|
This forum is read-only now. Please use Forum 2 for new posts
xml |
No replies possible in the archive |
Author: rbr
Date: 25-06-2006, 13:26
| given the ridiculous rankings some countries have ,mexico 4th america 5th , yet argentina 9th is it not about time fifa overhauled the whole points system and brought one in place almost like the golden boot system where points are allocated depending on dificulty ie in italy they get 2 points per goal where as in scotland you get 1.5 , this could and should be brought into international football as the seedings need seriously over hauled . Does anyone agree |
Author: panda
Date: 25-06-2006, 13:32
| yes, for sure something is wrong- a discussion on this is running at the 'continental coefficients for WC' a couple of threads down from here.
But I don't know how to do it, when inter-continental matches are not so frequent or meaningful. Maybe there should be a continetn-coefficient the way uefa has country-coefficient. You can add a % of your continent coefficient to your country score.
But I see political protests. |
Author: badgerboy
Date: 25-06-2006, 13:42
| rbr
Also here is a link to the official FIFA website which details how they calculate the rankings. Or at least how they used to calculate them as this is apparently changing!
It shows that they did already try to take into account the different strengths of the regions, importance of the matches etc. It just didn't quite work.
It will be interesting to see what the new calculation method is exactly. One thing they should do is only count a certain number of matches in a year (a bit like Tennis used to do with the ATP rankings) so that teams like Mexico - who simply play lots of internationals aren't artificially bumped. |
Author: ignjat63
Date: 25-06-2006, 13:43
| panda, your suggestion with continental coeff is the only one that makes sense, I think. Continental rankings are fairly easy to make and more less fair. Once continental rankings are known you apply "weighing" coefficiants and lump all the ranking lists into one.
The only questions are those concerning weighing coefficients. For example this question: When calculating weighing coefficients would we meet with the same problems comparing continents as we meet now. Also the weighing coefficients would be based on average principle per continent. So it should not be the same for Brasil and Argentina as for the weakest SA countries. Especially in Africa and Asia which have a lot of countries. These are laymen questions as I am not statistician. But I see no better way of makind one world wide list. |
Author: Max_F
Date: 25-06-2006, 15:28
| IMHO fifa ranking have not bad (but not very good) idea but bad realization. Good is that received points depends on "strength" of the opponent. Bad is that difference isn't very essential. So, say, USA can win almost all their matches in CONCACAF and get 15 points for each without problems, but in Europe matches much more difficult and brings in average only 20 points for win (10 for draw, 0 for loss). |
Author: Addy_G
Date: 25-06-2006, 18:20
Edited by: Addy_G at: 25-06-2006, 18:21 | The World Football Elo Ratings can prove more accurate than any other ratings/rankins.But it`s dozens of times harder to make the calculations. Here`s a sneak preview : Team team Team rating rank 1 Brazil 2062 2 Netherlands 1986 2 Spain 1986 4 Argentina 1975 5 Italy 1956 6 England 1955 7 Portugal 1947 8 Germany 1946 9 France 1919 10 Denmark 1866 11 Czechia 1851 12 Switzerland 1831 13 Sweden 1823 14 Uruguay 1818 15 Romania 1816 16 Australia 1810 17 Mexico 1797 18 Ireland 1794 19 Croatia 1790 20 Ukraine 1784 You can access the full listing {a href="http://www.eloratings.net/">here{/a>. |
Author: ferdi
Date: 25-06-2006, 19:46
Edited by: ferdi at: 25-06-2006, 23:22 | {i>The World Football Elo Ratings can prove more accurate than any other ratings/rankins.But it`s dozens of times harder to make the calculations.{/i>
I don't know what you mean by "accurate" in this context. The main feature of the Elo Ratings is that they are extremely simple to calculate and to understand.
You do not need weightening coeffs for continents. You just need to know the supposed strength of both opponents before the match, i. e. their ranking points before the match, and from that you calculate their ranking points after the match. There is some weightening introduced so that World Cup matches have heighest weight and friendlies have lowest weight, and high victories lead to a higher exchange of points, but all very straightforward and clear.
The disadvantage of this list is that everybody can calculate it from the results and publish it, so that FIFA cannot hold a copyright on this list. |
Author: rbr
Date: 25-06-2006, 20:28
| thanks to all who replied , now how could that be introduced successfully to our uefa tournaments . I know this is an old topic but the arguement is the same re- the number of coeficient points for uefa cup and c/l . If you look at berts table i think that the number of top point winning clubs from the uefa cup is disapropriate to the competition i.e points are too easily won in the uefa cup ???? |
Author: panda
Date: 25-06-2006, 23:48
| @rbr
Yes, this argument runs on a lot opf threads - but the main defence of the existing system is that it is not hard to calculate and understand.
@ferdi
Yes, you are right (as usual) - I forgot the Elo system as a possible means of calculation. However, if I take the analogy of chess, where there are immense numbers of matches, as the obvious use of the Elo system, is it possible that this very large statistical sample gives a better result than for international football? |
Author: Max_F
Date: 25-06-2006, 23:55
| ELO seems to me slightly unbalanced because 1 game can move teams by 5-10 places even in top 20! |
Author: rbr
Date: 26-06-2006, 00:07
| cheers panda , but surely the main reason for keeping the existing system cannot be based on the fact that it is easy to calculate and understand . For the last three years fans like yourself amongst many others have very intelligently debated and even come up with ways of vastly improving the current set up which can only be of benifit to clubs and fans alike . Each year country's are holding on to ,or dropping down the rankings, due to in a lot of cases freak results and runs in the rankings , russia and romania are 2 examples of this. there must be a better and fairer system |
Author: panda
Date: 26-06-2006, 00:15
| I know what you mean, rbr, but I guess people who defend the existing system say that e.g. a system where the strength of the opponent is considered, or a system where results are weighted to matter more nearer the present would be more complicated.
I don't feel strongly myself - I am relatively new to this debate. |
Author: ferdi
Date: 26-06-2006, 01:20
Edited by: ferdi at: 26-06-2006, 09:26 | Max_F wrote:
{i>ELO seems to me slightly unbalanced because 1 game can move teams by 5-10 places even in top 20!{/i>
panda wrote:
{i>is it possible that this very large statistical sample gives a better result than for international football?{/i>
What is a "good result" in football ranking? How would we measure the quality of a ranking?
I see two big problems:
1.) It is not clear that an order exists at all. An order would mean that if team A is better than team B and team B is better than team C, then team A is as well better than team C. But I could imagine that 3 teams use 3 different types of tactics such that team A is superior to team B, team B is superiour to team C, but team C is superior to team A.
(Note that I do not talk about the fact that an inferiour team may beat a better team sometimes by chance.)
2.) It is very well possible that the quality of a team changes quite rapidly: See for example the Czechs: They recently had a very high ranked team, but performed quite poor at the World Cup. And now several players end their careers, and we simply don't know what we will expect from them in the near future. So in which position would you rank them correctly?
I think a good ranking should have two features:
It should - in most cases - give results that we intuitively expect. (For example immediately after a world cup the world champion should be ranked very high.)
It should be simple to calculate it. (By simplicity I mean: Not too many sophisticated rules like in the current FIFA ranking. For the ELO ranking you have to understand how to calculate powers, this is IMO not withstandig to that it is a simple method.) |
Author: panda
Date: 26-06-2006, 09:59
| @ferdi
As usual, I bow to your analysis.
In fact, the point about team A beats team B beats team C is especially relevant when teams play in their own country or in other continents. PLus, again to take example of chess, obviously the player stays the same person, the team can change completely.
So- back to the badger post - any ranking must be 'correct', BUT maybe no ranking can be very good. |
Author: exile
Date: 26-06-2006, 10:55
| The fact that a single result moves a team 5-10 places is not an argument against the ELO system.
1. There is really not much difference between a team in 40th place and one at 50th place. You would not be very surprised if the lower ranked team were to win.
2. Look at the Netherlands in the early 70s. They came from 34th position in Jan. 1970, 27th in 1971, 25th in 1972, 16th in 1973, and still 16th at the start of the 1974 World Cup - and up to 3rd by the end. Very often a team emerges from obscurity to be one of the World's top teams in a very short time. It just takes the coach to find a star player or hit on the correct tactics, and victory in a couple of games can lead to increased confidence.
As it happens Mexico didn't do too badly against Argentina! Nevertheless I think the ELO position of 17th is probably closer to reality.
From a Scottish point of view the ELO rankings are of great interest. According to the listings the greatest Scottish teams have been
1887 - reached 1950, 1st place 1930 - 1926, 3rd 1963 - 1883, 7th 1978 - 1949, 4th
and the worst Scottish teams
1934 - 1746, 12th 1948 - 1699, 17th 1955 - 1711, 18th 1961 - 1716, 21st 1971 - 1661, 35th 1987 - 1713, 28th
then Scotland dropped below 1700 in 1998 and have stayed there with new lows
1660 in 1990 1649 in 2003 1627 in 2003 1605 in 2003 1590 in 2004 1583 in 2004 1573 in 2004
Now we're at 1676, 44th place. |
Author: Max_F
Date: 28-06-2006, 00:56
Edited by: Max_F at: 28-06-2006, 01:08 | Yes, on 40th-50th places moving to 10 places isn't very strange, but in top 20 (even top 10) it slightly wonderful for me. Of course, maybe teams in top 10 so close each other, but why then them receive/lose so many points and permanently move by 5-10 places up and down after each game? I think it should decrease game effect to make rating more stable. And always exist problem of insignificant mathes in which team may not play at full potential, so weight of these mathes should be low (hard to determine such matches!). Of course it should take into account if we want to reflect real strenght of teams in our ranking. But if we consider rating only as "reward" (to use in draw etc.) it shouldn't do. |
Author: exile
Date: 28-06-2006, 12:39
| If rankings are made too stable, then a team like Holland or Poland in the early 70s, both of whom suddenly hit on a great set of players, doesn't get a chance to move up to the top of the rankings.
National sides are very reliant on the talent available - they can't buy players, so a great team in one year can be mediocre a year later, and vice versa. |
Author: Edgar
Date: 06-07-2006, 10:47
| The principles of calculation for the modified World Ranking will be announced to the media and general public today in Berlin. |
Author: Edgar
Date: 06-07-2006, 20:09
| From Fifa.com:
The most important change to the calculation of the ranking is that it will no longer take into account the last eight years of results but only the last four. At the same time, all of the other factors previously taken into account (result, importance of match, strength of opponents, regional strength, number of matches considered) were tested, analysed and, in some cases, totally revised. In fact, two of the factors that were previously used (goals scored and home advantage) will no longer have any impact on the ranking.
Next year, the official website of the football's world governing body, FIFA.com, will also offer an online tool that will make it possible to carry out hypothetical calculations for all teams based on the consolidated database of FIFA's Information Services Department.
Full article -> link |
Author: Edgar
Date: 07-07-2006, 08:48
| BERLIN (AP) - Italy should be the big mover when FIFA releases its revised ranking system after the World Cup final.
Brazil, already well clear using the old format, is expected to remain atop the FIFA rankings when the new formula is released next Wednesday.
Italy plays France in Sunday's World Cup final.
Marius Schneider, FIFA's head of information services, said Thursday that Italy could move into the top three depending on the final outcome. He expected Brazil to remain No. 1 regardless.
If the new method for calculating rankings was applied before the World Cup, Brazil would have led Czech Republic, Argentina, England and Spain, with France sixth and Italy 10th.
"Italy and Germany will move well," he said. "The teams that lose the most will be the ones like Cameroon and Nigeria, that would have been in the Top 10 under the new system but didn't qualify for the World Cup.
"The United States, which had good results at the 2002 World Cup, will lose those points and drop back."
Under the new system, results will be counted over a four-year period, with weight given to the strength of opponents and the calibre of the tournament.
The existing system, introduced in August 1993, calculates results across eight years and factors in the number of goals scored and conceded. It has long been criticized for being too complicated.
From now, every result will count, with three points given for a win, one point for a draw and nothing for a loss. Various factors like tournament level and standard of regional competitions also are accounted for.
Points from all matches will lose half their value after 12 months, meaning if Italy gained the maximum points for winning the World Cup, it would lose 50 per cent of those by July 2007.
The system is designed to give maximum weight to the most recent matches.
"It's a change, we can certainly expect there to be some surprises, but don't expect a massive revolution," Schneider said. "The teams that will benefit most are the ones that win important matches against higher-ranked teams."
FIFA president Sepp Blatter said the new system was more simple.
"We have acknowledged the need for a substantial revision. I am convinced our experts have come up with a satisfactory solution for a new way of calculating rankings," Blatter said in a statement. "It is difficult to meet everybody's expectations, but we are confident that the new system will provide an accurate measure of the strength of each member association." |
Author: skivaz
Date: 07-07-2006, 09:12
| finally a good ranking system, is what i was waiting for... now no more Greece as seeded teams and very difficult that a group with Ukraine France and Italy will be drawn again! |
Author: Osiris
Date: 07-07-2006, 09:21
| Where we can see a new formule of calculation?
It is not clear:
1. How regional strenth to be calculated? I think the range should be wider than 0.07 may be 0.4-0.5 2. How status of match to be calculated? I think the maximum difference should be 4 instead of 2. 3.What will be coefficients for previous years? current year- 1.0 one year before-0.5 two years before-? (may be 0.5x0.5),three years before- ?? (may be 0.5x0.5x0.5) |
Author: Osiris
Date: 07-07-2006, 09:28
| Skivaz,
FIFA ranking has nothing to do with a draw. Draw is based upon coefficients calculated of points wone in 2 previous qualification tournaments devided by number of games played. |
Author: Edgar
Date: 07-07-2006, 09:45
Edited by: Edgar at: 07-07-2006, 09:46 | @Osiris
The next FIFA Rankings will be published on July 12th. I guess your questions will be answered then.
@skivaz
As Osiris said, UEFA doesn't use the FIFA Rankings for EURO and WC qualifiers. |
Author: executor
Date: 07-07-2006, 10:00
| @Osiris
weights: 100% for previous year, 50% for the results of 2 years ago, 30% for 3 years ago and 20% for the oldest.
strength of the opponent: 200 - ranking / 100. So a team placed 30th will have a coefficient of 1,70.
regional value: 1 for UEFA, 0.98 for CONMEBOL and 0.85 for the others. |
Author: skivaz
Date: 07-07-2006, 10:12
Edited by: skivaz at: 07-07-2006, 10:12 | excuse me everybody for my lack of knowledge... but if they don't use for draws then what is their use?? nobody is going to be relegated in second division or so... it would be better if they apply the same criteria to establish the seeded teams in draws... i don't care if Italy is N. 2 or 20 in the FIFA ranking if at the end it still can have as opponent a strong team... right? |
Author: Edgar
Date: 07-07-2006, 10:13
Edited by: Edgar at: 07-07-2006, 10:17 | The FIFA Rankings are only used to determine the WC seeded teams. CAF uses the FIFA Rankings in the WC Qualifiers - CAF zone. The CAF teams qualified for the last WC are at the top, the rest are ranked using the FIFA Rankings. AFC uses the FIFA rankings only for the bottom teams playing in the preliminary round (before the group stage), then it ranks the team using the results in the last qualifying cycle. The same for CONCACAF. |
Author: Osiris
Date: 07-07-2006, 13:35
| @Edgar, @executor,
Thanks a lot. Regional strenth is better now. Difference 0.15 instead of 0.07 IMHO will be better like that: UEFA-1.00 CONMEBOL-1.00 Africa-0.85 Asia-0.80 N.America-0.80 Oceania-0.60
or at least Africa,Asia, N.America-0.85; Oceania-0.70
Downscale coefficients for previous years are exellent.
What are coefficients for WC finals:Continental finalls:qualification:friendlies? In the past were 2.00:1.75:1.50:1.00 IMHO the difference should be increased.
Eloranking has its own drawbacks: it is very shorttermed. Teams go up and down sharply after every match played. |
Author: Max_F
Date: 07-07-2006, 19:55
| I think that rating based only on game results isn't very good, because, as i said above, exists insignificant mathes in which teams may not play at full potential. I think it is necessary to take into account tournament achievements also. |
|
|