I agree with most of what you say with the exception of "everyone save SCO quite far outside" given Finland is ranked 28th...
Sweden vs Denmark should be treated the same way as Germany vs Italy because it`s just a technical necessity (goal difference could be important) to give the game a concrete score instead of a walkover victory. Another question: How is India treated? They played 0:0 against Iran at the Asian Cup but were excluded from the tournament after they couldn`t play against Chinese Taipei when they hadn`t enough players after positive covid tests. Their previous result was declared null and void by AFC. So what happens with the game against Iran with regard to rankings? 0:0 as World Cup qualifier/Continental finals match, 0:0 as friendly or is it not considered at all for the World ranking. I think the first option should be the appropriate solution.
That's a very good question - a game original played and certified but later voided. I agree that the appropriate thing to do *should* be using that original final score, and with a WCQ weighting. I don't think they would ever demote the weighting of the match, even if it ended up effectively being a friendly in terms of its competitive purpose. It's the only match India played in this window, so we'll see if their rating holds steady (game not considered) or if their rating drops, which would also tell us how the game was weighted.
Partly off-topic (meaning that it's perfectly on topic on this thread, but has nothing to do with the cancelled games or awarded forfeits you're lately discussing), but Tahiti's NT are having a tour in Europe, that's one of the few instances of inter-federations clashes for teams sitting below #100 (thus not having a chance to compete in big international tournaments), and there are some interesting scores so far. Luxembourg (#118 in FIFA rankings) just had two back-to-back wins vs Tahiti (#104 in FIFA rankings), the first one by 5-0 and the second by 11-0! How often does it happen that a team beats by 11-0 another team sitting 14 positions ahead of them? What kind of boost is Luxembourg expected to get from these scores? Also, I am not sure how exactly FIFA rankings work: have the rankings been adjusted, between the first win by 5-0 and the 11-0 win happening some days later? And if yes, could we anyway make some kind of an out-of-the-envelope calculation of how much is Luxembourg supposed to gain from these two matches? Despite the fact that they were just frriendlies, I expect quite some points from the size of the scores and the difference in rankings.
It's not that much. Only 15 points on the line per friendly. Luxembourg probably have the home advantage for ranking and are slight favorites. My computation says a 5-0 win gains Luxembourg 6.27 and the 11-0 gains them 6.42. Thus gaining about 13 points, and Tahiti losing 13 points. Tahiti will stay above them in the rankings. edit: There is very little difference, if you compute after each match, or after the whole bunch of games in the time span.
Two things to remember here. First, the difference in ranking is never (directly) important in calculating the points shift - rather, it's the difference in rating. And the further you get down the rankings, the smaller the rating gaps are between ranks. So the two teams in question are only 30pts apart despite being 14 ranks apart, meaning the difference between ranks is only about 2pts/rank in this region of the rankings. (Conversely, there are 130pts' difference between #1 USA and #7 BRA.) Second, considering the lack of play between confederations means that the ratings are less accurate for comparisons in this region of the rankings. While I think that UEFA teams in the region are relatively accurate on their own within UEFA because of how often UEFA teams have competitive matches, I don't think any other confeds play enough for the ratings in this region of the ranking to be highly accurate. Rankings are calculated immediately per game. So, in this case, the 11-0 game would be evaluated based on points already transferring between the two teams based on the 5-0 results earlier. That's a general rule for Elo system, and it's how I've done my points calculations posts here on occasion. (It's been a while since I've typed calcs for many games in a window all together for people to see, admittedly.) In this particular case, it's true that the order of calculations doesn't matter that much. However, that's because of several things happening at once: it's just friendlies, so points won't change much anyway; with HFA, the teams aren't super-close in points and the favored team won in both cases; and it's two teams facing each other twice in a row. Throw out any or all of those considerations, and when you do the calculations starts to matter more.
Although I am aware that Algarve Cup is just a set of friendly matches, is Italy gaining anything significant from beating #14 Denmark and #12 Norway and drawing vs #2 Sweden?
Italy will have gained enough points to overtake Denmark. However, the gap between Italy and Japan was noticeably bigger, nearly 40pts, and I think Japan probably held mostly steady through their five qualification matches. Italy, by 6-0'ing all three Algarve opponents, could've gained at most around 25pts, and they didn't do that, so no passing Japan unless Japan's draws pulled them down much more than their wins helped them (and I don't think that's the case).
It's ok like this: having Italy pass Japan in FIFA women's rankings would create a conflict of interest for me.
Alright, time for some closing thoughts/predictions now that all of the February tournaments are over. No change in the top 2, including the relative difference in rating between the two - both went 2-1-0 with the USA have weaker opponents than SWE but stronger scorelines. FRA will likely pass GER, with the latter falling out of the top three for the first time in history. With the exception of FRA-FIN, all five of the six games these teams played were between top-ten teams; the match weighting for those five is thus double a normal Friendly match, so with FRA earning three wins and GER going 0-1-2, that should make up the 35pt difference between the two. (GER fans might need to worry a bit - since a 1-0 victory over SWE in the 2020 Algarve, GER is on a 0-1-4 streak against current-top-ten teams, having lost to FRA and NED in 2021 as well.) #5-#9 in the rankings will get tighter. I'm not sure of exactly what all shuffling will happen there, but ENG will pass BRA for sure.
Will be interesting to see how the "Ada" effect impacts the Norway ranking going forward! I am excited and optimistic but always cautious and knocking on wood. Also, Norway should be top 10 in this latest ranking (but coddling North Korea, not by statistics but by allowing them to stay in the table, prevents Norway from being top 10). Ugh.
So India lost 11,24 points, Sweden improved by 2,41 points and Denmark lost 8,35 points. How have India vs Iran and Sweden vs Denmark been rated?
Yes, I am curious too about these two matches that we were pondering about last month. Anyway, after a good Algarve Cup's result, Italy exchanged positions with Denmark, as expected.
Looking at India first since they only played a single game: The game was obviously considered for the rankings since India's points changed, so in theory the only question is what weighting the game was given. India came into the match with roughly 60pts more than Iran, which coincidentally also gave them an "expected result" of roughly 0.60. A 0-0 draw is officially given an "actual result" of 0.47. Unfortunately, the expression [-11.24 ~ (0.47-0.60)*K ] gives a k value well over 80, when the highest weighting in the FIFA system is K=60, so something else is going on (hence the in theory earlier). Either India was credited with a loss instead of a draw, *or* they were given the draw but also credited with losses in their other two games. I'm guessing the latter because I can't find any single combination of K-weighting and "actual results" that would combine to give a bit more than -11 points. (Another possibility is a clerical error where the draw was counted twice, since the draw alone with the appropriate continental championship K=45 weighting would be worth slightly less than -6 points.) Denmark played two games, so we'll look at them next: Denmark and Italy came into Algarve with nearly identical ratings, so they both had an expected result of roughly 0.50. The 0-1 loss (actual result of 0.15) combined with a K-weighting of 15 (friendly with at least one team outside the top 10) gives a drop of -5.25 points, meaning the forfeited game had to be considered for the ranking to account for the remaining -3.1 points of Denmark's drop. The roughly 200pt gap between Sweden and Denmark gives Denmark an expected result of roughly 0.24. The expression [-3.1 ~ (a-0.24)*15] back-calculates an "actual result" of 0.0333..., which thankfully tracks very nicely with the "actual result" of 0.04 from an 0-3 loss. TL;DR The 0-3 Denmark forfeit was treated as such with a weighting for a friendly match The 0-0 India does not have an easy explanation at the moment from what I can see
Ah, derp, you're right! Not something you normally consider for teams at tournaments, so it completely slipped my mind there. Be right back!
Okay, yeah the home team hosting adjustment pushes the original points delta from roughly 60 to 160, making the expected result around 0.72 instead of 0.60, and (0.47-0.72)*45 = -11.25. So the 0-0 draw was treated as such with the continental championship weighting.
I guess Denmark was `punished` for their withdrawal because they had enough players left to play the match (only four covid-19 cases). If this makes sense depends on why the rankings are important: Do we need them for seedings in draws or do we want a realistic reflection of the strength of a team?
Good question. I think that from a proper perspective, the purpose of FIFA ratings if for seedings in draws. That is reflected in how they handle the Denmark situation.
Tahiti has lost 6,28 points after their defeats against Luxembourg (0:5, 0:11) and have dropped below Cook Islands but why was their 0:0 against Andorra not considered in rankings? Andorra`s 749 points as worst UEFA team didn`t change.
That's a good question... Something is up with with Andorra's ranking in general. FIFA went to two decimal points in their rankings several updates ago, and there are only a handful of nations whose ratings are still at just whole numbers and no recent change, mostly in OFC - presumably because those nations haven't played games since the shift. But Andorra obviously *has* played a game, so it's weird to see it not updated. It's either an error on FIFA's part, or something about the game meant it wasn't eligible to be considered for the rankings. =edit= In fact, Andorra has played *four* games since the shift to using decimal points, and yet their rating is the same as it was back in Dec.2020. Looks like FIFA simply isn't counting any of Andorra's recent games. I presume this is a compliance issue of some sort.
Spain stayed around the 13th place a lot of years until august 2021, finally integrating the top ten due to Japan and Australia 's poor performances at the olympic games (among other things). Then keeps on progress and takes the 7th place in the last ranking (1980 points) Since April 2021, less than one year, Spain got 50 points which is huge at the top levels ! We might see Spain passing the 2000 points symbolic bar next days after the two matches on Scotland and Brazil 's ground. If i am not wrong : For the SCO-Spain match (coefficient 3), it is an expected 30%/70% win. Spain can get about 10 points winning by just 2 goals (and we have to admit it's easily possible as Alexia and her friends scored 8 -0 the first leg ) For the Brazil-Spain match (coeff 2), it is an expected 62/38% win. But if Spain wins (it is also possible and anyway necessary to reach the 2000 bar), by 1 goal for example, it will be around 14 points more. The fastest progress for a country in a long time ? .
Record 181 associations are in these rankings. https://www.fifa.com/fifa-world-ranking/women?dateId=ranking_20220617