Alright as long as we get a ranking that suits our run our performances and our wins against nations such as Spain Argentina and England in major tournaments the last 2 years
I read somewhere maybe in this thread the reason for not losing points in knockouts and at the time of reading it made sense. Here is the hypothetical scenario... let’s say 2 teams in a group are evenly matched and are close in the FIFA ranking. Sweden beats Costa Rica to advance... Sweden is now just above CR in the rankings. Now they face Brazil in the knockouts and lose. Sweden’s loss drops them below CR in the ranking, even though they are the better team and advanced farther in the tournament.
why should a team get extra points simply for playing extra games ? a ranking is supposed to be a measure of how good a team is, not how many games they play. If you get a result it shouldn't matter whether it is in the first round or in the final. I've already referenced France v Croatia and France v Australia. Under this new Elo Australia loses points for a one goal loss to France, while Croatia doesn't lose a thing for a 2 goal loss. That is simply illogical and does not represent what a team achieves from match to match. Another example that this time favours Croatia, when I compare the rankings I devised to FIFA. Lets look at Croatia v Argentina against France v Argentina. 3-0 and 4-3 respectively. Under my rankings Croatia would get more points for the fact that they win by 3 goals. Under the new Elo France would get more points based on it being a knockout game. Again this is simply ridiculous and illogical, and does not help to reflect the true strength of teams which should be based on nothing but RESULTS. Not how many matches you play, or what round you play them in. To me this formula stinks of attempting to make the rankings exactly mirror the final World Cup results, which quite frankly is silly because as you have admitted in the past, it is a very small sample size. The rankings should be an overall assesment of a teams results, yes weighing the Worldcup very highly, but removing knockout round losses goes way too far. And here again is where my rankings are superior because instead of getting 0 points for ANY LOSS, you actually get points for losses in my rankings based on the strength of the team you lose against, and based on how many goals you lose to them by.
The rules for 2026 have not been established... maybe there is just one seeded team. Here is what FIFA has stated about this topic... “* The host country would also automatically qualify for the FIFA World Cup, and its slot would be taken from the quota of its confederation. In the event of co-hosting, the number of host countries to qualify automatically would be decided by the FIFA Council.” Based on this it’s posdible Canada does not auto qualify. It sure would be weird if they hosted but didn’t make the WC.
Results in football aren't all equally the same value, even when they are equally the same result! The value of results in a tournament is tied to the ultimate objectives: advance from the group, then advance as far as you can. In trying to achieve these objectives, you sometimes will be alright with a draw in the group stage and sometimes will need to win. Some losses don't mean much while others mean a lot. The balance between all of this to me is best achieved through what I suggested: add/subtract points in the normal ELO manner when a side wins/loses. And for achieving the objective, which means getting to play more games, give them the extra reward points. Incidentally, that is basically the only extent to which I would give goal difference all that much value. If the goal difference helps you play more games because of the tie-breaker rules, then you get the extra reward points. Otherwise, I wouldn't try to tell people a false story based on the margins of victory, which often reflect things that have little to do with comparative strength. For instance, Croatia could have defeated France or that game might have ended in a draw/pk shootout but for some unfortunate incidents. But, that game against France, might have also ended 6:1 in favor of France. That wouldn't mean Croatia is weaker than all the other teams that France barely beat by 1 goal, including Australia.
Well first of all Sweden and Costa Rica are not close in the rankings to begin with. So that is a bad example. But I will make up an example of how I believe it should be done, this is what I do with my rankings. Lets say Uruguay and Denmark are in the same group. Under my current rankings before the World cup both teams are ranked just inside the Top 20. Uruguay beats Denmark and would get around 80 points, while Denmark would get around 27 for a one goal loss. Uruguay goes through and loses to Brazil who is ranked #1. From a one goal loss to Brazil they would get 33 points. The average points from those 2 matches are 57 points. While the points gained by Denmark in their loss against Uruguay is 27. So Uruguay's average points from those 2 games including one loss is still much higher than Denmark. I dont know how Elo would rank this, but I believe under the old FIFA, Uruguay would still come out on top over those 2 matches than Denmark.
Yes it would. For that SINGLE match Croatia was weaker than Australia. Remember a ranking is a combination of results, not one single result. So Croatia absolutely was weaker than Australia in that one match. All we can go by is THE RESULT. We cannot get into subjective things like "this team actually played better" or "they lost by more because they were trying to equalize" or "it was a worldcup final" that is all subjective nonsense that has no place in a ranking. Otherwise why not just give Morocco points for "beating" Spain because they were cheated by the refs, and why not award them a draw against Iran because they "deserved" it.
I would still go with what I have said. My view is that the most methodologically sound ranking system is the one by ELO. I would, however, not use non-FIFA match day results and take out the garbage from the data they use. For major tournaments, especially the World Cup, I would give points for each game that is played and, thereby, reward sides that advance further and play more often, but I would also do the normal system of adding/substracting points for each individual result using ELO's system which looks to who you lose/draw/or win against to determine the points you win/lose etc.
I see the justification for not losing points in the knockouts, and having them be worth more. The teams that make it this far should be rewarded and FIFA is making it so that there is no negative impact for losing in the knockouts. There will be 8 first round losers in 2022 and 16 in 2026. These teams shouldn’t drop points for advancing to the knockout. It may not be a perfect way to do it, but I can understand why they are doing it this way.
Maybe Sweden and Costa Rica are not close in ELO, but they are 23 & 24 in the current FIFA rank. Your system seems to work well. Now how would this same senario work in FIFAs new proposed system with and without dropping points in the Knockouts?
But the teams are ALREADY rewarded for getting the points from results that put them into the knockout rounds IN THE FIRST PLACE. What you are essentially doing is rewarding a team TWICE, and that frankly is not logical or reflective of reality. Why shouldn't there be a negative impact for losing in the knockout rounds. It means you aren't good enough, so why should you be rewarded for LOSING. The knockout round teams would still be likely ranked higher than the teams who didn't make the knockouts. What you guys are essentially saying is that "team X" is better than "team Y" simply because they are playing in the knockout rounds. Well this is simply a faulty assumption because we all know that groups are unequal. But we cannot deny if "team X" beats "team Z" and gets points for that victory. Whether that victory lands them in the knockout rounds or not is irrelevant, what is relevant is that they beat "team Z" Again what this ranking is doing, is rewarding teams TWICE for one action. That is a misrepresentation of reality and a scewing of results to favour teams that did well, instead of just allowing their already positive results to simply speak for themselves.
Well the new fifa/elo doesnt do averages they do adding and subtracting from each match. I dont necessarily mind that but I find their formula a little hard to understand. But with my ranking it leaves Uruguay with around 80 points for a win over Denmark and around 33 for a loss to Brazil averaging at 57 for 2 matches and still far above Denmark at 27 points from the one loss to Uruguay. Now lets just assume that the new fifa/elo did averages instead of adding substracting points, just to make my point. The 33 points for the loss against Brazil would not be counted and the average from 2 matches would still be 80 points. So it basically awards teams for losing simply because they are in knockout rounds. Under my system Uruguay would average 57 points from those 2 matches to Denmarks 27 from the one in the scenario we made up, but under the new FIFA/ELo it would be 80 points to 27. In my ranking Uruguay is still well above Denmark even though it lost the match to Brazil, while in FIFA/elo it is way way above Denmark. I think that is an unrealistic disparity given that Uruguay would have just lost a match and wasn't penalized for it. Hope this makes sense.
Ultimately the only way we can really test any ranking is through a subjective look at it. If we look at a list and it looks fairly accurate, we can say that it is probably a pretty good measure. When we see teams like Nigeria, Japan ranked near 50 and teams like Switzerland in the top 10 we gotta ask some serious questions.
FIFA has nine Asian teams ahead of Canada, with Qatar ranked fifteenth in Asia. ELO has Canada tied with China, who is tenth in Asia, and has Qatar twelfth in Asia. I think it's close enough that the travel matters. If there was a two leg series, hosting the first leg could be an advantage because that team would need to travel once and the other team would need to travel twice. If you want to determine which team is better, maybe there should be one game at a neutral site in between such as in England. I think Canada is much better than Australia's C team. In World Cup 2014, USA won the goal differential tiebreaker over Portugal after both teams went 1-1-1 and drew against each other. USA lost to Belgium in the Round of 16 2-1 in extra time. Portugal lost 24 ELO points. USA, who started with fewer points, had a net gain from the Group Stage, but lost 27 points for losing to Belgium to lose 20 in the World Cup, which was a change of 4 better than Portugal. I think USA's change should have been more than 4 better than Portugal. Furthermore, in November 2014, USA lost a friendly to Ireland 4-1 and lost 18 ELO points. Why did losing to Belgium by 1 make them lose more points than losing to Ireland by 3?
Here's what you get with the new method if you simply double the match weights. It makes more sense at the top. You could argue Croatia are still a little hard done but two of their knockout victories were shootouts. We'll see what FIFA does but I think this would be a simple and logical fix. For example, Mexico takes 90 points from Germany with their win instead of 45.
This is what the rankings would have looked like before the 2018 World Cup if the new formula was applied with double weight to the results after the 2014 World Cup. It still has the Netherlands too high, but 11th is a lot better than 5th. I say just double the weights and let it fly.
I am working on my world rankings update post World Cup. I have finished the rankings for all 32 world cup sides, now I must update the rest of the teams who didn't make the world cup, and include all of their friendly matches in the pre WC period of June. So the 32 teams rank as such with their points per match totals 1) France - 89 2) Brazil - 82 3) Belgium - 82 4) Croatia - 74 5) England - 74 6) Spain - 73 7) Germany - 66 8) Denmark - 64 9) Colombia - 64 10) Sweden - 63 11) Switzerland - 62 12) Portugal - 61 13) Uruguay - 60 14) Argentina - 59 15) Peru - 57 16) Mexico - 54 17) Russia - 53 18) Senegal - 53 19) Nigeria - 52 20) Morocco - 51 21) Iran - 48 22) Poland - 48 23) Iceland - 48 24) Japan - 46 25) Serbia - 46 26) Tunisia - 44 27) S. Korea - 44 28) Australia - 41 29) C. Rica - 40 30) Egypt - 39 31) S. Arabia - 32 32) Panama - 29
Are those numbers rounded? I notice some teams have the same points tallies, but you don't indicate that they are tied in the rankings.
We should see the Elo ratings as a basket of points which is supposed to remain steady overall. At each game, the number of points earned by team A is exactly the same as the number of points lost by team B. This system has lots of virtues, but indeed, as pointed out by Iranian Monitor, the big issue comes from the fact that most international games are played within a confederation. As a result, there is very little points exchanged between confederations overall. This results for instance in Conmebol having a lot more points per team on average than CAF. As a result, no matter the outcome of games, the Conmebol lower teams will be naturally brought up in regular games within their confederation whereas the better teams in CAF will be pulled down in the same frame. I first thought the issue would be solved with lower confederations getting better, so I've calculated the sum of Elo ratings for each confederation and checked how it evolved over time (from the 1950's to nowadays). And things are actually getting worse! Conmebol steadily strengthened, UEFA does so too but at a lower rate. All other confederations are actually getting weaker. The reason for this is actually rather simple. As long as Conmebol and UEFA teams will have better results than other confederations on average during intercontinental games, they will keep earning points (even if at a slower rate over time). Now, another way to look at things is simply to not see it as a flaw and accept it as a fact. What I mean is that indeed, African, Asian and North American teams don't really perform better now than they did in the 1990's. And after all, if Japan had actually won against Belgium, it would have earned significantly more points than what it eventually lost because of the Red devils comeback. This doesn't really change the fact that, when we look at the global ranking at any given time, Bolivia will be overrated compared to Nigeria, but now that's also the limit of the exercice of a global ranking in the first place. National teams simply don't meet each other enough regularly and on a wide enough geographic basis to get any form of ranking method totally consistent in the first place.
While the rules might not have been established, there is nothing to suggest that a 100-year tradition will change. The expansion of the WC makes it even less likely that co-hosts would'nt qualify automatically. I imagine the only reason its not a set rule is to just guard against a potential scenario where we might have, I dunno, 8 co-hosts sometime in the future. But as long as there are 3 or less, they will all qualify for sure.
It also might be to only have one host team be seeded. My guess is FIFA will auto qualify all 3, but won't auto seed all 3.
If the 16 seeds in 2026 are USA and the top 15 in the FIFA Rankings, Mexico might be seeded based on their ranking, but Canada would need a giant improvement to be seeded.
So far all are speculations . Always the host is seeded. I do not see a reason why the 3 host countries are not seeded on this occasion. Especially in a tournament of 48 countries.