What are the greatest upsets in international football based on difference in the rankings?

Joined: February 9th, 2011, 2:58 pm

January 28th, 2013, 3:36 pm #1

Kaizeler wrote:
If one goes by the Elo ratings (which consider also the competition
and margin of victory) that would be Sweden beating Belgium 8-1 at the
1924 Paris Olympics. But feel free to take your pick from the list.

nfm24 wrote:
I think North Korea 1-0 Italy (1966), Japan 3-2 Sweden (1936), or USA 1-0 England (1950).  Maybe also Indonesia 0-0 USSR (1956)


Luca wrote:

I remember other astonishing cases:





30/10/1963 Luxembourg-Netherlands 2-1 (Camille Dimmer 20', 68'; Piet Kruiver 35')



16/06/1982 Algeria-West Germany 2-1 (Rabah Madjer 54', Lakhdar Belloumi 68'; Karl-Heinz Rummenigge 67')



14/07/1988 Australia-Argentina 4-1 (Paul Wade 4', Charlie Yankos 42' and 67', Vlado Bozinovski 80'; Oscar Ruggeri 31')



08/06/1990 Cameroon-Argentina 1-0 (François Omam-Biyik 67')



13/10/1993 France-Israel 2-3 (Ronen Harazi 21', Eyal Berkovic 83', Reuven Atar 93'; Frank Sauzée 29', David Ginola 40)



28/03/1999 South Korea-Brazil 1-0 (Kim Do-Hoon 90')



23/07/2001 Honduras-Brazil 2-0 (Juliano Belletti - owngoal - 57', Saúl Martínez 88')


31/05/2002 Senegal-France 1-0 (Papa Bouba Diop 30')
Quote
Like
Share

Joined: April 7th, 2007, 12:28 pm

January 28th, 2013, 4:47 pm #2

China 0-1 Hong Kong (1985 WCQ) also springs to mind.

1948 Olympics, South Korea 5-3 Mexico

1942 Korea 5-0 Japan (note that Japan occupied Korea and had already "borrowed" its best players for their own team).
Quote
Like
Share

Joined: February 9th, 2011, 2:58 pm

January 29th, 2013, 2:13 pm #3

Actually, the Olympic Games have often been the stage for many "giantkillings".
In 1924, Egypt liquidated Hungary 3-0. The Magyars were strong at that time, as they had humiliated Italy 7-1 just some weeks before. They also fielded important characters like Béla Guttmann and Ferenc Hirzer, the latter was considered one of the best foreign players in the Italian championship between the two World Wars.
Quote
Like
Share

ctr
Joined: December 16th, 2011, 8:04 am

January 31st, 2013, 5:47 am #4

So much concentrated expertise is obviously daunting to other members of
the forum! You were therefore so far the only ones who wrote a post on this
topic.



I think nfm24 is right, not to rely on the numbers of eloratings; for
historical considerations, they are not sufficiently reliable.



The teams from Asia / Oceania were
classified with far too high ‘start’ ratings (based on arbitrary decisions).
This fact made the ratings of the most Asian teams for a very long period
totally unusable. In eloratings e.g. Cameroon 2-1 Romania (1990) was a bigger upset
than North Korea 1-0 Italy (1966).




 


Other big upsets in eloratings: Turkey
4-0 China
(OL 1948, China was clearly favourite…). Compare Turkey
and South Korea
before the World Cup 1954! Turkey’s
maximum rating after World War II was 1600 (in March 1954, after they
eliminated Spain
– albeit by lot). South
Korea's ratings were ever higher until
05/07/54! Only after a 2-5 defeat against Taiwan
- according eloratings - it seemed possible, Turkey might be a little bit stronger…




 


I spoke to eyewitnesses of the match Turkey
7-0 South Korea
(WC 54). No one came even remotely on the idea that he saw two teams playing at
about the same level.




 


So, it is very difficult, to give an answer to “What are the greatest
upsets in international football based on difference in the rankings?”




 


In June last year the Oceania Nations Cup was held. In FIFA rankings as of
05/09/12 – immediately before Oceania Nations Cup started - Samoa was ranked on
156 – Tahiti 179, Vanuatu
172 and New Caledonia
155. The actual results



Samoa 1-10 Tahiti and Samoa 0-5 Vanuatu



were big upsets. New
Caledonia
9-0 Samoa was not a surprise. Such a result is quite normal
when playing 155 v 156 ... (But maybe it
was a surprise: in March 2012 Samoa was at 150, New Caledonia 162 – Since then,
both teams had no matches).





 


What often looks like an upset is sometimes only an
indication of the absurdity of a ranking system.




 


The rankings are mostly useless if you want to analyze the history of
football. I have therefore developed years ago my own computational model. In
combination with a created analysis database, it provides a lot of information.
(Also, information that would be suitable for wakkipedia.com… Who is the bogey team of Belgium – or Malawi? Who makes, even Germans doubt the Endsieg? – Luca, what
do you think…?
).




 


But I can also search for: what was the most unexpected draw? Eloratings
finds more than 300 games in which a draw was less expected than Indonesia v USSR (1956). In my model, this game
is far ahead - approximately the same as Belgium 1-1 Luxembourg (WC Qualifiers
1989), which stands at eloratings on rank 1.




 


Slightly ahead (in my model) was only Dominican
Republic 0-0 Anguilla
(WC Qualifiers 2004). Perhaps interesting: in both cases, there was a repeat of
the game two days later…



USSR won 4-0, Dominican Republic
6-0.




 


That in major tournaments (World Cup, Olympics) Japan
3-2 Sweden
(1936), was the biggest upset even says eloratings (if you compare wins
expected with actual result).




 


The four biggest upsets (WC) have already been mentioned. North Korea 1-0 Italy, USA 1-0 England. Luca, you have
added Algeria 2-1 Germany, France
1-0 Senegal




 


More, if I am asked …
Quote
Like
Share

Joined: April 7th, 2007, 12:28 pm

January 31st, 2013, 1:26 pm #5

Very interesting. When I asked the question I realised immediately that the question soon becomes "how good are the rankings".

Cape Verde advancing to the African quarterfinals has been widely considered a "shock", but apparently they ranked the highest in their group according to FIFA.

If two teams at the lower end of the rankings play each other, more "unexpected" outcomes can be found simply because rankings find it difficult to differentiate between such weak teams.



Other questions about upsets: does your system consider home advantage, or the importance of the match?

Biggest upset between 2 European teams? South American? (I can think of Colombia 5-0 Argentina and Bolivia 2-0 Brazil, but I'm sure there are others perhaps based on the "bigger" nation fielding a weakened team)



One could also take into account population size (although any match involving India or China would suddenly feature in the list).

E.g. 1980 Olympic qualifying : Indonesia 2-3 Brunei
Quote
Like
Share

Joined: July 4th, 2011, 6:46 am

January 31st, 2013, 2:03 pm #6

Are you also considering Non-FIFA teams or 'non-ranked' FIFA nations? I know that last year alone Iraqi Kurdistan had a few good results against FIFA teams, and even though they didn't win, South Sudan drawing 2-2 against Uganda (ranked 85th at the time) must be worth something. Then there were nations who were joint-last (such as Sao Tome and Principe) getting wins, even if their opponents weren't too high themselves there was probably a decent difference to consider...
Quote
Like
Share

Joined: October 31st, 2006, 5:16 pm

January 31st, 2013, 9:08 pm #7

mattsanger92 wrote:




Are you also considering Non-FIFA teams or 'non-ranked' FIFA nations? I know that last year alone Iraqi Kurdistan had a few good results against FIFA teams, and even though they didn't win, South Sudan drawing 2-2 against Uganda (ranked 85th at the time) must be worth something. Then there were nations who were joint-last (such as Sao Tome and Principe) getting wins, even if their opponents weren't too high themselves there was probably a decent difference to consider...
In any serious discussion of the original question, FIFA's rankings should not be considered at all, because they are completely rubbish.  It's impossible to tell what is a shock or not using this system (as shown by ctr's fine example of Samoa above).  These results were not a shock to anybody except FIFA and those people who believe their rankings are good (there are a lot of people who do, believe it or not!)
Quote
Like
Share

Joined: October 31st, 2006, 5:16 pm

January 31st, 2013, 9:10 pm #8

I think New Caledonia beating Fiji 11-0 in the 1969 South Pacific Games would be a shock (at least the size of the win) - if we ignore all the other factors and concentrate on the pure result.
Quote
Like
Share

Joined: April 7th, 2007, 12:28 pm

February 1st, 2013, 6:11 am #9

TheRoonBa wrote:except FIFA and those people who believe their rankings are good (there are a lot of people who do, believe it or not!)
Don't they all live in Zürich?  It seems obvious that the
badness of any ranking system is exaggerated at the bottom end,
particularly for teams which don't play often (and when they do play,
only against other teams at the bottom), so "upsets" at the bottom end
are usually artefacts.  Would Gibraltar 3-0 Faroes be an upset?  Difficult to say until Gibraltar is able to play more regularly against the same sort of opposition as the Faroes plays against.
Quote
Like
Share

Joined: October 31st, 2006, 5:16 pm

February 1st, 2013, 11:27 am #10

nfm24 wrote:
Don't they all live in Zürich?  It seems obvious that the
badness of any ranking system is exaggerated at the bottom end,
particularly for teams which don't play often (and when they do play,
only against other teams at the bottom), so "upsets" at the bottom end
are usually artefacts.  Would Gibraltar 3-0 Faroes be an upset?  Difficult to say until Gibraltar is able to play more regularly against the same sort of opposition as the Faroes plays against.
Agreed.  But it is possible to do (much) better than FIFA.  It's possible to make a half-decent ranking of all teams, even those who don't play often and those who only play against other teams at the bottom.  FIFA is just not interested in doing this while other people are.  Every attempt on the internet at a ranking system for football (by professionals or amateurs) performs better than FIFA's ranking.  FIFA's ranking is at the bottom of the rankings ranking ;-)
Quote
Like
Share

Joined: April 7th, 2007, 12:28 pm

February 1st, 2013, 2:46 pm #11

So the biggest upset in the meta-ranking would be for FIFA's ranking to outperform yours!
Quote
Like
Share

Joined: February 9th, 2011, 2:58 pm

February 1st, 2013, 2:48 pm #12

I have two other cases. I don't know if we can define them "upsets" or just "surprises". Judge yourselves.



17/05/1959 Peru-England 4-1 (Juan Seminario 10', 40' and 80', Juan Joya 69'; Jimmy Greaves 58')



What astonishes is probably the result. England had a good team (the young Bobby Charlton, Jimmy Greaves and Billy Wright), but also Peru fielded great players like Juan Joya (multi champion with Peñarol), Juan Seminario and Víctor Benítez (protagonists in Europe) and so on...



06/11/1996 Bosnia and Herzegovina-Italy 2-1 (Hasan Salihamidžić 5', Elvir Bolić 43'; Enrico Chiesa 10')



I remember it was considered a very bad defeat for Italy. It was also the last appearance of Arrigo Sacchi on the Italian bench.
Quote
Like
Share

Joined: July 4th, 2011, 6:46 am

February 2nd, 2013, 10:23 am #13

TheRoonBa wrote:
mattsanger92 wrote:




Are you also considering Non-FIFA teams or 'non-ranked' FIFA nations? I know that last year alone Iraqi Kurdistan had a few good results against FIFA teams, and even though they didn't win, South Sudan drawing 2-2 against Uganda (ranked 85th at the time) must be worth something. Then there were nations who were joint-last (such as Sao Tome and Principe) getting wins, even if their opponents weren't too high themselves there was probably a decent difference to consider...
In any serious discussion of the original question, FIFA's rankings should not be considered at all, because they are completely rubbish.  It's impossible to tell what is a shock or not using this system (as shown by ctr's fine example of Samoa above).  These results were not a shock to anybody except FIFA and those people who believe their rankings are good (there are a lot of people who do, believe it or not!)
It was just an example, the same principle can apply regardless of FIFA's list (South Sudan had never played a match before their draw with Uganda, the reigning champions of the region, as can the Gibraltar example above).
Unrelated note, the FIFA Rankings definitely aren't perfect and there are plenty of inconsistencies (mainly when tournament hosts are out of competitive action for a few years, but it can't really be helped and technically fits the formula FIFA have), but it is still a quick and easy list to check each month for a general overview, so I still consider it a trustable source for that purpose (I don't really look too much into the formulas of each ranking I come across, and most appear to be 'about right'), but again I know it has problems and I'm not looking to start an argument about this.
The main issue I have with it is when media sources or fans who don't know any better dismiss the rankings or call it unimportant whenever it doesn't fit their opinions (e.g. England have been in the Top 10 of the FIFA Rankings since November 2008 and deserve to be there with consistent results (if you're not losing many matches then it is a good run of form), but you could expect plenty of comments like 'Why are England Top 10 they're worst team in the world they always lose on penalties lol '). Then it gets followed up with 'examples' that don't have any evidence (usually ending by suggesting that the team in question 'play a match against Spain')... I'm just glad you're hating the FIFA World Rankings for the right reasons, is all I'm trying to say .
Quote
Like
Share

Joined: July 4th, 2011, 6:46 am

February 2nd, 2013, 10:58 am #14

Back on topic, what about results from some of the very smallest European nations? Faroe Islands beat Austria 1-0 in their first-ever official competitive international (played in neutral Sweden) in 1990, and Liechtenstein drew 2-2 with Portugal in 2004. And another example of a nation winning their first-ever match - Montenegro 2-1 against Hungary in 2007.
Is the size of competing nations also a point being considered, or would that just leave any bad result from China/India/USA open to questioning (although the result was impressive on-paper given the circumstances, Bahrain 10-0 Indonesia last year would not be considered for this topic given their respective form/ranking). Looking more at the population of the 'smaller' team, obviously, and is a useful factor when there are no rankings to consider...
Quote
Like
Share

Joined: April 5th, 2012, 10:54 am

February 2nd, 2013, 1:34 pm #15

Eusébio scored on his international debut for Portugal, but it didn't prevent them going down 4-2 against Luxembourg at a 1962 WC qualifier.
Quote
Like
Share

Joined: October 31st, 2006, 5:16 pm

February 2nd, 2013, 7:15 pm #16

Yes - I get your point.  I don't just bash FIFA rankings for the sake of it.   

They aren't even useful for a general overview, though.  That is my point.  And all of the problems with the FIFA ranking CAN be helped (and made better).  It's just that they are so ignorant that they won't listen to common sense.  There is just nothing good or useful about FIFA's rankings unless you want to know how NOT to rank teams (which I don't think is useful).  I've told them this and they made quite a bad attempt at defending them (basically they said it was HARD to rank teams).  It is quite hard, but it's doable - with all their money there is no excuse to have such a pathetic ranking.  None at all.


You'd be better looking at eloratings.net or any other rankings you can find - they are all just as quick and easy to access and much better (and accurate) to use as a reference.  You'll learn a lot more about relative strengths of teams than you would by looking at FIFA's rankings. 


I don't think Montenegro beating Hungary was a major shock.  They already had considerable footballing heritage as part of Yugoslavia.  Size of country is completely irrelevant.  It's more about the culture of the place.  If there is a small island with 1,000 people who have always played football, they would be favourites to beat a team from a country of 100 million who had only just started playing football. (Taking an example from another sport - Falkland Islands would be expected to have a good chance of beating Russia at cricket).


The best way to define a shock is to look at the difference in ranking points between 2 teams at any given time.  So, basically, if you beat a team, that is, say 900 points better than you, that would be a bigger shock than beating a team who is 500 better than you.  Eloratings provides such a list.  It is not perfect, but is a useful guide.


http://www.eloratings.net/Upsets.htm


Margin of victory is also considered here (and status of match).  There are no friendly matches listed.
Quote
Like
Share

Joined: December 26th, 2008, 11:44 am

February 3rd, 2013, 6:40 am #17

FIFA also changed its formula sometimes which makes their list even more debatable. It seems they want to change their formula just that way that the new Worldchampions are in the first month ater the World Cup finals the #1 at their list and the Vice-Worldchampions the #2. Where I come from they call this manipulation.
Quote
Like
Share

Joined: February 9th, 2011, 2:58 pm

February 5th, 2013, 5:15 am #18

By coincidence, I've discovered a funny case - not for the Mexicans, of course... 
14/12/1973 Trinidad and Tobago-Mexico 4-0 (Everald Cummings 11' and 39', Steve David 52', Warren Archibald 62')
This match was valid for the 1973 Concacaf Championship, that is to say the 1974 World Cup Qualifiers, and Trinidad and Tobago astonishingly liquidated Mexico, eliminating them from the World Cup. The newspaper El Informador defined this defeat as "a tragedy", while, according to several Mexican sources, this débacle was caused by witcheries and voodoo, as the tournament was held in Haiti...

Quote
Like
Share

ctr
Joined: December 16th, 2011, 8:04 am

February 5th, 2013, 5:55 pm #19

mattsanger92 wrote:




Are you also considering Non-FIFA teams or 'non-ranked' FIFA nations? I know that last year alone Iraqi Kurdistan had a few good results against FIFA teams, and even though they didn't win, South Sudan drawing 2-2 against Uganda (ranked 85th at the time) must be worth something. Then there were nations who were joint-last (such as Sao Tome and Principe) getting wins, even if their opponents weren't too high themselves there was probably a decent difference to consider...

Yes, my
model considers many non‑FIFA members. For example, Zanzibar, which participates in the CECAFA
Cup and there periodically plays against FIFA members. Zanzibar
on the other hand has played a few times against Northern
Cyprus. Therefore, I also look at their results. And then I look
to a ranking and ask: is it reasonable what the ranking says.

FIFA ranked South Sudan since August, Roon Ba after the match against Uganda, which
you pointed. My model considers only the outcome of a match (Wins, Draws, Losses).
Meanwhile, I have four results (data). In order to make a meaningful assessment
with statistical significance, I need nine. Therefore, these four results are
considered in the model and processed internally. But South
Sudan will add in the rankings when they will have played nine
games.




  
Quote
Like
Share

ctr
Joined: December 16th, 2011, 8:04 am

February 5th, 2013, 5:58 pm #20

Luca wrote:I have two other cases. I don't know if we can define them "upsets" or just "surprises". Judge yourselves.



17/05/1959 Peru-England 4-1 (Juan Seminario 10', 40' and 80', Juan Joya 69'; Jimmy Greaves 58')



What astonishes is probably the result. England had a good team (the young Bobby Charlton, Jimmy Greaves and Billy Wright), but also Peru fielded great players like Juan Joya (multi champion with Peñarol), Juan Seminario and Víctor Benítez (protagonists in Europe) and so on...



06/11/1996 Bosnia and Herzegovina-Italy 2-1 (Hasan Salihamidžić 5', Elvir Bolić 43'; Enrico Chiesa 10')



I remember it was considered a very bad defeat for Italy. It was also the last appearance of Arrigo Sacchi on the Italian bench.
Peru v England, 1959



I think the
Englishmen were shocked ..., the Scots not even surprised. Seriously, including
the (normal) home advantage the chances were slightly better for England. In
intercontinental matches the home advantage has a (much) stronger impact. So,
the defeat was not a surprise – maybe the size.




 


Bosnia and Herzegovina v Italy,
1996




I also remember
this game, although I'm at an age, the short‑term memory is beginning to
decline. I fear the joke is understood
only by residents of retirement homes with a sense of humour
. Based on win expectancy Italy
was the clear favorite.



One note: a win expectancy e.g. 80% does not necessarily say that a team will
win in 80% of cases. It must be considered the counter probability (the
difference is 60 percentage points).

Correct solutions are thus: 8 wins ‑ 2 losses, but also 7 wins ‑ 2 Draw ‑ 1
loss and 6 W ‑ 4 D ‑ 0 L. Win expectancy e.g. 80% says, in 10 matches the
difference of wins and losses will be 6 (and nothing else).



Italy
has traditionally had a high draw tendency. So I think the fans would accept if
their team will win 4 times (plus 6 draws). Nine wins (best with 9 scored goals) would make them happy ‑ and only one
defeat angry ... even when the overall balance would be better. Therefore, it
was the least surprising: It was also the
last appearance of Arrigo Sacchi on the Italian bench.



One other note: your long list of astonishing cases began with Luxembourg v Netherlands. This result is just an upset from today's perspective. In the early 1960s, Luxembourg had
almost reached the maximum of skill level, what is possible due their
infrastructure. See also the post of Kaizeler
(Luxembourg 4‑2 Portugal. 1962)

Luxembourg
was then scarcely weaker, but the level of professionalism in other countries
increased steadily. Therefore, Luxembourg
became less and less competitive. A few years ago they had some professionals
and the results were better again (temporarily). Luxembourg will never have a
professional league ‑ how could they? If fate but it adds, they might have from
time to time some professionals in other leagues...




  
Quote
Like
Share

ctr
Joined: December 16th, 2011, 8:04 am

February 5th, 2013, 6:02 pm #21

nfm24 wrote:Very interesting. When I asked the question I realised immediately that the question soon becomes "how good are the rankings".

Cape Verde advancing to the African quarterfinals has been widely considered a "shock", but apparently they ranked the highest in their group according to FIFA.

If two teams at the lower end of the rankings play each other, more "unexpected" outcomes can be found simply because rankings find it difficult to differentiate between such weak teams.



Other questions about upsets: does your system consider home advantage, or the importance of the match?

Biggest upset between 2 European teams? South American? (I can think of Colombia 5-0 Argentina and Bolivia 2-0 Brazil, but I'm sure there are others perhaps based on the "bigger" nation fielding a weakened team)



One could also take into account population size (although any match involving India or China would suddenly feature in the list).

E.g. 1980 Olympic qualifying : Indonesia 2-3 Brunei
First to the other questions:
my system, which is based on normal distribution function, consider home
advantage, the importance of a match, some ‘fitting’ functions, number of games
in a tournament and some other things. It does not include any
kind of ‘Bayesian’ and it does not
consider the size of a match result. The system is designed to be very
flexible. Therefore, it could be extended to other factors. One example: many
people (mostly older) believe that in previous years the international matches (friendlies)
have been "more important". The truth is that the informative value
of a result in a friendly match was greater before 1970 than today. (For historical considerations, this is
important).
But not I decide whether and (if so) how much higher was this
value, but the model. And that is the basic principle of the whole system. I personally
do not make any pre‑determinations. I define only the goal. My definition is:
the assessment of the skill level of a team at some point (time, moment), must
vary as small as possible from the real measurement (= result of a match). The
accuracy is checked a) for all games
and b) for all World Cup matches. All
the parameters (and their size) are then calculated iteratively. The structure
of the system and the parameters are optimized when the average deviation of a and b is the smallest.

This all sounds very weird. But it is a very good description of reality. At
the end of each season (e.g. League One, Season 2011‑12), all teams will be
classified according to their average performance. (Let the bonus point for a
victory disregarded; in practice, the impact is minimally). The results of all teams vary in the course of a
season, but they always vary around their average. This can be understood
perhaps, if my horrible English does not make it impossible.

A normal table with the final standings of a league is convertible in my system
‑ the reverse is also possible. It is then simply a different form of
presentation.



The reason Cape Verde
advancing has been widely considered a "shock": The rankings do not
allow really a assessment by feeling. The group was so tight that every outcome
was no real surprise. In order to have a scale of gaps, the following real‑life
example: at the end of last season, Birmingham
City, Blackpool (host), Middlesborough
and Burnley would have played a round‑robin‑tournament.
The gaps in the group with Cape
Verde were comparable closely. (Compare
final standings Division 2, 2011‑12). In this group, FIFA was not entirely
wrong. But ‑ what about Algeria?...
When I look at all the data, The Roon Ba is right. Stevan, a member in this
forum, asked me in November 2010, to check the accuracy of all the rankings of his
Aggregated rankings. Therefore I possess data of more than 2,300 games.

AQB, Chance de Gol, eloratings, Rankfootball, Soccer Power Index (ESPN), The
Aggregated List itself, The Roon Ba and even the rankings of Voros McCracken
(latest update 10/07/2010) – all of them are much more accurate than FIFA. In the area of the top 100 (FIFA), in
competition matches, the FIFA rankings is not quite so bad. But ‑ only the
distance is smaller; it is also still less accurate than all the others. Games
of two teams at the lower end of the rankings are perhaps additionally a
special problem … But this issue is not the cause, that FIFA ranking is
inaccurate.



All of the other rankings are better – but all of them a (very) different.



Biggest upsets South America

In addition to those already mentioned:

Uruguay 0‑3 Venezuela (WC Qualifiers 2004)

Bolivia 6‑1 Argentina (WC Qualifiers 2009)

Brazil 1‑3 Peru (Copa América 1975)



Biggest
upset between 2 European teams

If the scale is the most unexpected victory,
eloratings says:

Switzerland 1‑2 Luxembourg (WC Qualifiers 2008)



That is maybe a surprise. But – it was the 54th WC Qualifier away game of Luxembourg.
Record: 1 draw, 52 defeats. Head of the list of opponents they had their best
chances: Liechtenstein (in 2005) 0‑3; Faroe
Island (2001) 0‑1; Cyprus (1997) 0‑2; Norway
(1965) 2‑4; Finland
(1976) 1‑7.



So I agree, when eloratings expected a chance of winning less 1%. And don’t
forget: after winning the qualifying group – Switzerland
defeated Spain
1‑0 (WC 2010)!

When searching for upsets we should not forget: there are actually none at all!
They are all in the field of statistical range.



Final note: I personally do not like rankings.
Quote
Like
Share

SDb
Joined: January 24th, 2013, 5:10 am

February 5th, 2013, 7:30 pm #22

In my personal opinion, the rankings are just a way to show the current form for a specific team. A lot prefer eloratings as they go back further and take into effect more variables. I prefer fifa ranking as it shows more recent strength of a team. Still, it's not a table of who's better than others etc.

Back to topic, I found it more shocking to see Sweden draw with Germany last year after being four goals down. Watching the game no one expected that at all! Based on rankings however the 1993 qualifier Turkey vs Norway comes to my mind. Turkey were ranked 63 spots below Norway (3rd) but went on to win the match.
www.soccer-db.info - football internationals
Quote
Like
Share

Joined: April 7th, 2007, 12:28 pm

February 8th, 2013, 2:28 am #23

ctr wrote:Biggest
upset between 2 European teams

If the scale is the most unexpected victory,
eloratings says:

Switzerland 1‑2 Luxembourg (WC Qualifiers 2008)

So I agree, when eloratings expected a chance of winning less 1%. And don’t
forget: after winning the qualifying group – Switzerland
defeated Spain
1‑0 (WC 2010)!
I forgot that one.  In that case both teams would have been well ranked (in almost any system) because they play regular matches vs the same sort of opponents.   Compare to North Korea 1-0 Italy : whatever ranking N.Korea had initially was based on very very few games.
Similarly USA 1-0 England.
Quote
Like
Share

ctr
Joined: December 16th, 2011, 8:04 am

February 9th, 2013, 6:12 am #24

nfm24 wrote:I forgot that one.  In that case both teams would have been well ranked (in almost any system) because they play regular matches vs the same sort of opponents.   Compare to North Korea 1-0 Italy : whatever ranking N.Korea had initially was based on very very few games.
Similarly USA 1-0 England.

Of course, you
are right! A small base of data, such as in North Korea poses the risk of high
margin of error. But I use additional other information
that greatly limit the error range.

I transform time-independent data in a time-dependent system. Time-independent
data are funny. (Where else you can see a ranking: 11 Hungary, 15 Scotland,
16 Austria?)
From all the available results the system calculates an all-time ranking (=
average). The system ‘knows’ 245 teams: ‘historical’ teams like USSR, Yugoslavia,
or e.g. Ireland
(British) are included. A crosstab from a database provides the data, who
played against whom ... how often ...

If one is able to solve an equation with 245 unknowns, then you can measure the
performance gaps accurately. Equations with several unknowns can be solved by
iteration. Time-independent data, of course, have a limited informative value. As
a rough guide, they are still useful.

A practical example of how the system works: based on all-time average would
have been expected that Scotland
achieved in the first ten games mathematically 4.865 wins. The system compares
always ‘expectation’ and actual result; On the basis of the difference is
calculated a new rating score. In the case of Scotland: after ten games are all their
ratings time-dependent (since 1878).

The system processes more than 5,000 intercontinental games. A much
differentiated view on a very large base of data is possible. The skill levels
of teams from different continents are comparable! At this point I stop for
now. I would not want anyone writes: Football is for everyone - except
mathematicians, physicists and intellectuals.



  
Quote
Like
Share

Joined: October 31st, 2006, 5:16 pm

February 9th, 2013, 4:29 pm #25

SDb wrote:In my personal opinion, the rankings are just a way to show the current form for a specific team. A lot prefer eloratings as they go back further and take into effect more variables. I prefer fifa ranking as it shows more recent strength of a team. Still, it's not a table of who's better than others etc.
.
As FIFA rankings are worse than all other rankings (for defining who is better than others, and also for showing current form), I am not sure why anyone prefers them.  Are Algeria really the 2nd best in Africa (in strength or on current form?)  I doubt it.  ELO ratings is better than FIFA's rankings in all aspects - strength of teams and also current form. It really is a very bad attempt by FIFA to rank strength of teams or current form.  I think people just prefer it "because FIFA did it", because no rational thought can come to the conclusion that FIFA's ranking is better than other rankings (for any purpose)..
Last edited by TheRoonBa on February 9th, 2013, 4:31 pm, edited 1 time in total.
Quote
Like
Share

Joined: October 31st, 2006, 5:16 pm

February 9th, 2013, 5:15 pm #26

Rankings are a tool to show, in a more accurate way, what people think they know. Most people only keep track of the big teams, so they might know things like "Spain, Brazil, England and Germany are good" and "San Marino, Liechtenstein, Malta and Luxembourg are not good". Many people have no idea how good certain teams are (for example, Bermuda, Afghanistan, Samoa, Bolivia), and so a ranking is a much better system than just relying on what people think, because it includes ALL matches and ALL teams and allows comparisons to be made between teams who might never play each other. Many people who criticise ranking systems know almost nothing about half of the teams in the ranking, and probably do not know the last 9 or 10 matches of each team - mostly people have memories of teams being good in the past and expect them to always be at the same level. These people are not really in a good position to give constructive criticism on ranking systems.
Quote
Like
Share

Joined: November 3rd, 2006, 11:49 am

February 10th, 2013, 2:14 am #27

I critcise ranking systems and I know very much about Samoa, Bermuda and Afghanistan.....I do not need rankings to know which are strong teams and average teams.....and I really know when a result is an upset.....
Quote
Like
Share

Joined: October 31st, 2006, 5:16 pm

February 10th, 2013, 6:49 pm #28

pieter wrote:I critcise ranking systems and I know very much about Samoa, Bermuda and Afghanistan.....I do not need rankings to know which are strong teams and average teams.....and I really know when a result is an upset.....
Yes - that may be true.  But many people criticise and DO NOT know. These are the people that I am talking about.  These people need rankings, because it is quicker than explaining to them the entire history of football and all the results ever played.  It's a quick and easy way to show that in one list.  Of course it is not perfect, but it does a much better job of ordering teams than humans can do with their own minds.
Quote
Like
Share

ctr
Joined: December 16th, 2011, 8:04 am

February 10th, 2013, 7:47 pm #29

I do not
criticize rankings! But I do not like them particularly. Why should I criticize
rankings?

With pieter’s arguments I could say: I oppose, teams will be classified at the
end of the season according to their average performance. Yesterday e.g. lost
in the German Bundesliga Borussia Dortmund at home against Hamburg with 1-4. I look at the current
standings and then I felt confirmed. Dortmund is
ranked on 2, Hamburg
on 5. Rankings are
bullshit! So it would be much better to abolish all the championships, and to
replace them with the expertise… (I
did not find the smiley, which symbolizes a mildly ironic smile)
.

Why I don’t like rankings anyway? The normal viewer - "normal" is not at all meant to be derogatory
- has no sense of proportion of what is reasonable and appropriate. Any idiot
can perform any calculations and then declare that this is a world ranking. Only
with an objective review may be clarified whether an idiot was at work … or
someone, who has found meaningful measurement criteria (e.g. Mark Cruickshank).

But - how to make this the normal viewer who looks always only a short view on
numbers? One ranking system used that formula, another one another formula. A
viewer is overstrained, to identify which formulas are useful and reasonable?

The first view of an observer will always be on the ranking position. Instinctively he wants to see
a change, if a serious event occurred (Dortmund
1-4 Hamburg
…). Is it reasonable to conclude from the result that Hamburg
is the best club team in the world at present - or Dortmund nationally level 2? And therefore it
is not surprising that people look at the FIFA rankings. Those, the jumps in
the ranking are too low, will maybe search in internet for rsoccer.com. Based
on data from Alexa.com, this is the most widely used ranking (worldwide) –
behind FIFA rankings. It does not bother the users of the site when there Tahiti
and New Caledonia are ahead The Netherlands ...The site allegedly represents the
current form of teams.



Upon
request by Stevan since 17/11/10 I
saved all the world rankings - always parallel with the release of FIFA rankings. Instead of statistical indexes here are some numbers ‘you can touch’.

Compared are games on neutral ground. (For such games I do not need take into
account a home field advantage). And now I see the games in which the FIFA
rankings and the other ranking system had different ‘opinions’. Who was right? A
ranking that reflects better the current strength / form, will more often
correctly assess the outcomes of games in next month.

The results (from the
perspective of the FIFA rankings): against RSoccer 47-38. FIFA was more right!
It is possible, therefore, to be worse than the FIFA rankings. Other
comparisons (results): 33‑46, 36‑50, 31‑45, 35‑50, 32‑48, 38‑60, 29‑52. I call
here not the respective other rankings. A dispute as to whether this or that game can
be taken into account is unnecessary. The results look like handball results (after
two legs). A press secretary of a club, who still claim to be the best (in face
of such unambiguous defeats), would even picked up in Zurich as a helpless person...

The FIFA ranking
succeeds precisely not to depict the
current strength / form of teams. But - the use of the FIFA rankings should not be criticized too harshly. A clever
and knowledgeable football expert does not understand why eloratings considers
in reality fewer games than FIFA (in the very most cases). I think even the
designer of eloratings didn't know that Arpad Elo's system contains a very
powerful ‘cybernetic component’.

The FIFA rankings can be forgotten. But even with statistically ‘high-end
systems’, there are always serious problems. How a ranking would look like (and
develop), if it reflects the reality? Like the reality - extremely boring.
That’s one of the reasons many of us watch football… From ranking positions
often is not recognizable the actual size of skill level gap of two teams.






I have checked
the Roon Ba ranking on a broad basis (the published versions until 07/10/12).
The quality is outstanding. Within the regions of the ranking (top 100 teams,
non-FIFA members, etc.), there are small differences in quality ("makes
sense" to "excellent"). So I'm looking forward to the new
version. It will certainly have a valuable piece of information, but
unfortunately only a very few will be interested.
Last edited by ctr on February 11th, 2013, 7:40 pm, edited 1 time in total.
Quote
Like
Share

Joined: December 26th, 2008, 11:44 am

February 16th, 2013, 7:34 am #30

Also a kind of an upsted. In 1963 Holland beat Worldchampions Brazil 1-0, within half a year later they were eliminated by Luxembourg in a round of 16 (1-1 and 2-1 defeat; both matches played at home!).
Quote
Like
Share

ctr
Joined: December 16th, 2011, 8:04 am

February 17th, 2013, 8:06 pm #31

Fast Midfielder wrote:Also a kind of an upsted. In 1963 Holland beat Worldchampions Brazil 1-0, within half a year later they were eliminated by Luxembourg in a round of 16 (1-1 and 2-1 defeat; both matches played at home!).
  

You are right!
The contrast between the results is tremendous!



And just to
complement: the gaps of the teams at that time were different than nowadays. Holland is one of the
great football-nations in history; added, with a very long tradition. But -
especially in the 1950s, they lost touch with the top teams. Only on the
professionalization of the clubs in the very late 1960s, they could also build
a strong NT. The first exclamation mark of new Dutch football was probably the
winning of the European Champions' Cup by Feijenoord, in 1970. (I think people
living in Vlaardingen know it!) 1963, with semi-professionals and
amateur structures, Holland
was at best average. Luxembourg
had probably its best team of all time. Brazil
was on tour in Europe (and the Middle East) and had some very poor results
(e.g. in Belgium).
After the 1962 World Cup they began to build a new team.



Therefore: any
result - taken alone - was at most a surprise. The overall consideration of the
results is quite an oddity ... or what you call “a kind of upset”.
Quote
Like
Share

Joined: December 26th, 2008, 11:44 am

February 18th, 2013, 4:48 am #32

ctr wrote:Holland is one of the
great football-nations in history; added, with a very long tradition. But -
especially in the 1950s, they lost touch with the top teams. Only on the
professionalization of the clubs in the very late 1960s, they could also build
a strong NT. The first exclamation mark of new Dutch football was probably the
winning of the European Champions' Cup by Feijenoord, in 1970.
Dutch NT results were very poor in the first half of the 1950s as the best Dutch players went abroad to play as a professional and became suspended for NT. In the 1960s most of the time Ajax did not released its players for Dutch NT, sometimes Feyenoord did the same. In 1962 even all Feyenoord (Dutch League champions of the 1960-1961 and 1961-1962 season) players were suspended for Dutch NT. Results of Dutch NT became better since 1971 when everyone realized the importance of a strong NT. Finally.
The first exclamation mark of new Dutch football was in 1966 when Ajax beat Liverpool 5-1 in the European Champions Cup tournament and drew 2-2 in the return (after being 2-0 up).
Quote
Like
Share

ctr
Joined: December 16th, 2011, 8:04 am

February 18th, 2013, 12:36 pm #33

Thank you
for additional information. I knew too little about the reasons why the Dutch
NT had especially in the early 1950s
very poor results. (Absolute low point: 1954).

I also remember Ajax 5-1 Liverpool. Of course, it was a exclamation mark in the
sense of “Yes, we can...”. In the next round Ajax was eliminated by Dukla Prague, a
venerable club, but never really had a European top team. EC1, 1967: Ajax v Real Madrid was
very tight. In 1968-69 Ajax eliminated Nuremberg (German
champion). The 4-0 in the second leg was widely seen as an upset (If you would
say, you were not surprised about "typical Moffen's ignorance",  I
would accept it). AFC Ajax even reached the final ‘69, where it then but had no
chance against AC Milan. What the world thought (…luck of the draw, …Trnava in
the semifinals, …which won its part in 1/4-final against Reipas Lahti), was
irrelevant.

All great achievements of Ajax
were still not sufficient to change completely the general mood in the Dutch society. First after
Feyenoord's triumph in the European Cup, the traditional ideas were no longer
tenable. (And that’s I called ‘exclamation mark’). Traditionally, professional
football was considered disreputable. (A decent man goes regularly to work.
Professionals, however, are like gamblers). Such views, well pronounced e.g. in
Sweden
similar, were the fundamentals of a very special amateur ethos. Such views were
the views of the ‘elites’; for a long time they have been generally accepted. (No
country wants to be represented by ‘seedy’ gamblers).

Given the fact that Holland
had in 1970 two 'giants' in
international club football, a new question arose: what if? The question found an answer: Netherland’s
NT was quickly a major power! The rise of the Dutch NT was not heralded by continuously
good results of the NT (the rise was too fast …); ‘infrastructural changes’ were
crucial.

I think we have no real dissent. Different opinions would be also not bad. Only
in a dialog I can get to know new (additional) aspects.
Quote
Like
Share

Joined: February 9th, 2011, 2:58 pm

February 18th, 2013, 2:56 pm #34

However, if we look at the players fielded by Netherlands during the two games against Luxembourg, we can say that the level was very good.


Eddy Pieters Graafland, Guus Haak, Coen Moulijn and Cor Veldhoen became European champions with Feyenoord in 1970, while Sjaak Swart and Piet Keizer conquered the European Cup in 1971, 1972 and 1973 with Ajax.
Quote
Like
Share

Joined: December 26th, 2008, 11:44 am

February 18th, 2013, 3:49 pm #35

In the 1962-1969 era the level of the Dutch players was good, however there was room to improve their mental qualities if they were playing for the NT. It is known both Ajax and Feyenoord did not always released all their best players, but if they released them the clubs asked these player not to give 100%. So Dutch NT suffered more embarrashing defeats. They had quite a lot of problems with teams which showed a fighting spirit. This explains defeats against Norway and Denmark (2) which was regarded as very embarrashing at the time. East Eeuropean teams were their bogey teams. Their poor records against Belgium (1 win, 2 draws, 6 defeats) however is not surprising as at the time Belgian NT was stronger.
Quote
Like
Share

Joined: August 4th, 2012, 4:06 pm

March 22nd, 2013, 7:45 am #36

TheRoonBa wrote:
Size of country is completely irrelevant.  It's more about the culture of the place.  If there is a small island with 1,000 people who have always played football, they would be favourites to beat a team from a country of 100 million who had only just started playing football. (Taking an example from another sport - Falkland Islands would be expected to have a good chance of beating Russia at cricket).
I'm going to have to respectfully but aggressively disagree with you on this point. Part of what you said is true, that culture is the most important factor, but population can be a huge factor also. The statement "size of country is completely irrelevant" is what I'd like to counter. There are small countries like Croatia and Serbia who are very athletic, but will never perform as well as many larger countries throughout a wider spectrum of sports plain simply because they do not have enough population to do so. The reason is is because for one, genetically, only a certain percentage of the population is going to be able to perform athletically. Secondly, generally the larger a country is the more people it can potentially have to contribute to leisurely programs like sports for both direct and indirect economic reasons. Lastly, the chance of trends of playing new sports is less with smaller countries, because there is less population to possibly take up that sport. So can we say that a team with a smaller population who beats a team with a much larger population creates an upset? not necessarily, but very possibly. It all depends on the specific situation.
Of course someone could counter my argument and use exploited countries with large populations like China, India, Nigeria, and Iran as evidence. But this will be getting deeper into the topic of socioeconomics which I won't tackle in this post, but if anyone is interested to hear I can tell them.
But my point is simply put here: Suppose there are 2 countries that both have an equal athletic culture (meaning their programs, dedication, etc) are exactly the same... also, both of these countries have the same economic equality globally... but the only difference is one of the countries has 3,000,000 people and the other has 300,000,000 people. The larger country is going to dominate in more sports than the smaller country without question.
And to add to this, let's change it around a bit. In this scenario lets suppose everything in the above scenario is the same except that the country with 3 million people are more dedicated athletes and have better athletic programs than the country with 300 million. Even with this the country with 300 million will usually perform better in a wider net of sports simply because it is easier for them to. If you need a real world comparison to this use Uruguay and the United States. I think it is pretty safe to say that Uruguay is a more athletic nation than the United States in their programs and culture, but there is just no way they can play as good as the United States in as many sports simply because of population, and also, the economic effects that come with having a smaller population.
If you look into my stats using a combination of all contact and collision sports, the United States is in second place with the most brute power of all countries (with Spain being in first place). Spain has a population of 47 million and the United States has a population of  314 million. THIS IS A VERY RARE HAPPENING! I have some socioeconomic theories of how this happened, but it is not something you will see in history very much among countries with similar economic development. Most countries with 40 to 50 million people are far far below this ranking.
On a final note population advantage does have a limit, and the higher the population goes the less and less effect it has. So the United States would only perform very slightly better if they have 1 billion people compared to what they have now if their economic and cultural attributes stayed the same.
Last edited by abramjones on March 22nd, 2013, 7:56 am, edited 2 times in total.
Quote
Like
Share

Joined: October 31st, 2006, 5:16 pm

March 22nd, 2013, 11:39 am #37

Of course, the points you make are valid. But I was only talking about football.
Quote
Like
Share

Joined: February 9th, 2011, 2:58 pm

June 11th, 2013, 4:43 pm #38

11/06/2013 Italy-Haiti 2-2!
Quote
Like
Share

Joined: October 31st, 2006, 5:16 pm

June 11th, 2013, 5:30 pm #39

2 good results for Haiti within a few days - most teams wouldn't be capable of losing 1-2 to Spain and drawing 2-2 with Italy. I didn't expect Haiti would get such good results, especially as they'd lost their last 5 games to Cuba, Dominican Republic, Oman, Chile and Bolivia.
Quote
Like
Share

Joined: November 3rd, 2006, 11:49 am

June 12th, 2013, 1:21 am #40

I wnet to sleep when they were 2-0 down, I regret  I did not watch the rest of the game
Quote
Like
Share

Joined: February 9th, 2011, 2:58 pm

June 12th, 2013, 5:26 am #41

And just a handful of days ago, another upset: Armenia-Malta 0-1.
Quote
Like
Share

Joined: October 31st, 2006, 5:16 pm

June 12th, 2013, 6:03 am #42

I think Denmark 0-4 Armenia last night beats this one.
Quote
Like
Share

Joined: April 7th, 2007, 12:28 pm

June 12th, 2013, 7:53 am #43

I watched some of this game, probably the most unimaginative performance I've seen from a "favourite", they had no tactics other than long balls up to the 2 big strikers. Armenia easily passed through their defence and probably could have scored more goals if they needed.
Quote
Like
Share

Joined: April 7th, 2007, 12:28 pm

June 13th, 2013, 10:49 am #44

Also Croatia 0-1 Scotland was a severe upset at least in the FIFA rankings which we all respect and aspire towards.
Quote
Like
Share

Joined: February 9th, 2011, 2:58 pm

October 23rd, 2013, 5:11 am #45

I admit I'm not too skilled about rankings, but I suppose this result represented a clamorous upset at that time:

31/03/1963 Bolivia-Brazil 5-4

Brazil were "bi-campeões do mundo" at that time, even if they didn't field their best guys in the game, while Bolivia didn't have an encouraging tradition before that moment. Anyway, that year, Bolivia conquered the South American Championship. 
Quote
Like
Share

ctr
Joined: December 16th, 2011, 8:04 am

October 27th, 2013, 7:31 pm #46

Luca wrote:
I admit I'm not too skilled about rankings, but I suppose this result represented a clamorous upset at that time:

31/03/1963 Bolivia-Brazil 5-4

Brazil were "bi-campeões do mundo" at that time, even if they didn't field their best guys in the game, while Bolivia didn't have an encouraging tradition before that moment. Anyway, that year, Bolivia conquered the South American Championship. 
In doubt the ability of a good
judgment (based on expertise) is much more important than profound knowledge of
rankings and/or rating systems. Your posts are always very impressive in terms
of expertise (and they also do not stand in contrast to proper rating
systems).
1962/63: this is exactly the time when I (really) began to be interested in
football (during this time I was infected and since then I’m terminally ill
...).
So, I can confirm: this result represented a
clamorous upset at that time.
Some relativisations
you have already made.
Some other ones: South American Championships (or Copa América) are not as
important as many media and football fans believe… The factor altitude was at
the beginning of the 1960s, even more important than it is today (the game was
played in Cochabamba, >2,500 m). ‘Medical advances’ brought helpful insights
and ‘support’ for the lowlanders

Just to complement: 17/04/49 Bolivia 3-2 Uruguay
(in Brazil)
was another upset, maybe even a bigger one. Compare it with other results Uruguay v Bolivia
(all games were played on neutral ground): 1945 2-0, 1946 5-0, 1947 3-0 and one
year later, 1950, Uruguay
won in Brazil
(World Cup) 8-0.
  
Quote
Like
Share

Joined: February 9th, 2011, 2:58 pm

October 29th, 2013, 5:59 am #47

Many thanks for your remarks. In the past, the South American
Championship used to be a bizarre competition, where many variables
intervened to upset everything.Height is surely one factor; another
one is the fact that the most important countries often renounced to
call up their best players, because of many reasons (strikes, contrasts
between clubs and FA, lack of motivations and so on...). Last but not
least, referees sometimes had a finger in the pie...
Quote
Like
Share

Joined: April 7th, 2007, 12:28 pm

November 6th, 2013, 1:18 am #48

Some big potential upsets could potentially come over the next weeks. Not in the same league as some of the greatest upsets mentioned here, but still reasonable.

Iceland vs Croatia (2 legs)
Jordan vs Uruguay (2 legs)
Ethiopia needs to win in Nigeria
Egypt needs to win 5-0 vs Ghana.... only the referee can save them.

On recent form I wouldn't consider NZ eliminating Mexico to be a big upset. Perhaps Mexico is still favourite, but not by so much.
Quote
Like
Share

Joined: October 31st, 2006, 5:16 pm

November 6th, 2013, 9:43 am #49

Equatorial Guinea v Spain on 16th November could provide an upset...
Quote
Like
Share

Joined: April 7th, 2007, 12:28 pm

November 6th, 2013, 5:26 pm #50

Also Italy 2-2 Armenia last month would ordinarily be considered a massive shock.
Quote
Like
Share