria
Continuing with the top heavy (light) Group A/B half of the draw, in which there are only two of the top nine teams in the tournament. This is the group in which there is the clearest delineation between the top two teams and the bottom two teams in the group. It is more of a coin flip to determine the winner of the group, but we’ll go with Germany to win and Poland to advance.
Group C:
(1 ) Italy
(2) France
(4) Romania
(9) Netherlands
Wow. Which of the football draw gods did these teams piss off in order to get such a strong group. With three of the top four clubs in the RPI, as well as the number nine team, it is perfectly conceivable that the best team in the world may not even advance to the Euro 2008 quarterfinals. Despite it being unfair, it should make for some awesome games in the group stage. Although Romania might claim a spot, we’ll go with Italy as the group winner and France as the runner-up.
Group D:
(3) Spain
(5) Greece
(7) Sweden
(15) Russia
First of all, let’s just take Russia out of the equation. I saw the Spain team, which was announced earlier today, and it’s hard to imagine that team being stopped by any other team in this group. That notwithstanding, the numbers indicate that this is a three-way battle for two spots and that any one of those three might spend the knockout stages at home. To keep with the tradition began above, we’ll go with Spain as the winner and Greece as the runner-up.
Knockout Rounds
In the quarterfinals and semifinals, Group A and B are in one draw while Groups C and D are in the other, with the winners of one playing the runner-up of the other in the quarterfinals. If the numbers are accurate at all, a few things are clear: (1) The bottom half of the draw is superior to the top half of the draw, with seven of the top nine teams in the tournament in Groups C or D. Within the two draws, Group B appears to have the best two teams in the top half, while the best teams from the bottom half appear to come from Group C, though Group D is also loaded. If the RPI holds up (and I have no reason to believe it will – remember, this is just a random Saturday afternoon experiment), the knockout stages will play out as follows:
Quarterfinals:
Poland over Turkey
Germany over Portugal
Italy over Greece
France over Spain
Semifinals:
Germany over Poland
Italy over France
Finals:
Italy over Germany
Is this how it’ll turn out? I doubt it, but I had fun plugging all of the data into the spreadsheets and seeing how the formulas played out. Again, I hope you find it half as interesting as I found doing it. It should be fun and I can’t wait!!
Nevertheless, it’s a relatively slow Saturday on the world club football scene (outside of Germany and France) and we haven’t created a new spreadsheet in some time, and we do love ourselves some math, so we decided to apply some old fashioned college basketball formulas to the Euro 2008 teams because, well, why not?
Here’s what we decided to do. We applied the RPI formulas to the sixteen teams participating in Euro 2008. We decided to use the pre-2004 RPI formula because (a) it is much simpler and time is always limited and (b) it makes it easier to account for draws. How is the old school RPI determined? I’m glad you asked. Basically, it takes team records and breaks them into three components, does a little math, and then adds them together.
What are the three components? You ask a lot of questions. The first component is team winning percentage. For the purposes of this experiment, a draw counts as half of a half of a loss. Winning percentage accounts for 25% of the RPI. The second component is opponent’s winning percentage. This is pretty straight forward – it’s the combined record of your opponents. If a nation plays another nation three times, the opponent’s record is counted three times. Opponent’s winning percentage accounts for 50% of the RPI. The third component is opponent’s opponent’s winning percentage. Same process, taken a step further. Opponent’s opponent’s winning percentage accounts for the final 25% of the RPI.
For the game pool, I went back five years. Why five years? It gives me enough data to give meaning to the math. There’s no magic to the number – it’s just what I decided to use. Also, for the purposes of the team records, I decided to apply all games each nation played against any other nation in Euro 2008. I didn’t count games against non-Euro 2008 nations because, to do that, I’d have to expand the RPI to over 200 nations instead of just 16, which would take forever and, frankly, I have a little bit of a life. Also, again, I believe I was able to get meaningful data from using these sixteen teams. Also, I included all games I could find, including friendlies and games that did not necessarily mean anything at the time they were played. Does this skew the numbers a little bit? Probably. It really doesn’t matter, because this is basically just an experiment and should give us interesting, if not exactly accurate results.
Before I get to the results, let me just say that this is merely an experiment to see how a commonly used formula in one sport applies to Euro 2008. It doesn’t take into account home/away games (like the post-2003 RPI) and is not meant to be definitive. The data I found included 138 games. If I missed one or two (though I tried to be careful), it’s not the end of the world. If you’re looking for any national bias, I am not European, so I don’t really have a dog in this hunt. I’ll be rooting for the home of my grandfather (Germany), but no bias that I am aware of enters the formulas. Finally, this was just done to kill some time on a boring Saturday afternoon. I hope you find it at least half as interesting as I found it to do.
THE RESULTS
Component One – Winning Percentage (25%)
This component is pretty straight forward. Again, assuming most people didn’t bother to read the admittedly verbose prose above, the records are based on all internationals games played between two teams that are in Euro 2008 over the past five years, with a draw counting as half a win and half a loss.
Winning Percentages:
1. Italy (.6500)
2. France (.6429)
3. Poland (.6071)
4. Germany (.5833)
5. Greece (.5750)
6. Netherlands (.5682)
7. Spain (.5667)
8. Romania (.5417)
9. Croatia (.4643)
10. Turkey (.4615)
11. Czech Republic (.4444)
12. Sweden (.4412)
13. Portugal (.4211)
14. Russia (.3750)
15. Switzerland (.3333)
16. Austria (.1538)
Component Two – Opponent’s Winning Percentage (50%)
This component is the combined record of each of the opponents for each national team.
Opponent’s Winning Percentage:
1. Sweden (.5518)
2. Portugal (.5442)
3. Romania (.5318)
4. Czech Republic (.5208)
5. Spain (.5194)
6. Turkey (.5180)
7. Italy (.5168)
8. Switzerland (.5113)
9. Austria (.5042)
10. Greece (.4865)
11. France (.4849)
12. Russia (.4794)
13. Croatia (.4729)
14. Germany (.4712)
15. Netherlands (.4629)
16. Poland (.4435)
A few observations: Teams that not only played above .500, but also did it against teams above .500 were Italy, Spain and Romania. Poland had a great winning percentage, but did it against weak competition. The opponents of both Switzerland and Austria played over .500 – primarily one assumes because those teams got to play Austria and Switzerland.
Opponent’s Opponent’s Winning Percentage (25%)
I don’t think I can explain it better than that. It is what it sounds like. It may be a tad convoluted, but I didn’t invent the formula. I just applied it to something to which it was never intended to be applied.
Opponent’s Opponent’s Winning Percentage:
1. Greece (.5131)
2. Poland (.5127)
3. Netherlands (.5118)
4. Spain (.5074)
5. Germany (.5065)
6. Russia (.5034)
7. France (.5018)
8. Croatia (.5018)
9. Turkey (.4996)
10. Italy (.4983)
11. Czech Republic (.4880)
12. Portugal (.4869)
13. Romania (.4857)
14. Sweden (.4856)
15. Switzerland (.4856)
16. Austria (.4816)
I won’t even pretend to know how to analyze these numbers. Now, we take the three components, put them together, add a pinch of salt and stir so nothing gets stuck to the bottom, and see what we get for the final RPI ranking.
EURO 2008 TEAM RPI RANKING:
1. Italy (.5454)
2. France (.5286)
3. Spain (.5282)
4. Romania (.5227)
5. Greece (.5153)
6. Germany (.5080)
7. Sweden (.5076)
8. Poland (.5017)
9. Netherlands (.5014)
10. Turkey (.4993)
11. Portugal (.4991)
12. Czech Republic (.4935)
13. Croatia (.4780)
14. Switzerland (.4603)
15. Russia (.4593)
16. Austria (.4110)
I think that’s pretty interesting. Now, let us apply it to the groups in Euro 2008:
EURO 2008 PREDICTIONS – THE EXPERIMENT
Group A:
(10) Turkey
(11) Portugal
(12) Czech Republic
(14) Switzerland
If Group C is the group of death, Group A has to be considered the group of the dead. I have to say that I was surprised to see any team with Cristiano Ronaldo ranked this low. The fact that all of the teams rank between tenth and fourteenth indicate that this group is wide open and any one of the four teams could advance. With home field advantage, I would not be at all surprised to see Switzerland go through. For the purposes of this experiment, however, let’s say Turkey wins the group and Portugal also advances.
Group B:
(6) Germany
(8) Poland
(13) Croatia
(16) Austria
Continuing with the top heavy (light) Group A/B half of the draw, in which there are only two of the top nine teams in the tournament. This is the group in which there is the clearest delineation between the top two teams and the bottom two teams in the group. It is more of a coin flip to determine the winner of the group, but we’ll go with Germany to win and Poland to advance.
Group C:
(1 ) Italy
(2) France
(4) Romania
(9) Netherlands
Wow. Which of the football draw gods did these teams piss off in order to get such a strong group. With three of the top four clubs in the RPI, as well as the number nine team, it is perfectly conceivable that the best team in the world may not even advance to the Euro 2008 quarterfinals. Despite it being unfair, it should make for some awesome games in the group stage. Although Romania might claim a spot, we’ll go with Italy as the group winner and France as the runner-up.
Group D:
(3) Spain
(5) Greece
(7) Sweden
(15) Russia
First of all, let’s just take Russia out of the equation. I saw the Spain team, which was announced earlier today, and it’s hard to imagine that team being stopped by any other team in this group. That notwithstanding, the numbers indicate that this is a three-way battle for two spots and that any one of those three might spend the knockout stages at home. To keep with the tradition began above, we’ll go with Spain as the winner and Greece as the runner-up.
Knockout Rounds
In the quarterfinals and semifinals, Group A and B are in one draw while Groups C and D are in the other, with the winners of one playing the runner-up of the other in the quarterfinals. If the numbers are accurate at all, a few things are clear: (1) The bottom half of the draw is superior to the top half of the draw, with seven of the top nine teams in the tournament in Groups C or D. Within the two draws, Group B appears to have the best two teams in the top half, while the best teams from the bottom half appear to come from Group C, though Group D is also loaded. If the RPI holds up (and I have no reason to believe it will – remember, this is just a random Saturday afternoon experiment), the knockout stages will play out as follows:
Quarterfinals:
Poland over Turkey
Germany over Portugal
Italy over Greece
France over Spain
Semifinals:
Germany over Poland
Italy over France
Finals:
Italy over Germany
Is this how it’ll turn out? I doubt it, but I had fun plugging all of the data into the spreadsheets and seeing how the formulas played out. Again, I hope you find it half as interesting as I found doing it. It should be fun and I can’t wait!!
Posts
