Saturday, 7 June 2014

Key Question: Does the 2014 FIFA World Cup contain a Group of Death??

TeamPldWDLGFGAGDPts
 Italy220053+24
 Brazil210154+12
 Argentina200225−30

As mentioned in a previous post, one way to determine if a Group of Death exists is by summing up the World Rankings of all countries within a group which in turn will give an indication of how tough that group is: the lower the summed number of a group, presumably the harder the group is.

But that is just one definition of a Group of Death (ie. lots of highly ranked teams). However is this the proper approach? For example, might we find another definition for a Group of Death? Are seedings or world rankings even accurate in the first place? Even if they are accurate, is merely adding up rankings enough?

Let's take a look..

If we use the latest rankings as at 5 June 2014, we get these results:

Group A: 3 (Brazil) + 20 (Mexico) +  18 (Croatia) +  56 (Cameroon) = 97
Group B:  1 (Spain) +  14 (Chile) +  15 (Netherlands) + 62 (Australia) = 92
Group C:  8 (Columbia) + 12 (Greece) + 23 (Ivory Coast) + 46 (Japan) = 89
Group D:  7 (Uruguay) +  9 (Italy) + 10 (England) +  28 (Costa Rica) = 54
Group E:  6 (Switzerland) + 17 (France) +  26 (Ecuador) +  33 (Honduras) = 82
Group F:  5 (Argentina) + 21 (Bosnia and Herzegovina) + 43 (Iran) + 44 (Nigeria) = 113
Group G:  2 (Germany) + 4 (Portugal) + 13 (USA) +  37 (Ghana) = 56
Group H:  11 (Belgium) + 19 (Russia) + 22 (Algeria) +  57 (Korea Republic) = 109

In short, under this methodology Groups F and H are relatively weak, whilst Groups G and D are the "hardest groups" and could be considered Groups of Death in that sense.

If we use rankings that were used for the World Cup Draw (October 2013) we still get Groups G and D as the harder groups with slight variations to the strength of other groups:

Group A:  11 (Brazil) + 24 (Mexico) + 18 (Croatia) + 59 (Cameroon) = 112
Group B: 1 (Spain) + 12 (Chile) + 8 (Netherlands) +  57 (Australia) = 78
Group C: 4 (Columbia) + 15 (Greece) + 17 (Ivory Coast) + 44 (Japan) = 80
Group D: 6 (Uruguay) + 9 (Italy) + 10 (England) + 31 (Costa Rica) = 56
Group E: 7 (Switzerland) + 21 (France) + 22 (Ecuador) + 34 (Honduras) = 84
Group F: 3 (Argentina) + 16 (Bosnia and Herzegovina) + 49 (Iran) + 33 (Nigeria) =101
Group G: 2 (Germany) + 14 (Portugal) + 13 (USA) + 23 (Ghana) = 52
Group H: 5 (Belgium) + 19 (Russia) + 32 (Algeria) + 56 (Korea Republic) = 112

And using the rankings as at last month (8 May 2014):

Group A:  4 (Brazil) + 19 (Mexico) + 20 (Croatia) + 50 (Cameroon) = 93
Group B: 1 (Spain) + 13 (Chile) + 15 (Netherlands) + 59 (Australia) = 88
Group C: 5 (Columbia) + 10 (Greece) + 21 (Ivory Coast) + 47 (Japan) = 83
Group D: 6 (Uruguay) + 9 (Italy) + 11 (England) + 34 (Costa Rica) = 60
Group E: 8 (Switzerland) + 16 (France) + 28 (Ecuador) + 30 (Honduras) = 82
Group F: 7 (Argentina) + 25 (Bosnia and Herzegovina) + 37 (Iran) + 44 (Nigeria) = 113
Group G: 2 (Germany) + 3 (Portugal) + 14 (USA) + 38 (Ghana) = 57
Group H: 12 (Belgium) + 18 (Russia) + 25 (Algeria) + 55 (Korea Republic) =  110

G and D are therefore consistent Groups of Death. But some comment should be mentioned here.



I would suggest that this methodology of adding up world ranking values is extremely flawed because the FIFA Ranking System itself is flawed for the following reasons:

First, it takes into account friendly games, which whilst retaining a limited degree of competitive nature, often fail to provide a true reflection of both teams' quality as there is no trophy or qualification at stake (and teams are often experimenting with their line up). It's true that friendly matches are already worth less because FIFA currently awards a multiplier of 1 for those matches (as opposed to, say, 2.5 for a World Cup Qualifier), but on the basis I have mentioned above, I would argue that a lower multiplier should be awarded (if at all).

Second, the worth of some of these ranking points is also questionable on the other basis that most World Cup Qualifying matches are played within confederations - whilst the rankings do consider the relative strength of the opponent and discount or multiply points (depending on the strength of the confederation and the importance of the match respectively), FIFA's world ranking formula still does not adequately consider the weakness of a particular confederation. For example, as a baseline measurement, no deduction is given where the country played was in UEFA or CONMEBOL (and this is fine - as they are the strongest confederations). However, a ridiculously generous 0.15 discount (ie. multiply points earned by 0.85) is granted to points earned against an Oceanic team - one would, and should, argue that the discount ought be higher as Oceanic teams are incredibly weak.

However, this is not to say that the rankings aren't useful. There are, however, other ways of measuring a Group of Death - though all of them can and should be combined with the "Sum of World Rankings" approach.


Perspective Approach ("What is the sum of the rankings of the other three teams?")

This approach goes like this: Pick one team in the group. Now, from the perspective of that particular team selected, would this group be a Group of Death?

The answer is obviously a resounding YES if one picks the lowest ranked team in a group of otherwise highly ranked teams. Thus, going on this logic, in addition to Groups G and D, Group B certainly looks like a Group of Death from the perspective of Australia - when adding up rankings of the other three teams Group B has the lowest summed "3-team" rank out of all other groups (apart from G and D, depending on which rankings you use above).

Team
PldWDLGFGAGDPts
 Spain32010006
 Netherlands32010006
 Chile32010006
 Australia30030000
Figure A: A possible final outcome of this group...

Group D is clearly also a Group of Death from the perspective of Costa Rica, as is Group G from Ghana's perspective.

Underrated or Overrated Approach

This approach is quite subjective, or at least, more subjective than other approaches.

It asks: Are any teams UNDERRATED or OVERRATED such that its ranking does not properly reflect its true competitive status?

For example, on that definition, Group C could be a Group of Death. Japan's ranking of 47 is arguably underrated given their strong displays at the Confederations Cup (despite their losses) and their strong and consistent record at previous important tournaments (World Cups and AFC Asian Cups). One might also reason that Ivory Coast has a star-studded team and should do well this World Cup (of course, a team is more than the sum of its individuals..).

Team
PldWDLGFGAGDPts
 Colombia00000000
 Greece00000000
 Ivory Coast00000000
 Japan00000000

The difficult issue though is assigning values to underrated teams. For example, should we halve Japan's ranking if we are so confident that they are underrated? Or do we just subtract 10 places? Do we compare Japan to another team that plays like them?

On this approach though, Group G to me definitely seems like a Group of Death as Ghana is arguably an underrated team given their consistency in the last 2 World Cups.

Comparisons

One might even compare teams under the Underrated or Overrated Approach. For example, if one makes the assumption that Australia is just as strong as Ghana because it drew 1-1 with Ghana in the 2010 Group Stages (a very dangerous assumption to make), then arguably, Group B might also be a candidate for a Group of Death as Australia is "underrated" on that logic.

But, again, this logic is quite flawed and subjective.


Competitive Group Approach ("Is the race for Second Spot close?")

One rogue or outlandish method of determining a Group of Death is to ask: Are all the countries of a SIMILAR world ranking? If so, you have a Group of Death because anyone can win any selected game given the close proximity of all the world rankings.

Hence, the focus of this approach is not on Absolute Strength but Relative Strength - how strong are the teams relative to each other? Does everyone have a real, credible chance to fight it out for Second Place?

In this instance, Group F and Group H (along with Group E) could arguably be Groups of Death as Second Place is closely contested, given the relative closeness of the ranks of the non-seeded teams. Indeed, closeness may not even be the right term for it - rather, it is more the fact that, within a group of lower-ranked teams, they all arguably have a reasonable probability of winning their matches. Of course, Groups D and G might still be Groups of Death under this method, but of a different class.

Team
PldWDLGFGAGDPts
 Argentina00000000
 Bosnia and Herzegovina00000000
 Iran00000000
 Nigeria00000000
Team
PldWDLGFGAGDPts
 Belgium00000000
 Algeria00000000
 Russia00000000
 South Korea00000000

However, this approach isn't favourable because the traditional use of the term "Group of Death" refers to the Absolute Strength of teams in a group.  In fact, under this method, depending on how you define Relative Strength, almost all groups in theory might be Groups of Death.


Conclusion:

In my view, under all the approaches, and this is likely to be an extremely obvious statement but, Group G is a true Group of Death. I do not consider Costa Rica's rank of 28 to be accurate - indeed, it seems overrated which should inflate the real value of Group D's Summed World Ranking.

Andre Lim

No comments:

Post a comment