FIBA World Rankings

[mcruic]

mcruic is there a possiblility that you include more sports?

specifically realy popular team sports like Volleyball and Baseball

There's always a possibility - but it relies on being able to get results data. It takes quite a long time to collect all the results necessary for a meaningful ranking. I'd need to go back to at least 2005, but even then it's not always possible to find full results. In a sport such as baseball, for example, the mercy rule makes things complicated. A 10-0 victory after 8 innings is not the same as a 10-0 victory after 5 innings, in my opinion. But sometimes, only the final result is known, and there is no indication of when the match was stopped for the mercy rule.

For volleyball, even only going back to 2005, there are around 5,000 international matches - I'd need all data (including individual set results) - and this is sometimes not available.

I've added the sports where full results have been easy to get (but still, it requires time to type them all into a computer). If you know somewhere full baseball international matches are listed, let me know and I'll add a ranking for baseball too and any other sport you can think of - the only thing that is required is availability of results.

[sinobball]

The ELO system itself does have a few flaws - the main one being that it is not accurate for teams who lose all the time (but not by many points). For example, let's imagine India lost only 70-69 to China. With the ELO ranking, China would gain points, and India would lose points. This just doesn't make sense. The end result is that the gap between China and India has got larger, but in reality, a result like this shows that the gap is smaller. So, with my system, I've modified the ELO rating to take this into account.

First, I would like to say kudos man. I understand in a general way how you do the ranking now. This must be a lot of work to compile all the data. I mean how many "East Asian Games" are there in the world? To me it's a very minor tournament but I do think you have a good point. Even if China doesn't treat it seriously, Mongolia for example does (or at least I believe they do.)

But I respectfully disagree with "This just doesn't make sense." in your hypothetical example. So if China beat India by 1 it loses points? If India is really improved, or if China really regressed it would show in their W-L records for the other games.

Moreover, I don't think score margin is indicative of basketball strength, especially in more-or-less meaningless games. For example, if Jordan loses to China by 30+ or to Philippines by 10+ in group plays the results should not have the same weight as the resuts in more important rematches later. So I take it Iran got a lot of points by massacring opponents by 90+ despite losing in the most important match? But I do realize the difficulty in taking into account all the factors.

Again, good work man.

[Ciscorlie Horford]

Moreover, I don't think score margin is indicative of basketball strength, especially in more-or-less meaningless games. For example, if Jordan loses to China by 30+ or to Philippines by 10+ in group plays the results should not have the same weight as the resuts in more important rematches later. So I take it Iran got a lot of points by massacring opponents by 90+ despite losing in the most important match? But I do realize the difficulty in taking into account all the factors.

Again, good work man.

Yeah I think sinoball is right on this aspect, I also think that home court advantage should be taken into consideration because taht can determine the outcome of a game at times, like say if you beat a team that has home court advantage you should get more points than if it is just an average win, and but if you lose having home court advantage then again you should lose more points than normally or say, you beat a team and you had home court advantage but it was an ugly win, then that has to reflect too.


mcruic, if I can get you the data for baseball, could you do the ranking?

[damelo]

2005 - Copa América - Campeão (40 pontos)
2006 - Mundial - 19a colocação (5 pontos)
2007 - Copa América - 4a colocação (12 pontos)
2009 - Copa América - Campeão (40 pontos)
2010 - Mundial - 9a colocação (50 pontos)
2011 - Copa América - Vice-campeão (32 pontos)

That is 179 points for Brazil. So, why do we have 184 points in the FIBA rankings? Some could say my mistake?

I read somewhere only 18 teams could get points. So you should not count 5 points in 2006 if it's the case.

[mcruic]

Yeah I think sinoball is right on this aspect, I also think that home court advantage should be taken into consideration because taht can determine the outcome of a game at times, like say if you beat a team that has home court advantage you should get more points than if it is just an average win, and but if you lose having home court advantage then again you should lose more points than normally or say, you beat a team and you had home court advantage but it was an ugly win, then that has to reflect too.

mcruic, if I can get you the data for baseball, could you do the ranking?

About margin of victory - that's a hard one. I wrestled with NOT including margin of victory as a factor - but the rankings just don't make any sense any other way. It makes it very difficult to differentiate between teams outside of the top 20 or so. I wanted to make a system that could rank EVERY team, not just the top ones.

Margin of victory IN GENERAL does show something. Good teams generally beat really poor teams by lots of points. If margin of victory is not included, then there's no incentive for teams to score. They may as well just win all their matches by 1 point. In my opinion, a ranking should award teams who score points, as ultimately, the point of the game is to score.

Also, about China beating India by 1 point. There is no way of knowing how long it will be before India's improvement would be reflected in their W-L record. If India played 3 good teams immediately after (let's just say Philippines, Jordan, Iran), and lost all these games also by 1 point - this would obviously show they were improving - yet the ELO formula would punish them for these 4 matches.

About home court advantage - I have actually already factored that into the calculation as well (statistically, as you say, it does provide an advantage to the home team).

About baseball - I can get the data for the major tournaments no problem (world tournaments, Pan American Games, maybe even Asian Baseball Championship) - but for some tournaments, complete data is harder to find (even for the European Baseball Championships). I don't want to make a ranking unless it is global. I'll PM you sometime later this week with a list of tournaments I have, and a list I need results for - that might make it easier.

Thanks again for the constructive criticism (sinobball too) and for the offer of help.

[JGX]

Like JGX says - teams don't always bring their full teams. But I've taken the stance that if it's a national team tournament, teams should bring their full team. The minute a ranking system says "we can't include this match because Team X brought their B team", then teams will want their matches taken out of the rankings every time they get a bad result, using the excuse that it wasn't their "full A team".

Sure, but if you want to use your rankings in order to predict some future results you'll have to make some mental calculations about whether a team is underrated due to missing players in previous tournament. I'm not criticizing you here, just pointing out that there isn't really a perfect way to do it.

The ELO system itself does have a few flaws - the main one being that it is not accurate for teams who lose all the time (but not by many points). For example, let's imagine India lost only 70-69 to China. With the ELO ranking, China would gain points, and India would lose points. This just doesn't make sense. The end result is that the gap between China and India has got larger, but in reality, a result like this shows that the gap is smaller. So, with my system, I've modified the ELO rating to take this into account.

Yes, I was happy to see this in the explanation on your site!

China finished ahead of Iran in the Asian Championship, but apart from Iran's defeat to Jordan, Iran had better results (For example, Iran beat Lebanon by 31 points, China beat them by 20, with home advantage - Iran beat South Korea by 17, China beat them by only 13).

Also, Dominican Republic finished ahead of Puerto Rico at the FIBA Americas Championship, but that was due to one game where they defeated Puerto Rico. In the previous game, Puerto Rico beat Dominican Republic in group play and in general, performed better than DR throughout the tournament.

Agree with this. Hard to get people to understand though.

[JGX]

mcruic, if I can get you the data for baseball, could you do the ranking?

I already did the baseball ranking. It's a few years old but probably still accurate:

http://forums.interbasket.net/f38/national-teams-in-america-9164/p2.html

[Hepcat]

Interesting the current FIBA national team boys ranking:

1. U.S.A.
2. Lithuania
3. Serbia
4. Argentina
5. Canada
6. Croatia
7. Spain
8. Greece
9. France
10. Turkey
11. Australia
12. Russia
13. Brazil
14. China
15. Puerto Rico

FIBA Boys Comprehensive Ranking

[Dtown]

Stuff to expect post Olympics in the rankings

Guaranteed to lose points
Italy
Greece
Puerto Rico
New Zealand
Serbia
Angola

Guaranteed to gain points
Tunisia
Nigeria
Great Britain
Brazil
France
Russia

Potential Gains/Losses
Argentina loses points unless they gain gold
USA gains points with a Gold or Silver
Spain gains points with any finish better than 7th
Lithuania gains with any medal, loses if they don't make the semifinals.
Australia gains with a trip to the quarter-finals
China loses points if they don't make the quarter-finals.

Biggest loser: Italy, going to lose 200 points and plummet in the rankings finally.
Biggest winner: Brazil seems likely, especially if they contend for a medal. If not them Spain almost assuredly

[Dtown]

New FIBA Rankings out.

New Rankings

Top 5 exactly the same, Russia (6), France (8), Brazil (9) join the top ten. China, Serbia fall out the top 10, Italy falls out of the top 10 and 20.

[JGX]

Gotta love the FIBA rankings. GB women qualify automatically as hosts, lose all their games, move up 25 spots. GB men qualify automatically as hosts, fail to make the quarters, move up 20 spots.

[Bosnian 23]

Can someone tell me how Bosnia is ranked 24th? We have competed at Eurobasket for so many years, there's no way we can be that far down.

[Dtown]

Can someone tell me how Bosnia is ranked 24th? We have competed at Eurobasket for so many years, there's no way we can be that far down.

Actually Bosnia's ranked 50th currently.

Reasons why they are ranked so low
1: The rankings are based on an 8 year sliding scale. Meaning only results from 2005-2012 are counted currently.
2: In that time frame Bosnia's only been to two Eurobaskets and finished 15th and 17th.
3: While Eurobasket participants are awarded more points than any other region, they are awarded less points than teams playing in the Olympics or World Championships.

[soulis79]

The "Winners" of the Ranking: Russia +5 Places,France +4,Brazil +4,Nigeria +4,Great Britain +20 and Tunisia +10. All of them did not participate at Athens Olympics 2004. Brazil will grow more at Rio Olympics because they didn't make it to Beijing Olympics.

The "Loosers" on the other hand: Greece (lost points,no positions), Serbia -4,Puerto Rico -3 and of course Italy -14. Italy will drop more places because they didn't participate at Eurobasket 2009 and they had poor places at Eurobaskets(2007,2011).

[Bosnian 23]

Actually Bosnia's ranked 50th currently.

Reasons why they are ranked so low
1: The rankings are based on an 8 year sliding scale. Meaning only results from 2005-2012 are counted currently.
2: In that time frame Bosnia's only been to two Eurobaskets and finished 15th and 17th.
3: While Eurobasket participants are awarded more points than any other region, they are awarded less points than teams playing in the Olympics or World Championships.

lol I don't know why 24 was written, I meant 50. Thanks for the explanation thoug, but Rankings should be updated every week not every eight years!