I agree that it's not appropriate to assess the sporting prowess of a
country with only medal counts, Chris, but why stop at the top 8, and what's
the basis for assuming a bronze plus a silver equals a gold (and so on down
the line)?  Some would argue that it's not worth assessing countries at all,
but there is a need for a reasonable evidence-based approach to assess how
well athletes are performing within and between countries when decisions
need to be made about how much money to spend on high-performance sport.

We've recently come up with what we believe is the definitive approach using
the world ranks of countries' athletes. In brief, you add up the inverse of
the ranks, an approach that makes 1 gold worth 2 silver, 3 bronze, 4 fourth
places etc.  It's not something we just dreamed up, either: incredibly, 1
over an athlete's present rank is the approximate probability the athlete
will ever get to top rank in their sporting career, so when you add up all
the 1/ranks for a country's athletes, it's an estimate of the country's
top-rank ability.  Well, that's what we found with swimming, anyway, but
we'd be very surprised if it doesn't work with all sports. Rita Malcata
presented this prizeless prizeworthy work at the ECSS conference
http://sportsci.org/2012/ECSS.htm , and we have just submitted it for
publication.  See below for the abstract.

With the Olympics, you would use the final rankings as far down as they go.
You would end up with an estimate of each country's gold-medal ability. Use
whatever population or GDP denominator you like to make your country look
better, or in the case of the US, China, Britain and Russia, pretend that
population and wealth don't count.  Does anyone know if anyone has counted
the individual US states as separate countries for this purpose?  Either
that, or go the other way and call all countries in the EU (plus Finland,
Norway and Switzerland) the United States of Europe?  Wow, just a quick look
at http://www.london2012.com/medals/medal-count/ shows that  the USE would
trounce the USA.  Hmmm... United States of Oceania?  Aussies, you are
seriously short of medals this time, so the Kiwis' haul would be a big help.
What about the United States of Asia and the United States of Africa?  I
suppose it would have to be the United States of North America and the
United States of South America to make it all work.  Canadians and Mexicans,
how would you like to be lumped in with your big bully brother?  

Will

Using Athletes' World Rankings to Assess Performance of Countries 
Rita M. Malcata, Tom J. Vandenbogaerde, Will G. Hopkins

The sporting performance of countries needs to be evaluated with a fair
method for including performances of athletes not winning medals. Purpose:
To develop country rankings based on athlete world rankings in the sport of
swimming. Methods: Annual top 150 rankings in the 16 female and 16 male
Olympic swimming events were downloaded for 1990 through 2011. Using
logistic regression, we modeled the proportion of athletes on each rank who
ever achieved the top rank over the 22 years. The modeled proportion was
closely approximated by 1/rank (e.g., rank=7 implies one athlete in 7
achieves top rank). Each country's performance score was then derived by
summing 1/rank of all its athletes in all 32 events. The measure's
convergent validity was assessed by correlating country scores with medal
counts at major competitions over the 22 years; the reliability was assessed
by calculating the year-to-year intraclass correlation coefficient (ICC) for
2008 through 2011. Measures of country performance were also derived using
the modeled proportion for achieving top rank, any top-three rank and for
winning gold at major competitions. Results: For scores based on 1/rank,
validity and reliability correlations were high: validity, 0.82 (90%
confidence limits, 0.79 to 0.85); ICC, 0.96 (0.94 to 0.97). The three
countries with the highest scores in 2011 were USA (a score of 41), China
(21) and Japan (16). Results were similar for the other measures.
Conclusion: We have successfully developed fair measures of country
performance that are more comprehensive than measures based on medal counts.
The inverse-rank measure is the most practical for the sport of swimming,
but all measures should be useful for any sport with world rankings.




------------------------------------

Post messages to sportscience@yahoogroups.com.  To (un)subscribe, send any 
message to sportscience-(un)subscr...@yahoogroups.com. View all messages at 
http://groups.yahoo.com/group/sportscience/.Yahoo! Groups Links

<*> To visit your group on the web, go to:
    http://groups.yahoo.com/group/sportscience/

<*> Your email settings:
    Individual Email | Traditional

<*> To change settings online go to:
    http://groups.yahoo.com/group/sportscience/join
    (Yahoo! ID required)

<*> To change settings via email:
    sportscience-dig...@yahoogroups.com 
    sportscience-fullfeatu...@yahoogroups.com

<*> To unsubscribe from this group, send an email to:
    sportscience-unsubscr...@yahoogroups.com

<*> Your use of Yahoo! Groups is subject to:
    http://docs.yahoo.com/info/terms/

Reply via email to