Is FIFA Ranking useless? Is there a better way?

FIFA ranking system has improved quite a bit. It used to be really bad and more people complained.

I've studied the latest ranking system once and drew conclusion that it can be improved in one area. If I were a FIFA chief, I would adjust it so that older match record is counted less. This will have FIFA ranking reflect the current state of each NT.

 

I haven't studied the system at all, so was interested to hear that others have taken a look at it.

So, to count older match records as less would give teams with momentum (teams that have grown/improved in a relatively short period of time) to keep their momentum -- in other words encouraging the continued growth of football in that country as well. But what happens if, like Korea or Japan in the next five years, a country suddenly hits a bump...? Would it take away from what success it has built in the past decade? I'm not sure how far back these records go, or how the match records count per year...

Looking at the listing itself, I'd say it looks okay up to the 20s... but then it's hard for me to judge after that since the countries in the 30s and over bracket are hard to gauge... they don't have the same level of comprehensive international play (playing against tougher teams) experience.

 

Here is one. That is popular. http://www.eloratings.net If one scoll down past monthly matches. You can see history of nation. Then you can compare to FiFa, to get better feel.

 

We should use a combination of judges and point-ranking system. Obviously, it is impossible to rank 200 teams subjectively. However, they could have a panel of 10 people from diverse background to rank top 30 teams.

31 through 200 can be calculated through point system. But i'd like to see a rating system where u get more points for ranking teams higher than you, Not just different points for different continent. Right now, u can more points for beating a nameless south american country than beating Mexico or Korea right?

 

I don't like judging in sports. It invariably leads to corruption, bribes, error in judgment, disagreements, etc.. I know what your point is and it is a good idea if humans make perfect judges. The don't

 

yeah i guess... but even if humans dont make perfect judgment, i think it'll be better than the current pure statistics. we wouldnt have had usa and mexico ranked at top 10 if any subjective method was used.

 

I for one am glad that you arent the FIFA chief

Giving such great weight to recent mathces and not accounting for past performance as much will skew results greatly and allow for a very volatile table. You see with the current system you only move once you can achieve considerable improvement of lack there of over a period of time. In my opinion it would be wrong to elevate a team's standing immediately after they win 3 consecutive mathces. You should be consistent to be where you are.

 

But what if Mexico and USA were THAT good, but we refused to acknowledge it because they have not historically been powerhouse teams? If the numbers say Mexico is a Top 10 team, then Mexico is a Top 10 team whether we like it or not.

 

that's why we need subjective judgment.

If you've seen US or mexico play soccer, you should know they arent the top 10 footballing country in the world.

or maybe you dont know much about soccer? or you're just saying take what FIFA gives me and believe it to be true.

It is quite easy to argue that France, England, Portugal, Spain, Germany, Czeck, Italy, Holland, Brazil and Argentina are all better than US and Mexico.

 

No, we shouldn't.

There are problems with the FIFA rankings, but I would prefer them to just about any subjective judging system. They serve a usefulness for tournament seeding, and if we open seeding up to subjective judgments, we open the door wide for corruption of the process.

FIFA rankings could be improved, particularly with regards to comparisons across confeds, but they are better than people give them credit for.

 

Why? Why should historical bias supercede results on the field? Please explain.

I have. Mexico outplayed their group opponents in Japan - yes, even Italy. The Azzuri were more talented, but they never got in sync.

USA did well enough to advance out of their group and outplayed Germany before losing in the quarters. They may not have the pure talent that top Europeans teams have, but they take care of business on the field.

It's not about how much talent a team has. It's about how well a team plays. There's a reason teams play matches instead of just adding up player attributes from FIFA 2004.

I have never argued that the FIFA ranking is perfect. You just haven't sufficiently explained why historical bias is more important than results.

Who cares what you can argue? Talk is cheap. You can get a more accurate comparison by having an algorithm that doesn't depend on personal bias, and just deals with results and strength of schedule.

 

i didnt say we should rank teams historically. i said we need to use subejctive judgment, and u assume all subjective judgments are more biased then point system.

But no. i dont think algorithms and statistics can overcome the superiority of human judgment. Let's say Brazil and Nicaragua plays. Brazil dominated 99% of the time, then brazil does a back pass and ball goes into the goal. Brazil loses 1-0. Under statistics, this 1-0 would equal to Argentina's 1-0, where they dominate brazil and scores a goal after dominating.

Subejctive judgment will say although both nicaragua and argentina only beat brazil 1-0, argentina is a better team.

Statistics will hold Nicaragua and argentina to be equal in strength.





And if you're trying to tell you can argue that US is better than Germany, you're smoking something. You need to look at overall result, not just one match. Germany was a semi-freaking-finalist. US lost to Poland etc. Germany is clearly better than US not because of subjective bias, but because they have better team.




edited

 

I wish you'd stop telling me what I think. You can create a point system that is more biased than a judging system, but you can fix them - judges will always be biased to some extent. And when you're dealing with as many members and different regions as FIFA is, it is near impossible to select judges and minimize bias.

Better to create a single system everyone plays by, free of human interference.

Irrelevant for two reasons.

1. Soccer is a game decided by number of goals, not who dominates. Even if Nicaragua plays poorly, they did what was necessary to beat a better team, while Brazil, for all their talent, couldn't figure out how to beat a minnow.

2. Of course, there will be aberrations - that's why we look at a series of restuls instead of comparing single matches. Both Brazil and Argentina will play enough matches that any blips, fluke results will cancel themselves out. Even if Brazil loses to Nicaragua, as long as they do well in other matches, the Nicaragua results will diminish in importance. But if Brazil makes a habit of dominating matches and losing, then Brazil isn't that good a team, is it?

I should have explained myself better. I didn't mean to say that US is better than Germany, and we both agree that single matches don't mean much.
It was just a response to your point that USA and Mexico shouldn't be Top 10 teams and it should be obvious by watching the teams play. Well, by watching those teams play, we know that both teams are more than capable of holding their own against Top 10 teams.

Are they really Top 10 teams? Well, that's what the records should tell us, not some judges saying "There's no way they're better than the Czechs so I won't rank them higher than the Czechs".

 

i guess the difference is, i dont think its ever to create a point system that is superior to genuine human subjectivity. You think statistics can be superior to human judgment.

and i disagree about one more point. Although score is very important, the way the match was played is important as well. Was it their home or not? Did they have a bad referee red carding many players? There are many factors that needs to be considered PLUS the score line.



and with your czech remark. Nobody would have ranked czech at top 10 three years ago. I do now because they deserve. What is their FIFA ranking right now anyways? Shows that history isnt everything huh.

 

It's not so much statistics as it is "Does this team win a lot? Does this team beat good teams? Does this team do well in big tournaments?" It's about a pattern of performance.

IIRC correctly, you do get more points for winning away or at a neutral site. If you don't, you should. But that's the thing - the point formula can always improve. Humans will always be stupid and biased.

In any case, you're missing the point. Stuff like bad referees, bad field conditions etc, even themselves out over the long term. The ranking system keeps fluke results from affecting the points too much.

It's not. But the Czechs are a glamorous team that underachieved in the past and are now playing up to their potential.

Now, if the Czechs keep performing well, they will break into the Top 10 - and that's the thing. The FIFA rankings take a longer look (though I think it's a little too long) so that momentary blips, both good and bad, don't swing the rankings too wildly. It's great that the Czechs are doing well - now, if they can keep it up, they'll earn their place.

But the long term measure is better than judges making emotional decisions when they see a hot, glamorous team or a powerhouse going through a slump. Even though Brazil slumped during the last WCQ, their long term performance warranted their place in the Top 5 when emotional judges may have kept them out.

 

i have no idea what you're arguing for.

werent you just arguing with me and justifying US and Mexico getting ranked top 10? If that's not a momentary blip what is? Please at least be consistent. So should ranking be more "immediate" or more history based? Its ok for US and mexico to break top 10 for a good showing in WC and maybe Gold Cup. And as for Mexico playing well historically, i can't remember the last time they went to quarterfinals.

i can quite assure you. Brazil would not have dropped more than 3 places even if subjective measures were used.


The point system of what you're trying to do is frankly impossible for many reasons.

Whatever point system you use, it is after all, still subject to human judgment. For example. As i understand, you get same points for beating brazil or columbia; Italy or Belgium. As far as i know, u get same points for beating Brazil in the World Cup or the continental cup.

So then, should world cup and continental cup victory be equal? no. Of course world cup victory should be weighted more heavily. Then, one will need to subjectively weigh the difference in between World Cup and continental cup. Or more relavantly, Euro and Copa America. Same level of tournament. However, any idiot would know euro is more "important" and prestigous than copa america.



The next is limited number of matches and opponent.

Let's say Korea garners up enough weaklings and wins every match. Since Korea hasn't played anyone worthwhile, how will you rate korea? The current system of adding up all the points? Since they've won all their matches, will you statistically rank them number 1?

THen now. you will have to put additional measures "too many matches with guys not your calibre don't count" Then again, that will be unfair to, let's say, Saudi Arabia. Who didnt get a free pass to the WC and had to play every single weak country to qualify. Their wins dont count. So they get unfairly ranked lower.


Do you get the jist of it? I dont believe you can create a point system better than best-effort-human-judgment.

But of course, that's the difference in opinion between you and me.

 

Did you really study carefully? It's opposite of what you said. Older games are counted less.

Read 5.2 section of Ranking Procedure

 

I'm not arguing for the FIFA rankings as much as I am arguing against your idea that judges should rank countries because you think countries like Mexico and

I am being consistent. I never said US and Mexico should be ranked in the Top 10. I'm not arguing one way or the other. I just know that they wouldn't be out of place in the Top 10. And if the numbers say they are, they are. And Mexico and US have both been decent on a consistent basis - otherwise, they wouldn't be ranked so high - they haven't been spectacular, but they've gotten results. Both have appeared in all recent World Cups and managed good results against good opponents.

You're wrong. Look, I enjoy these discussions but I don't appreciate you wasting your time by trying to construct inane arguments without reading the FIFA procedures first.
http://fifa.com/en/mens/statistics/rank/procedures/0,2540,3,00.html

 

i'm not wrong. maybe i was wrong factually, but my idea still holds

Friendly match x 1.00
Continental championship preliminary x 1.50
World Cup preliminary match x 1.50
Continental championship finals match x 1.75
FIFA Confederations Cup match x 1.75
World Cup finals match x 2.00

UEFA x 1.00
CONMEBOL x 0.99
CAF x 0.94
CONCACAF x 0.94
AFC x 0.93
OFC x 0.93


therefore, this is what FIFA decided. Under this, Korea beating Maldives in World Cup reliminary is worth more than beating Italy in a friendly, which i find quite absurd.

While beating an asian nation in WC gives you 1.86, beating a south american nation in Confederations Cup gives you 1.75. In the end, Current ranking system is quite subjective as well.

You've repeatedly said you're not arguing for the current rankings. If so, i dont know why you gave me a link. To prove my facts wrong? i admit i didnt look in to at all. So bashing my fifa-rating-criticism is irrelevant.

I think i sorted out my argument in a nutshell. With limited number of matches, it is impossible to statistically and accruately sort out 200 countries in order of strength.



Then why dont u devise a perfect ranking system. I assure you, someone will be able to find a hole and subjectivity in your ranking system.

As i said, how will you rank a team that only got to play crappy teams all year long?

 

I know. Perhaps, my English wasn't clear enough. I want FIFA to count even less of old results. As far as 8 year old results is counted. To me, it should go only back to 4 - 5 years at the most.

 

Were you aware that FIFA counts matches as old as 8 years to calculate FIFA ranking? What do you think about counting up to 4 year old matches instead of 8?

 

It wouldn't matter much in my confederation (OFC).

FIFA's ranking system also counts (ie totals the points of) the best 7 results (matches) for each & every one of those 8 years (4 in your model).

Too bad if you don't (or more to the point can't) play 7 or more matches each & every year.

That's the way it actually is in the OFC !

 

Actually there's more to it than that.

In essence, for the teams that play relatively regularly, only the best 7 matches of any one year count toward your FIFA ranking to any significance. In addition to the above multipliers, strength of schedule for individual games also counts meaning that in fact I believe the Italy win would be worth more points actually if it was by the same score.

By my count, South Korea would get 20.89 points for a 1-0 win over Maldives in your scenario, and 23.06 points for a 1-0 win against Italy in your scenario.

It's also important to note that really only the end of year rankings need to be accurate because those are the ones FIFA uses for seeding at the World Cup (as well as previous World Cup performance).

At the moment Mexico is ranked 10th and the USA is ranked 11th. While that would be higher than I would rank them, it's not an awful lot higher. Certainly it's reasonable to put both in the top 20.

 

The OFC is really the only instance where the FIFA rankings are pretty broken, but that's as much a result of the difficult task ranking OFC teams as it is the system. Any system is going to have trouble gauging the strength of the OFC due to the scarcity of meaningful inter-confederation matches by the OFC.

So Australia is underrated by FIFA who has them in the mid 50s, but who in reality probably belong in the mid 30s based on their results.

Now you can argue that the setup inhibits Australia from getting good results because of the nature of the playoffs, but it's difficult enough ranking teams by actual results much less ranking them by what you think the results would be under a different scenario.