How does Team USA rank against the other international teams?

You might have been one of those who watched the game between the United States and Canada women's basketball teams in person at Bridgeport, or saw it on television.  If you're reading this website you've probably read about it.

But did you know anything about the Khalipski Cup?  While the United States was taking care of Canada four national teams -- Spain, China, Turkey and host Belarus -- had a mini-tournament of their own.  All four of those teams are FIBA tournament teams and they were preparing in the same way that the United States prepared.

How did it go?  Each team played the other team one time.  Spain won all three games, with Belarus finishing 2-1, Turkey winning just one game and China losing all of them.  If you read lovewomensbasketball.com, you would know about the tournament but you would be one of a handful of people who did.

Americans only pay attention to national team play during the two major tournaments, the FIBA World Championships and the Olympic Games.  Each are held four years apart in even-numbered years.  Since the United States is so dominant -- and since a championship in one usually guarantees a spot in the other -- it doesn't have to play very many "friendly" matches with teams like Canada to tune up.  Given the nature of women's professional basketball and American players, it's hard enough to assemble a national team in time for the major tournaments, much less play a series of friendlies.

But other countries play such matches all the time.  For example, Turkey qualified for the 2014 FIBA World Championships because it's the host country, which always qualifies.  The United States qualified as the 2012 Olympic Winner.  The other 14 countries qualified by playing in regional tournaments for the right to take part in the FIBA World Championships.

Over the last four years, national teams have played approximately 900 games against each other.  I've tried to get as much information about these games as I can in order to give each of the national women's teams an Elo ranking. The link explains some of the mathematics about the Elo ranking, but a summary is that each team is given a certain number of points - I started with 1200 for each national team - and after a match, one team is awarded a certain number of the other team's points depending on the strength of the two teams and who wins.  Teams playing at home have a higher relative strength than teams on the road.

The current system I'm using puts a certain weight on games.  A championship match in FIBA is ranked three times as important in determining strength as a friendly match would be.  FIBA games are more important that games in FIBA feeder tournament.  This weighting counters the complaint that one team might not have its best players available for one kind of match but would certainly have its best player for the world championship.

Given all of this, the system isn't perfect.  Here's why:

a) It's hard to find information about all of the international games a team has played.  Sites like lovewomensbasketball.com have a lot of information, but I suspect the editor wouldn't claim that the information there is exhaustive.  FIBA has a historical database which is a big help but I have only gone back to July 2010 in gathering information.  The results of all major FIBA championships, regionals, and sub-regionals since July 2010 are represented.

b) Some teams don't play enough games. It usually takes between 30 and 40 games for a ranking under the Elo rating system to be accurate.

c) Teams play in different tiers sometimes, where teams in a "B group" of teams will play in one pool and the "A" teams in another.  After each group plays the teams in its pool, the worst players from the A pool are dropped and the best teams from the weaker B pool are added in...and then, they play all over again.  Games within the B group would be weaker games by definition, but the rating system does not reflect this.

Furthermore, Elo ratings are not meant to be predictive, or to be used to determine point spreads.  Rather, they give a rough judgment of how much better one team is over another.

With all of that in mind, here is what the FIBA rating system has come up with for games played since July 2010.

The results should not be surprising.  The United States is far above other national teams, and none of the results I was able to find are United States losses.   It looks like the USA's biggest threats might be Spain and France, and Turkey might sneak into medal contention.

Australia is only ranked fifth in the table above.  The rating system might unfairly penalize Australia here, as the Opals only have to deal with New Zealand in FIBA Oceania and they finished in third place in the 2012 Olympics.  But fans of Australia shouldn't fret too much - they'll have plenty of chances to improve that rating if they make it out of their group, and only Cuba might present a challenge.

Regardless of the relative ranking of teams, it seems the tournament always brings surprises.  In the last three major tournaments, the United States has faced three different countries in the finals.  No numeric rating system can calculate the twists and turns of international play...which is what make its so compelling.

For more on the upcoming World Championships, check out our FIBA section.