Editor’s note: Nik Werk is research manager at market research firm B2B International’s New York office.
It should be relatively straightforward to determine who won the Olympics but it’s not. All over the world, different news media have different ways of keeping score – some count gold medals, some count total medals, some allocate a value to the medals and some also take number of athletes or population into account – all giving completely different outcomes. In this article I will take a look at some of the common ways of keeping score.
- Counting
The most basic way would be to count the number of total medals or gold medals. It is uncontroversial, because there is no biased interpretation but also seems unfair, as the largest countries have a huge advantage. In this case, the top three are the U.S., China and U.K., depending on which type of medal we count.
- Weighting the medals
Some have tried to make this more sophisticated, like the New York Times, by weighing the medals on importance. They suggest that a gold medal is twice as important as a silver, which is twice as important as a bronze – the 4-2-1-point system. This is more controversial, because of the difficulty of assessing the relative importance of the three medals. However, it also seems fair in that it acknowledges the relative importance of medals, which the total number doesn’t. Again, the U.S., U.K. and China come out on top.
- Accounting for population
Because it takes the value of medals into account, this weighted medals method makes more sense than total number or just golds, and is therefore the method of counting we will use for the remainder of the article. However, our model is not complete, as total numbers give an unfair advantage to larger countries with more chances to win medals. For instance, Indonesia, a country of 257 million, won eight medals and would therefore be considered more successful than Bahrain who won six medals with a population of only 1.4m. We have to even the playing field.
To make up for this, we can plot the population per medal so as to not favor large countries. However, this seems to skew results too far towards the small countries. Could we really say that Grenada, who won one silver medal (worth two weighted medals) is really the winner of the Olympics?
- Participating athletes
One issue with population size is that it really doesn’t apply to athletics. A large country does not necessarily have a large number of athletes. The fairest measure, in our view, is the number of participating athletes. It seems unfair that a country whose athletes perform better on average falls behind one that just has more participants. For instance, Jamaica won 32 weighted medals with only 68 athletes, while New Zealand won just seven more with almost three times as many athletes.
This gives us what is arguably the fairest measure – weighted medals by number of participating athletes. It still seems to favor small countries too much.
- Context
This is the same graph as above – it just has the number of participating athletes plotted next to the weighted medals by number of athletes. While Tajikistan has a superior conversion, it still seems wrong. Surely it is more impressive to win 53 weighted medals per 100 athletes with a huge team of 554 participants, than getting a slightly better return with only seven athletes. Yet this model says Tajikistan won.
In searching for the fairest possible measure to calculate the winner of the Olympics, without biasing findings one way or the other, we still get what seems like a wrong result. Most people looking at this data would say that the U.S. performance is more impressive.
Models and formulas: keeping context in mind
This demonstrates a shortcoming of the formulas and models we often apply in market research. When we try to find a suitable, unbiased and accurate measure, we risk losing sight of the context. The “best” model we could come up with is suggesting that a country with 100 athletes, which takes home 99 weighted medals, is doing worse than a country which sends one athlete and takes home one weighted medal. Data analysts and market researchers plays a vital role in applying judgment and interpretation to the data. We cannot rely on models and formulas in isolation. Unless of course, you truly think Tajikistan won the Rio Olympics.
Data source: medalspercapita.com.