> Thanks also for ‘retranslating’ the graphs in words.
Well this is a sort of survival tactic in my profession; you have to pretend you understand what the numbers mean. Otherwise they will think that they can replace me with a computer, or an engineer.
In considering our questions, did you find a way to solve the inverse problem to identify for each question the best
example? Or does this just involve plotting every country's graph? I presume you also
plotted some other graphs of my other suggestions where the intuition
failed or the effect was weak, and you had the grace not to expose my
idiocy in those cases :-)
Actually that might be a fun "app" to produce - we could work out some sort of formulae and then incorporate a "Sim City" style "Sim FIFA" game where players can introduce their own political upheaval and other development factors to see what the effect on the rankings would be - you can market it to budding national FA administrators as a sort of industrial diagnostic tool.
Alternatively it may even be worth writing some sort of
proper academic paper based on this combination of statistical modelling
and football/geopolitical historical guesswork. We just need to put in a
bibliography and maybe add some more facts on top of the intuitive
Overall I agree with almost everything you wrote, so I've just picked out a few points of interest.
> Assuming, T & T and Jamaica have managed to organize the inclusion of in Diaspora living professionals this would correspond to the diagram
This and also, the "dedication" (or self-interest) of the administrators towards getting the national team to succeed, by tightening up a fairly lax and amateurish attitude previously. In otherwords, a new motivation.
> In respect of Venezuela but I'm not sure that the main reason of the 2nd development phase (amazing in its extent) was really the changing of the qualifying format. Well, the change of the format might have favored the motivation to start ‘activities’...
What it brought was continuity. Before this the World Cup qualifiers were short bursts, and very little in between (the Copa America was almost dead before the 1990s). So the South Americans, even including the big 2, did not necessarily run an ongoing national team with consistent coaching etc, instead only ad hoc arrangements and long gaps between meaningful fixtures. Now, perhaps this should not produce a strong effect on actual ability, since the playing standard would be more dependent on club football which was ongoing, but possibly just a lack of regular chances to gain ranking points was itself a handbrake.
> [Venezuela] Oil money was available to provide at least basal structures to develop football.
Football was popular only with immigrant communities, and there were some foreign professionals but it seems the locals did not benefit much. Maybe Venezuela is/was a bit like some of the Gulf countries, especially before Asian football became properly continental from about 1974 (Eastern dominated beforehand).
By the way, under "epiphanies" I forgot to mention Iran from the late 1960s onward, although perhaps this will not show up well on a plot because they had very few matches before this, and then also, post-revolution, the success dried up in the 1980s. But nonetheless it may be interesting, we could see two (small) ephiphanies in the same graph (beginning ~1964 and 1996), something like Mexico's graph?
> For many years the US fell in relation to competitors further and further back – but not by a decreasing level, but because the competitors improved.
Yes this is an important point - you are plotting skill level, not ordinal ranking, so a standstill in your graphs corresponds to a failure to improve which we may interpret as relative decline because as you say competitors (most countries) improve in skill level over time overall until saturation; in other words, a failure to keep up. [even ignoring the dynamic scaling you discussed]
Plotting just ordinal ranking would show additional noise due to third party effects e.g. other results from similarly ranked teams, and new teams joining FIFA (most visible 1950s to mid 1970s).
> For Asian teams a very different course. Extremely erratic
I think this is an artefact reflecting the erratic nature of Asian football's makeup until the mid-late 1970s. Also the large number of matches, and the fact that most ranking systems would not correctly rate the importance of the various tournaments in the region to the different nations involved, calling them all just "friendlies". To some degree this applies to Africa also. Although perhaps I have mistaken assumptions about the system used.
Malaysia's golden era (1970-82) corresponds to a unique generation and a well organised administration with stability relative to others in the region (though maybe the importance of that is overestimated). Here we can also see a country where professionalism (early 1990s) did not impact on the development at all - Malaysian football was already de facto full-time with organized coaching long before official professionalism, with a standing national team playing almost continuously. In Indonesia (early 1980s) professionalism had more of an impact.
Indonesia's early rise and subsequent dip is partly a "head start" effect, evolving before the "true" leading teams in Asia had got into shape. Some other former colonial teams should show a similar effect (e.g. Hong Kong, or in other continents maybe Curacao or Mauritius for example).
The Singapore graph appears to show the power of one star individual (Fandi Ahmad), the graph from 1979-97 reflects well his presence/absence from the national team (perhaps Wales has a similar Bale effect). Probably just a confirmation bias on my part, as Fandi was one of my favourite players growing up in South East Asia, but anyway we can ask a new question: which country's performance was most dependent on one player? Perhaps George Weah? This is a very difficult inverse problem.
> The reasons in the case of Mexico are not in my mind at the moment
Hmm. I can't think of strong enough political reasons other than general incompetence* so I thought only about results-based reasons. I would guess at weak neighbours (few competition fixtures against strong opponents from which to gain points), and also a general lack of competition fixtures in the era (?). The first peak roughly coincides with staging the World Cup (and the Olympics), a trough follows. Around this time CONCACAF decided to merge its championship with World Cup qualifying - this cuts down the competitive fixture list. No real motivation to run a continuous national team. Then in the early 1980s, suddenly a new incentive again as host in 86. Then another depression followed by the age-cheating ban and an underperformance in 94 & 98. Though I would have expected more impact from the invitation to join the Copa America in the early 1990s. And all this is just unjustified waffle anyway.
* Mexico post 1970 did suffer an economic drop due to a dramatic internal welfare program and other realignments (offset briefly by a kick from oil) but this kind of bungling was commonplace in Latin America so I don't think we can lean on it here.
Just an aside: if I was asked to draw by hand England's graph (of dynamic ability), without thinking hard, I would draw a similar shape to your Mexico graph :-) Albeit less exaggerated and with more of a decline from 2006 to now.
Other graphs I would still like to see:
Libya (reasons as mentioned above)
Israel (may reflect interesting features of football history, or just random noise)
Thailand (somehow they escaped the SEA humilation episode above ;-)
Costa Rica (maybe some resemblance to Mexico, if my thoughts above were valid)
Bermuda (just for fun)
And yet another question just to finish off : which country has the most ideal (sigmoidal) graph, i.e. the smoothest classical development? Above, the best were USA and Panama. Others could be (idle guesses)... Jordan? Burkina Faso? Republic of Ireland?
Or, which has the most consistent linear growth? Above the best were Japan and possibly Qatar. Others could be... Nigeria? Vietnam post 1989 ?
[OK OK, those were two questions...]