FIFA rankings 'loophole' likely to disappear after World Cup draw https://uk.reuters.com/article/uk-s...-disappear-after-world-cup-draw-idUKKBN1DG2TO (Reuters) - A loophole that allows national teams to boost their seedings for the World Cup by avoiding friendly matches is likely to disappear after next year’s finals in Russia, with changes due to the international calendar and a FIFA review of their rankings system. Currently seedings are based on the FIFA world rankings, using a points system which calculates average points for each game and is weighted against friendly matches. The result has been that some teams, such as Poland and Switzerland, who have played only a small amount of friendlies, have gained an advantage over teams who have played regular friendly games. Friendly internationals played in November have not counted towards the seeding pots for the World Cup as they are based on last month’s FIFA rankings. But two changes are set to impact the system significantly. From September 2018, European and CONCACAF region teams will begin playing in Nations League competitions which will be played on the dates in the FIFA calendar traditionally set aside for friendly games. That means there will be very little space available for friendly matches for the 90 national teams who are FIFA members in those two confederations. That trend could well spread with FIFA confirming recently that they are in talks with confederations about creating a global Nations League structure, which would further reduce the amount of friendly games. FIFA also said it plans to review the rankings system. “FIFA is reviewing the FIFA World Ranking system and will take a decision after the completion of qualification for the 2018 FIFA World Cup if any changes are be made to improve the ranking,” the organisation said recently. Asked if those changes would address the issue of nations avoiding friendly matches to improve their rankings and how Nations League games would be factored into the calculation, FIFA told Reuters: “The FIFA ranking will be reviewed in the near future. As long as this has not been done, we can’t speculate on future scenarios. The current ranking takes into account all official matches.” The World Cup draw will be held on Dec. 1 in Moscow.
https://resources.fifa.com/image/up...fifa-council.pdf?cloudid=ynviyg70bpzo1on3cdii has the agenda for the FIFA Council meeting on 10 June and the "FIFA/Coca-Cola World Ranking formula" is one of the items.
How about using the World Cup and continental cup results only? World Cup: 100 pts for winner, 80 pts for runner-up, 70 for 3rd, 60 for 4th, 50 for QF, 40 for R16, 30 for 3rd in group, 20 for 4th in group, 10 pts for repechage elimination. Continental cup: 50 pts for winner, 40 pts for runner-up, 30 pts for 3rd, 25 pts for 4th, 20 for QF, 15 for R16, 10 for 9th-12th, 6 for 13th-16th, 4 for 17th-24th.
What about all the teams that don't qualify for those tournaments? Antigua and Barbuda is 124th out of 211 teams in the FIFA Rankings. They have never qualified for a World Cup or Gold Cup, but they deserve to be ranked ahead of other teams rather than having a big tie with 0 points.
Yeah.......national teams already don't play many matches in a given year. You can't just factor world cup and continental cup results.
Revision of the FIFA / Coca - Cola World Ranking https://resources.fifa.com/image/up...orld-ranking.pdf?cloudid=jgxjkdrj1jfwyunjbkha
I'm no expert on rankings, but it looks a load of rubbish to me. Seems like it's Elo, modified to take into account "a number of FIFA's considerations", i.e. irrelevant crap.
There is no perfect system and a lot of people around here seem to think that ELO is a more accurate way of calculating the rankings. I think this was a relatively progressive move by FIFA all things considered and I don't know what else they could have done to make it more fair.
The problem is national teams play too few amount of games in a given year to get an extreme accurate ranking. So yeah the FIFA rankings needed improvement but in the end it's going to be hard to get a real accurate ranking system with this stuff. Small samples.
It's neither crap nor irrelevant. Lots of ELO systems have such modifiers and as long as the modifiers are sound (which they seem to mostly be) then they help get a better result. IMO they have a few too many levels in the system but then again, most people do think that performances in continental championships and the WC is more important so it's not ridiculous to do it that way. The one I find a bit strange is not losing points when you lose a knock-out game but I'm prepared to see how it works out before passing judgement on that one. Actually that isn't as much of a problem as one might think. There has been research done and both the ranking the eloratings.net site uses for mens football and the one FIFA uses for womens football have been shown to be very good at predicting the outcome of games, and that really is the best test there is for a ranking system.
They need to explain whether penalty shoot out loses are draws or loses. If a draw the team that gets knock out on penalties may have actually gained points. Will those be excluded or included.
This isn't really that much different to FIFA using their ranks from before the UEFA play-offs to determine the final seedings (ie October 2017 rather than November 2017) - because then teams that finished 2nd in their groups and then won a play-off could leap ahead of the team that finished 1st in the group because they got to play those extra games. This is the opposite case - if Team A beats Team B to get out of the group phase, but then loses to (say) Brazil in the Round of 16 - should Team A end up with a lower ranking because they lose points for their match against Brazil (these matches will probably have a higher weight/be worth more and so put more points at risk)? It seems that they are trying to make the latter stages of the tournament purely upside for rankings, with no chance of slipping back just because you advanced and then sucked. Of course, you could just scrape into the WC finals ahead of some other team, and then get crushed in the finals and end up ranked much lower than the team you originally beat - but no ranking system is going to be perfect. J
Also the fact that the team losing in knockout rounds of major competition doesn’t lose any Elo ranking points is a bit weird. It may mean that the number of points shared between all nations is constantly increasing, which isn’t the case in a normal Elo ranking. I guess more details will be fed out soon.
They could have just used Elo, no? It's looking to me like teams that make the World Cup finals are automatically going to earn shedloads of points regardless of how well they do, which will mean they are inevitably going to get seeded for the next qualifiers, and on it goes. I could be doing the calculations wrongly though!
Yea I can somewhat see the idea but IMO that isn't much of an issue so the "fix" isn't really needed. But as I said, I'll wait and see before I say it's a big problem. Higher weight yes but as they should be against better ranked teams the loss of points will should be lower so overall the points at risk would be lower too. That's the potential problem I see too. It's something they'll need to look closely at when the new system starts to be used.
It could be that the published formula is just wrong: @FIFAcom There is an error in your formula for calculating the new #FIFARanking system. The formula for We probably should look like in this screenshot, but in your document you are missing the superscript for the exponent. https://t.co/pNlaO2SkYP pic.twitter.com/C7ZWaYCq8W— Martin Bennedik (@bennedik) June 11, 2018
They are. Having a K-value that changes depending on certain factors is a common feature in ELO systems. Some, like most chess ratings, use it mainly for adjusting the values for players that haven't played that many rated games while some, like FIFA, use it to give different weight to different matches/tournaments and some use it even more. In the end there is a decent argument to be made that ELO systems produce a better rating with a K-value than without. But of course it is something to keep an eye out for in the future. Firstly, they already do that with the current system. Secondly, most people think that is a good thing. That teams should benefit from playing in high level tournaments. A common critique of the current system is that it places too little importance on performances at the WC/continental cups, that's why they have done all kinds of alternative temporary rankings to seed the WC through the years (and why UEFA have used its own rankings for Euros and such). And thirdly, I'm not sure the effect you are afraid of will be all that big with an ELO system (certainly not sure it will be any bigger than in the current system). But, again, it is something to look at when the new system goes live.
Yea it's wrong but not in any way that affects what we've been discussing here. It's just a graphical issue in the news release (I assume at least) because Word is piss poor for mathematical formulas.
What have we been discussing here then? I've been discussing the rankings that the published method produces. It looks like they are massively wrong.
Nah - it's the new ones. Effectively Elo with different weightings and a "floor" for knock-out rounds. It's a massively easier idea to explain - if only because it actually works like people seem to think the current rankings actually work. J
How come https://www.fifa.com/news/all-news.html doesn't have an article about it? Here are some comments: 1. The ratio of highest to lowest weighting is 12. It used to be 4. Weighting friendlies not during International Match Calendar windows might affect who the USA wants to play during Camp Cupcake. 2. We now know how much Nations League games will be weighted. 3. There is no more confederation weighting. USA fans have complained that they get fewer points from beating or drawing with a top team like France than a UEFA team would get from the same result. World Cup 2018 will not impact a confederation weighting like World Cups used to do. 4. There were months when a team was guaranteed to lose FIFA Ranking points. This won't happen if teams only change points after they play. 5. In the FIFA Rankings, Number 1 has 9.11 times as many points as Number 150. FIFA has 211 members, but I used Number 150 because the bottom teams have few enough points that the ratio would be giant. In the ELO Ratings, Number 1 has 1.74 times as many points as Number 155. I used Number 155 because the ELO Ratings include areas that are not members of FIFA, and I didn't feel like manually taking these teams out. Therefore the teams will seem closer together in the new rankings. 6. It will be interesting to see how much the rankings change after World Cup 2018 compared to after World Cups 2010 and 2014 when the formula did not change and after World Cup 2006 which was the last formula change.
That's my point tho. Whatever FIFA does to improve their system most likely still won't be better than what we currently have.