Friday, December 13, 2013

FIFA Ranking: Tweaking handling of wins

Should FIFA look into changing the handling of wins (friendly or competitive)?

As you might know, the Netherlands and Italy lost out on seeded status because they played (and won) against low ranked team. What will happen now? Will the fees required to play against top team rise because it could affect their ranking? Here's a possible solution: include result of wins in the calculation of the ranking only if it improves the average of the time frame that match belongs to.


After all, why should a team's average drop after a win. They got the maximum points from that game. It's not like they could have done better and failed.

Here's the October 2013 ranking (the one used to seed the teams for the final draw) computed using this modification.

The difference columns are compute using the October 2013 FIFA ranking as reference.

The only difference with regards to the list of seeds would be Netherlands replacing Switzerland.

Note: I didn't go back to 1993 or something, I simply computed the October 2013 ranking. The differences would be greater had I gone back to recompute all rankings using this method.

Rank - Team - Modified points - Rank difference - Point difference


1 Spain 1911 0 398
2 Germany 1462 0 151
3 Argentina 1420 0 154
4 Netherlands 1320 4 184
5 Colombia 1277 -1 99
6 Uruguay 1262 0 98
7 Brazil 1253 4 175
8 Belgium 1242 -3 67
9 Switzerland 1232 -2 94
10 Italy 1194 -2 58
11 England 1182 -1 102
12 Chile 1150 0 99
13 USA 1085 0 45
14 Portugal 1069 0 33
15 Ukraine 1059 5 188
16 Greece 1050 -1 67
17 Côte d'Ivoire 1001 0 84
18 Bosnia-Herzegovina 986 -2 61
19 Croatia 960 -1 59
20 Ghana 911 3 51
21 Sweden 898 4 48
22 Ecuador 895 0 33
23 Russia 893 -4 19
24 Mexico 883 0 29
25 France 881 -4 11
26 Denmark 841 0 17
27 Serbia 816 1 38
28 Algeria 808 4 67
29 Romania 807 0 40
30 Czech Republic 793 -3 10
31 Scotland 785 4 70
32 Costa Rica 774 -1 30
33 Slovenia 769 -3 17
34 Nigeria 747 -1 23
35 Iceland 727 11 94
35 Honduras 727 -1 7
37 Panama 713 -1 11
38 Armenia 711 0 24
39 Venezuela 709 -2 17
40 Peru 688 -1 2
41 Hungary 679 2 43
42 Mali 674 -1 6
43 Turkey 672 -3 2
44 Japan 671 0 37
44 Cape Verde Islands 671 -2 9
46 Iran 667 3 54
47 Burkina Faso 659 5 61
48 Norway 656 -1 24
49 Tunisia 650 -2 18
50 Egypt 641 1 31
51 Wales 636 -7 2
52 United Arab Emirates 621 19 125
53 Paraguay 620 -4 7
54 Australia 604 3 40
55 Montenegro 600 -1 16
56 Austria 596 -3 0
57 Uzbekistan 591 -2 9
58 Korea Republic 583 -2 14
59 South Africa 581 2 41
60 Finland 577 3 39
61 Albania 568 -3 5
62 Republic of Ireland 564 -2 14
63 Senegal 559 1 29
64 Cameroon 555 -5 1
65 Libya 550 -4 10
66 Zambia 538 1 25
67 Slovakia 529 -2 1
68 Bolivia 526 3 30
69 Israel 520 -3 5
70 Poland 515 -1 12
71 Jordan 514 -1 12
72 Guinea 512 -4 0
73 Bulgaria 505 3 18
74 Cuba 494 0 2
75 Sierra Leone 493 -2 0
76 Morocco 488 1 10
76 Togo 488 -1 0
78 Dominican Republic 482 0 8
79 Jamaica 470 3 14
79 New Zealand 470 0 0
81 Uganda 468 4 37
82 Haiti 464 -2 0
83 Trinidad and Tobago 462 -2 5
84 Belarus 445 -1 4
85 Gabon 442 -1 4
86 Azerbaijan 430 2 23
86 FYR Macedonia 430 0 0
88 Congo DR 417 -1 6
89 El Salvador 406 0 2
90 Northern Ireland 399 0 0
91 Congo 394 0 0
92 Oman 393 0 12
93 Ethiopia 386 2 10
94 Angola 383 -1 3
95 Moldova 380 1 11
96 Benin 379 -2 1
97 Botswana 373 1 19
98 China PR 371 -1 6
99 Georgia 352 1 2
100 Estonia 351 -1 0
101 Saudi Arabia 350 0 12
102 Zimbabwe 331 0 3
103 Korea DPR 327 4 17
103 Iraq 327 0 4
105 Lithuania 323 -2 0
106 Qatar 315 -1 2
107 Kuwait 312 2 5
107 Liberia 312 -1 0
109 Central African Republic 310 -2 0
110 Niger 306 0 0
111 Antigua and Barbuda 298 1 4
112 Canada 296 -1 0
113 Guatemala 295 -1 1
114 Guyana 293 0 7
115 Tajikistan 291 1 11
116 Mozambique 282 -1 0
117 Kenya 278 1 4
118 Latvia 277 -1 0
119 Equatorial Guinea 273 0 0
120 St. Vincent and the Grenadines 272 0 1
121 Burundi 271 0 4
121 Lebanon 271 0 4
123 Bahrain 268 0 2
124 Malawi 264 0 1
125 Turkmenistan 254 0 0
126 New Caledonia 249 0 0
127 Afghanistan 248 6 25
128 Luxembourg 247 -1 0
129 Namibia 246 -1 0
130 Rwanda 244 -1 2
131 Tanzania 242 -2 0
132 Suriname 239 -1 2
133 Grenada 233 -1 0
134 Cyprus 219 0 0
135 Sudan 216 1 1
135 Kazakhstan 216 0 0
137 Philippines 213 0 0
138 St. Lucia 203 0 0
139 Gambia 202 0 0
140 Malta 192 0 0
141 Syria 184 0 1
142 Lesotho 183 -1 0
143 Thailand 181 0 0
144 Tahiti 179 0 0
145 Belize 178 0 0
146 Palestine 175 0 0
147 St. Kitts and Nevis 172 0 0
148 Hong Kong 171 0 0
149 Myanmar 169 0 0
150 Kyrgyzstan 161 0 0
151 Vietnam 159 0 0
152 Mauritania 158 0 0
153 Nicaragua 155 0 0
154 India 151 0 0
155 Singapore 149 0 0
156 Chad 148 0 0
157 Maldives 147 0 0
158 Liechtenstein 144 0 3
159 Puerto Rico 141 0 2
160 Malaysia 137 0 0
161 Bermuda 127 0 0
162 Bangladesh 120 0 0
162 Indonesia 120 0 0
162 São Tomé e Príncipe 120 0 0
165 Nepal 119 0 0
166 Sri Lanka 108 0 0
167 Laos 105 0 0
168 Pakistan 102 0 0
169 Dominica 89 0 0
170 Curaçao 88 0 0
171 Guam 86 0 0
171 Solomon Islands 86 0 0
173 Aruba 82 0 0
173 Barbados 82 0 0
175 Faroe Islands 81 0 0
176 Chinese Taipei 79 0 0
177 Yemen 72 0 0
178 Mauritius 62 0 0
178 Samoa 62 0 0
180 Madagascar 57 0 0
181 Guinea-Bissau 56 0 0
182 Vanuatu 53 0 0
183 Mongolia 49 0 0
183 Swaziland 49 0 0
185 Fiji 47 0 0
186 American Samoa 43 0 0
186 Tonga 43 0 0
188 Bahamas 40 0 0
189 Montserrat 33 0 0
190 Comoros 32 0 0
191 US Virgin Islands 30 0 0
192 Cayman Islands 29 0 0
193 Brunei Darussalam 26 0 0
193 Timor-Leste 26 0 0
195 Eritrea 24 0 0
196 Seychelles 23 0 0
197 Papua New Guinea 21 0 0
198 Cambodia 20 0 0
199 British Virgin Islands 18 0 0
200 Andorra 16 0 0
201 Somalia 14 0 0
202 Cook Islands 11 0 0
202 Djibouti 11 0 0
204 Macau 10 0 0
204 South Sudan 10 0 0
206 Anguilla 3 0 0
207 Bhutan 0 0 0
207 San Marino 0 0 0
207 Turks and Caicos Islands 0 0 0

About me:

Christian, husband, father x 3, programmer, Romanian. Started the blog in March 2007. Quit in April 2018. You can find me on LinkedIn.

15 comments:

  1. Your proposal doesn't lead to a very transparent calculation method, would you say ? If I understand it correctly, a match could one month be excluded from the calculation and the next month be included, depending on the fact whether the match points are lower or higher than the time frame average (the match belongs to) calculated for each months ranking.

    I agree that the negative impact of friendly wins for top teams should be dealt with, because it is an illogical side-effect of the chosen calculation method. Which becomes all the more important now the ranking provides the sole seeding mechanism for qualification and final draws of the World Cup .

    There are a few different (and easier) methods proposed on this blog and other fora:
    - skip friendlies altogether in the calculation. This will have a major negative impact on the ranking of teams who don't play a lot of competitive matches. One could argue that for those teams the ranking probably is not the most important thing in the world, but hey, it is a world ranking !
    - give friendlies a higher weight (1.5 or 2) to diminish the negative effect.
    - use the sum of the importance factor over all matches played as denominator instead of the number of matches played to calculate the time frame average (Dorian's suggestion, which appeals to me).

    Anyone other suggestions ?

    ReplyDelete
    Replies
    1. To to the ones you already said I would like to add:

      -Change the weights of the time frames.

      Either I would remove the weights altogether or the weights should be constructed in such a way that they meet the following criteria:
      1) A world cup win against a team ranked in position X, zone Z, always gives more points that any friendly game against an XZ team during the 4 year span. With the current method, a friendly victory could account for more points than a WC victory after weights.
      2) A continental cup win in year Y against a team ranked in position X, zone Z, always gives more points that any friendly game against an XZ team during years Y and years Y+1.

      I think your proposals 1 and 2 are feasible. But the third one would make it harder for people to understand the calculation perhaps. Many don't even understand the current calculation. I think modifying the weights on each time frame is also something very simple that would enhance the calculation.

      Juan (Arg.)

      Delete
    2. An overall suggestion: When winning a game (regardless of the weight of the game), the minimum points to be acquired are the average points of the timeframe of the winning country. That way, winning a game against a (very) low ranked team shouldn't cost more than 1 or 2 points because the last timeframe is by far the most decisive one.

      Delete
  2. Just award the world cup seedings to the last 8 countries that won a World Cup - use rankings only for qualifying. There - everyone is happy.

    ReplyDelete
    Replies
    1. No - this is exactly why no one is happy. Making past success a guarantee of future success from non recent WCs will make sure that no new blood is ever found in football.

      Delete
    2. Abosolute nonsense. We are only talking about 1/4 of the teams in the world cup - it will have no effect whatsoever on 'new blood' - absolutely none.

      Delete
    3. You do realize there are only 8 WC winners right? And that by your method you would be awarding Uruguay a #1 seed from the results of the 1950 WC and England the same privilege from 1966???

      Delete
    4. This is about as logical as letting Blatter and Platini just draw up the groups completely by themselves in their own lavish offices without any press after all teams are known, or letting the same people decide who plays who in round two when all the games of round 1 have been played.

      Delete
    5. If you want to be a seed at the next World Cup, just make sure you win one. There is nothing wrong with rewarding success. As for it being like Blatter and Platini deciding, this makes no logical sense. My option is based on earning your seeding through hard fought victory against tough sides, not wining a qualifying group full of awful awful teams like the Swiss did.

      Delete
    6. Based on recent World Cups and world rankings combined, this probably should have been the 8 seeds: Brazil, Spain, Argentina, Germany, Netherlands, Italy, Portugal & Uruguay. I can't see that top 8 raising many eyebrows, plus it would have meant a neat 8 teams in the European pot and thus no need that ridiculous "Pot X".

      Delete
    7. Teams should be seeded on the basis of past World Cup wins only if they feild the same players from those teams. It will be great to see Bobby Charlton, Ghiggia and Maradona grace a World Cup again.

      Delete
    8. You have two choices here - either seed based on teams who have shown greatness and actually won something that is deserving of being a seed, or try and use some funky-but-flawed method to work out who 'might' show greatness - and end up with trully terrible teams like Switzerland in there.

      The 8 countries that have won a World Cup are all great footballing nations that won the tournament by right and are considered a big game for anybody in the world, irrespective of whether they are currently experiencing a lull.

      If countries like Colombia and Switzerland don't like this approach - then go and win the World Cup and knock Uruguay off the list. Until you have won it though, keep schtum.

      I suspect that from your last name you might not like this idea because it seeds England!

      Delete
    9. In the unlikely event that the 8 previous WC winners don't qualify for the WC. How would you seed the teams? I'm trying to be a constructive critic here. I think you should propose an approach that considers this scenario. You cannot simply say "if that happens we can use the ranking", because that wouldn't be a serious methodology. That is why there are formulas that consider the complete scenario. If you think about it, perhaps you can enhance your proposal.

      Juan (Arg.)

      Delete
  3. I have nothing against England at all, to suggest any grudge on my part based on my name reveals nothing other than your warped mindset. Seeding should be based on performance in the last two world cups, perhaps including qualfication matches. The suggestion that any winner should forevermore be seeded is just an extreme version at the other end of the spectrum from the FIFA rankings only debacle that results in an average Switzerland side being seeded. So what happens if say, the Netherlands or Colombia wins the World Cup which is far from inconceivable. Do we have to find space for 9 seeded teams? Increase the number of teams to 40 to accommodate? I don't think England should be rewarded for the performances of Bobby Charlton and Bobby Moore 45 years ago, nor Uruguay for 63 year old performances, that is even more farcical than the curernt system. They should be rewarded/ or rather/ ranked according to performances that bear some semblance towards current prowess. England would still stand a very decent chance of being seeded if they can recapture form that saw them regularly make the Quarter Finals of the World Cup and European Championships and who could begrudge them if a much less complex and more transparent seeding system less open to cynical manipulation was applied.

    ReplyDelete
  4. FIFA needs to listen to criticism and change the whole ranking system. Refusal to do so is just being ignorant and indifferent to public opinion. Tweaking little things like this is like putting plasters on a dead person. Pointless.

    ReplyDelete