Discussion in 'Women's International' started by jonny63, Mar 17, 2006.
Yes, but it damages its seeding for 2185 anyway...
Wouldn't they get to where they're expected to beat Tonga and wouldn't get points for doing so?
Yeah, but they could rise to the top on the shoulders of Vanuatu.
In the post you answerd to I did say they don't use the ranking for anything else. That do not neccisarly mean they will use it for even that. But it do make figthing for anything else but the top rating (apart from host and holder) useless , wich it what is what a replay to.
June ranking :
The next FIFA rankings are due out in mid August. Just out of curiosity, has anyone done a ranking using the FIFA formula in the last week or so?
It would be fun to see how teams change as a result of the Olympic Competition.
I guess you need to ask Edgar about this.
Edit: Oh, he already made a preview.
FIFA ranking 4 years from now (July 21, 2016):
8.) North Korea
12.) England (when the old-timers retire, they won't be quite as good)
14.) South Korea
16.) New Zealand
Yeah because all that movement will highlight all the quality progression taking place in the women's game, just to see Germany go on to lift the WWC .
Just concentrate on what's happening now, because FIFA rank and the women's game, still find themselves about as compatible as Windows Millennium, and any decent piece of hardware you try to run it with .
Both great on their own, but together you get nothing but errors and disappointment.
FIFA rank is not even good on its own. It's crap no matter what way you look at it.
Why is it crap?
It's not crap on it's own, as it clearly works correctly in it's assessments of who wins the most games throughout the footballing year. Problem is though, the women's game is still too irregular/disconnected a sport across all it's various confederations, for FIFA rank alone to be a barometer of true quality.
So while everybody works on development towards the next WC, you can see massively different priorities for all the various national programs, in how they choose to use friendlies, which is of course a major factor in any teams FIFA rank.
It is crap on its own. If it's going to include all its members in a meaningful ranking, it should be just that - MEANINGFUL for ALL or at least MOST of its members. It might be meaningful for a few of the better teams (even this is debatable), but for the vast majority of teams, the ranking does not reflect anything real. Imagine this scenario - Anguilla draw 0-0 with Montserrat in one match, then play the game of their lives and lose 4-3 to Spain. Anguilla get more points for the draw with Montserrat. Is this logical? Of course it isn't. You may think it is because Anguilla has technically gotten a better result (DRAW vs DEFEAT), but in reality, the performance in the second match should be worth more.
Expanding outwards a little - any team who loses narrowly to Spain will generally be ranked much higher than a team who draws with Montserrat.
Do you think American Samoa are 22 places better than San Marino in reality? No - they are there because they are in the weakest confederation and managed a single win, while San Marino are in one of the strongest confederations and find it difficult to win. The FIFA ranking completely fails to grasp this simple concept - and this makes it crap. There have been plenty studies of ranking systems - FIFA's comes consistently near the bottom in terms of its ability to predict the outcome of matches (i.e. - there are more 'shocks' when using FIFA's ranking as a predictor of team strength - in reality, it's because FIFA's ranking does a bad job of sorting the teams into a decent order). I will repeat again - A ranking is NOT a league table. Therefore, it should not simply reflect "who has won the most games". If that were the case it should be called the "FIFA League". A ranking is (or should be) a rating of teams in descending order of strength.
And as for the women's ranking - it's fairly easy to produce a barometer of true quality, regardless of the perceived irregularity/disconnectedness. It's just that FIFA's ranking system does not allow this to happen.
Take it from me - if you're willing to look outside the Top 20 or 30 teams (if you're not, you're not a true fan of world football, and you don't deserve to be debating world rankings), FIFA's ranking is both illogical and crap.
I'm not a FIFA basher for the sake of it. I've been trying to improve my own ranking for years - I have a vested interest in being as accurate as possible. Along the way, it's become obvious that FIFA's ranking is simply not good. I've been at FIFA HQ in Zürich and told them the same thing. They know there are problems with it, believe me. But it's only a select few at FIFA who can do anything about it.
No, you make fair point's, and like you, I also don't look to FIFA rank in assessing a teams quality either, but the fact remains that a lot of people out there live by simple nature of a ranking system, for a female sport few people still get to see very often.
I'm no specialist in all things FIFA rank, and neither do I care to be one too, but there's always going to be people out there who will fight you tooth and nail with reasons as to why the current FIFA system is beneficial to the women's game.
If they want to take on the task of attempting to educate as to why, so be it, but I'm not going to go against your strain of thought in regard to this matter, because addressing me your already preaching to the chior.
Sorry - all that rant wasn't directed at you in particular It was a general "you". The FIFA rank is not great. But in actual fact, the women's ranking is BETTER than the men's ranking (they use different systems). The top 10 in FIFA's women's rank is pretty much spot on. It's below that level where it starts to creak (mostly due to badly chosen initial ratings for teams in certain confederations - particularly OFC).
I have never understood this thrashing of the women's ranking. It's an Elo ranking with important matches weighted and the results of blowouts reduced. Simple concept, really.
What you need to do to get good ratings is no mystery. Play good teams an beat them. Playing crappy teams doesn't do it at the top of the ranking.
So if you want to get good rankings, make sure you get in the tournaments and win them.
If you want to do something else, fine, but don't complain the ranking suck if you don't accept what it takes to get the ranking.
How do you propose to rank this team then?
Just because Anguilla "played the game of their lives" and lost a close game, should you rank them very high? Are you suggesting a ranking system that based on teams playing the games of their lives? Is that a good predictor of their strength?
Logical and reality? Really?
It's not a thrashing - it's a nudge towards improvement of the bits that aren't working. Any half-baked ranking system could get the good teams to the top with a "play the good teams and beat them" algorithm. But, the concept of the ranking is to sort out all the other teams who don't consistently beat the good teams, and put them into some meaningful order.
Papua New Guinea, for example, has a fairly decent ranking (49th) - undeservedly. They've never beaten a decent team - ever. They've got there on the strength of beating the likes of Fiji and Tahiti and Samoa. And this has put them above Cameroon and South Africa - who have just finished their campaign at the 2012 Olympic Games. Yes - it's an ELO rating. But it needs some (fairly simple) tweaking to prevent anomalies like these.
The very best teams can get the rankings they deserve fairly simply. But the middle-ranking teams cannot. This is where the ranking is particularly bad (and also towards the bottom). I'm not particularly bothered if people don't care about anyone outside the Top 30 - that kind of attitude would just invalidate their entire argument.
Obviously, Anguilla's entire ranking wouldn't be based on just a single game. What I'm saying is that they don't deserve to have points added to their total rating for a crappy draw v Montserrat when they get absolutely nothing for almost drawing with Spain.
I'm not suggesting anything. I'm pointing out bits that need to be improved. I've done the legwork myself, and changed my own ranking formula numerous times. Because I'm willing to accept it isn't perfect, I have improved it gradually and noticed the bits that still need to be improved. A system that does not distinguish between a 1-0 loss to Montserrat and a 1-0 loss to Spain is not a good system. If the ranking distinguishes between wins (an identical win against a good opponent will give you more points) and draws (a draw against a good opponent will give you more points), then it's logical that it should also give you more points for an identical loss to a good opponent.
Logical and reality - really.
Understand the concept first - then chip in with your two cents.
I know there are people on here who will defend FIFA's ranking to the death. Even if it is not too bad - it's not the best it could be. As FIFA is the world governing body, it should surely have the best ranking system. But, as with governing bodies of other sports, it doesn't respond well to criticism and persists with a ranking that nobody takes seriously.
You have to keep in mind what the ranking is for. Nobody but those two teams cares if Vanuatu is better than Papua New Guinea.
The ranking are used to determine confederation allotments, seeds and draws for the World Cup and Olympics.
As such, you only really care about the top 24-40 teams. In that group, the only real anomaly is with teams that don't play much, and no system will no a good job with those. There just aren't the data points.
The answer is to require more games per year for teams who wish to be ranked. I know there will be an outcry that counties can't afford it. It is a tough world, but you can't blame the rating.
But in case you hadn't noticed, there are qualifying rounds for the Olympics and World Cup. And in case you hadn't noticed, teams outside the Top 24-40 teams constitute the majority of the ranking. If the rankings are not done properly, teams will be improperly seeded and may be unfairly disadvantaged in their quest to qualify for major tournaments. And also, in case you hadn't noticed, teams don't magically appear in the Top 40. They have to get there first. You can't have a top without a middle and bottom. There is also no problem ranking teams who play fewer games. Plenty of systems can do it just fine.
You most certainly can blame the rating where it is justified, and it is most certainly justified. Especially if you have a working model of a better system that more accurately reflects reality (which quite a few people do).
Put it another way - South Africa and Cameroon both qualified for the Olympics (2nd biggest tournament in women's football) and still couldn't get above Papua New Guinea. Are you trying to tell me that the opinion of one sixth of the total teams at London 2012 does not matter?
For every pro you give for FIFA's ranking, there are at least 2 cons. Can you see how that may be problematic?
The premise that some opponents are "worth" more than others sounds like Page & Brin's PageRank, which has a similar problem in sorting anonymous web sites according to relevance. The gist is to expand the directed graph of web links, and assign bonus points to sink nodes (or, roughly, relevance in proportion to in-degree).
So a Spain, who plays many other good teams, would accumulate credibility / bonus value for any minnows who later face it in a one-off, while minnows who only play each other would not. NCAA's various tournament selection committees do a similar thing when they consider strength-of-schedule, including opponents' records, opponents' opponents' records, and so on. That's essentially the same as a breadth-first expansion in the PageRank graph of connections, i.e. to nodes at distance 2, distance 3, etc.
You'd still have the problem of assigning a base value to the minnows who have no inbound arcs. Einstein suggests a Cosmos-logical constant ...
Qualifiers is a selection issue, not a rankings issue, and is driven as much by politics as by statistics. And it's the same with Allotments to confederations. You don't fix the politics with the rankings.
Perhaps the selection for tournaments should be by results from broader international play against ranked teams from around the world as well as from confederation tournaments.
UEFA just experimented with that idea. Don't you think that turned out well?
Not sure what you mean by "a selection issue". For example, in the 2010 Women's Asian Cup qualifiers (as part of 2011 FIFA World Cup qualification), the draws and seedings followed FIFA rankings. The 5 lowest ranked (in FIFA rankings) played the preliminary round. In the next round, the 3 highest ranked teams not automatically qualifying for the finals (Vietnam, Taiwan, Thailand) were seeded 1, the next 3 (Myanmar, Iran, Hong Kong) were seeded 2 and the Top 3 from the preliminary round were seeded 3. Nothing to do with politics or selection. X number of teams enter the tournament of their own free will (they are not selected), and they are classified into seeding groups based on their most recent FIFA ranking.
FIFA's rankings also decide seedings for confederation tournaments in the men's game. The reason some confederations still use their own in-house rankings for their own tournaments is that they do not think FIFA's ranking is good enough.
Separate names with a comma.