COSTAS: The Weighted Olympic Medal Count

By David Roher

Canada had a low raw medal count, but COSTAS gives them a resounding victory.

Disclaimer: This medal weighting system does not reflect the views or endorsement of any NBC broadcasters, and only reflects the view of the author.

U-S-A! U-S-A! According to the medal count that everyone’s using, we’re #1, unless you’re going by number of gold medals, in which case Canada is on top, or the EU, if they get all technical again.

At the risk of knocking the US off its pedestal, I wanted to develop a weighted medal count that not only adjusted for the importance of gold over silver over bronze, but also for the relative importance of one event over another. I also wanted its abbreviation to be COSTAS. I admit that the latter was the first requirement I thought of.

Thus was born the Congruent Olympic System for the Tabulation of Accolade Statistics. Here are its main components:

  • Gold, Silver, and Bronze are weighted four, two, and one, respectively.
  • Each of the 15 sports of the Winter Olympiad can be worth a total of roughly 17 points: if a sport features 5 events, its medals are worth twice as much as a sport that features 10 events. The 17 figure is so that the total medals in COSTAS add up to the real total.

The first point is isn’t really all that important; I can’t justify 4-2-1 over 3-2-1 and 5-3-1. But it’s the second one that I imagine will draw the most criticism. Surely, we value a speed skating medal over a skeleton medal. But value is subjective – every participant in the Olympics is going to have a different view of what sport matters over the other. By weighting each sport equally, COSTAS has no bias.

Without further ado, the results:

Country Unweighted COSTAS Diff.
Canada 26 46 20
USA 37 38 1
Germany 30 34 4
Austria 16 19 3
Norway 23 16 -7
Switzerland 9 13 4
China 11 12 1
South Korea 14 11 -3
Sweden 11 10 -1
Russia 15 10 -5
France 11 8 -3
Netherlands 8 5 -3
Great Britain 1 5 4
Poland 6 5 -1
Latvia 2 4 2
Australia 3 4 1
Finland 5 4 -1
Italy 5 3 -2
Japan 5 3 -2
Czech Republic 6 2 -4
Belarus 3 2 -1
Slovakia 3 2 -1
Croatia 3 1 -2
Slovenia 3 1 -2
Kazakhstan 1 0 -1
Estonia 1 0 -1

Canada’s balanced and gold-heavy results gave them an incredible 20 medal boost in the COSTAS medal totals. The US and Germany both got small boosts, but dropped to second and third respectively. Norway took the biggest hit in the medal count, dropping 7, while the Czech Republic dropped the greatest amount in the standings. Great Britain got the biggest standings jump from their Skeleton prowess.

In the end, no one system is better than another. But at the very least, I hope COSTAS provides a look at Vancouver that you might not have otherwise seen.

About these ads
This entry was posted in Olympics and tagged , , , . Bookmark the permalink.

15 Responses to COSTAS: The Weighted Olympic Medal Count

  1. alexander says:

    What is defined as a different sport? Obviously Hockey is a sport different from bobsled, but is skeleton different than bobsled? What are the different skating classifications?

    Perhaps most intriguing would be to multiply the weighted medals (4-2-1, 3-2-1, w/e) against the worldwide viewership!

    Of course that is harder to do.

    • David Roher says:

      I used the sport definitions as defined by the IOC. There are some weird distinctions – short track is different from speed skating, for instance. But it’s not my place to make distinctions.

      The worldwide viewership would be really interesting. The problem is the availability of the games in different countries relative to others and the populations of different countries. But yeah, I would love to see it.

  2. Bryan says:

    Love this analysis.

    Any chance you could provide comparison results between objective sports and subjective (judged) sports?

    And perhaps within objective sports you could compare “all other objective sports” vs. “sports where competitors are essentially hurling themselves down a hill.”

    OK, the second request is a bit sarcastic. :)

    Thanks for this great blog — keep up the good work!

    • David Roher says:

      Thanks, Bryan. Yahoo’s Fourth Place Medal has done something to that effect:

      http://sports.yahoo.com/olympics/vancouver/blog/fourth_place_medal/post/The-REAL-medal-count-Germany-ahead-of-U-S-?urn=oly,220910

      Dunno if they updated it after that post. My problem with distinguishing judged from non-judged sports is that all sports have referees.

      • Bryan says:

        Thanks for the link — that’s close enough.

        Sure all sports have referees, but in objective sports, the referees are there to make sure rules are followed so an objective winner is fairly determined. In subjective sports, the judges are there explicitly to decide the winner based (at least in part) on subjective criteria.

        It is possible for basketball teams to compete without referees and determine a winner. The same isn’t true for figure skating — given the current system, judges are required to determine a winner.

        I’m not someone who thinks of one type of competition as “more real” than another (whatever that means!), but I do think there is a distinction.

  3. Jimmy S says:

    The Winter Olympic medal tally heavily favours wealthy countries where it snows, so any adjustment should take both GDP & yearly snowfall into consideration.

    Not sure if snowfall measure is cumulative, per capita, or per sq.mile

    • Scott says:

      Snowfall is measured the same way rain is – it would be an average taken over all weather stations in a country.

      Which is interesting because in a country like say Kenya – we’d find no snow.. but yet it snows in Kilimanjaro. I’d throw snowfall measure out – but the idea leads to perhaps a mean temperature in “Winter” as a variable.

      In contrast though the Summer Olympics heavily favours wealthy countries where it DOESN’T snow.

      But an adjustment for population is needed I think. When Canada finishes 3rd with a population esssentially 1/10th of the USA… it is impressive. You’d expect some correlation to the medal count.

      Perhaps there is a way to find the numbers of people registered in sport organizations in each country and adjust accordingly.

      • Jimmy S says:

        Raw population is useless. The countries with success which punched above their weight are Austria, Norway, Sweden Switzerland. Two Alpine, Two Scandanavian. Better to use number of people with immediate access to rinks, pistes, and snowfields.

        So for Mt Kilimanjaro, unless there’s a population nearby, its snow doesn’t get counted. Nor do the populations of Florida and California (a 6 hour drive is not immediate)

        Also, disagree with COSTAS. If medals are awarded by the IOC, then the event is as equally important as the others. I don’t see that you’ve provided sufficient reason not to count events equally.

        • David Roher says:

          I just think that the number of medals usually has everything to do with the nature of the sport and nothing to do with the importance of it. Hockey is best suited as a round robin followed by an eight-team tournament. Biathlon is best suited as a series of different types of events at different distances.

          Is the level of competition 5x greater in biathlon? The quality of the athletes? If you reformatted the sport so that it consisted as a round robin followed by a tourney, would that that make it 5x less important? No.

          But I don’t think COSTAS would work in the summer, for those of you who think it works here. The IOC groups all track&field events into “athletics,” which is heavily problematic. When I do it for London (or retroactively for Beijing), I’m going to weight each sport according to the number of athletes that participate in it.

  4. Jimmy S says:

    See your point David – what about Ski Jump which doesn’t have a chick’s comp? COSTAS artificially inflates the value of the sports’ medals as there should (in fairness) be another set of 3 given out to the ladies.

    Also on Summer: The number of teams for football (soccer) is 16 for blokes and 12 for the sheilas due to pre-olympic qualifiers. Is this 28 “competitors”, 28 x 11 players on a team, 28 x 22 or 23 players in a squad, 1 competitor per team that attempts to qualify??

    • David Roher says:

      Good point about Ski Jump – definitely a flaw in COSTAS. Fortunately, I don’t think there are that many examples where there isn’t a comparable womens’ event.

      For athlete count, I don’t know nearly enough about the Olympics or those particular sports to know for sure. But if you go to London 2012′s official site, they have an athlete count for each sport. Since I last replied, I’ve been thinking and athlete count might be a good fix – it solves your Ski Jump problem, for example.

      • Scott says:

        Athlete count is a good and import factor here to put into each event IMO.

        For example, if Women’s Ski Jump had been included the field would have been incredibly small – although my heart goes out to them the lawsuits in Canadian courts by these women were a waste of money as a nation cannot force the hand of the IOC here.

        How do you adjust for team sports though? Will it be number of teams – does a relay with 4 athletes then have less weight than a full hockey team?

  5. Klaus Fix says:

    Here is the official version of who wins the Olympics based on country. “The convention used by the International Olympic Committee is to sort by the number of gold medals the athletes from a country have earned. In the event of a tie in the number of gold medals, the number of silver medals is taken into consideration, and then the number of bronze medals. If two countries have an equal number of gold, silver, and bronze medals, they are ordered in the table alphabetically by their IOC country code.” This is from the IOC and posted on Wikipedia. see these two websites; http://en.wikipedia.org/wiki/2010_Winter_Olympics_medal_table and this Reuters article at; http://blogs.reuters.com/china/2008/08/19/whos-top-of-the-medals-table/

  6. Pingback: COSTAS: The Weighted Olympic Medal Count | The Harvard College Sports Analysis Collective

  7. Pingback: COSTAS: the weighted Olympic medal count | BroLympics

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s