See sequel here: http://statisticalideas.blogspot.com/2016/08/indias-olympic-triumph.html
This week the International Association of Athletics Federations made a historic decision to fully ban Russia's track and field athletes from competing in upcoming global competitions, including the Rio 2016 Summer Olympics. The International Olympic Committee meanwhile "fully supports" this decision targeted against the Russia Athletic Federation, which now puts further pressure on President Vladimir Putin to declare these decisions as "unfair". Cheating obviously makes the games unfair. But so too is the implementation of punishment, when it seems apparent that other nations who cry foul are surely dishonest too. One need to look no further than celebrated American cyclist and cancer activist, Lance Armstrong. Mr. Armstrong won 7 consecutive Le Tour de France races (beating standing records by four Europeans who have won 5 times each). Instead of questioning this extraordinary achievement as a statistician would, people all over the world quickly idolized Mr. Armstrong as an American role model! He was engaged to singer-songwriter Sheryl Crow, and received major sponsorships from Radio Shack and United States Postal Service. Only after all of these too-good-to-be-true attainments did we disgracefully come to terms with the true connotation of the title of his best-selling book (It's Not About the Bike). Let's take a closer look at the Olympic performance of every winter competition in history, and see if the host country's accomplishments should be considered too-good-to-be pure. For these cold Olympics, could we have had a chance for other host countries, such as Russia's Cold War adversary the United States, to have engaged in more short-sighted "Lance Armstrong" moments, or have had any of a number of other deceptive violations that have been previously overlooked by the broader public?
We begin by looking at each nation's total medal count score in each of the 22 Winter Olympics ever held, both as a non-host (in blue) and as a host (in red), if applicable. All raw data is freely available here. In the first games in (France 1924) 49 medals were awarded, but by the most recent games in (Russia 2014) the medal count had blossomed 6 fold, to 295. So each game's country medal allocation has been rescaled out of 295. Additionally nation adjustments were made to make them comparable across time without losing much impact since rarely did any of these countries host under a former break-off territory. As examples, the Soviet Union is now aggregated under Russia, and East and West Germany are consolidated under Germany.
We see that there has generally been a nearly 11 medal count gain for the host country in their performance while hosting, versus during the games on either side of their host games. For example The United States (U.S.) in 2002 won 34 medals (43 when rescaled), while in 1998 and 2006 the U.S. won rescaled scores averaging 24 between those two years. Was something mischievous afoot that allowed the U.S. to win 10 additional medals (nearly 40% more) in their host year of 2002?
In this final chart below, we show the timeline of host nations with the Winter Olympics (incidentally despite often writing about financial risk, Olympics was the topic of our very first blog article a few years ago). For each nation we show the hosting performance "improvement" (additional medals scored then, versus normal). It is clear from assimilating all of these probability ideas here that generally the home country bias has petered down over time. Though not as sharply as the chart would suggest merely as fewer nations competed early on (10 then versus 26 now) and hence the amplitude of that curve may be just slightly larger. But critically, the U.S. is consistently pushing the envelope on the way down that trend. While Russia and other countries have generally operated more fairly (though obviously imperfectly) throughout the Winter Olympics' history, versus other host nations.
This week the International Association of Athletics Federations made a historic decision to fully ban Russia's track and field athletes from competing in upcoming global competitions, including the Rio 2016 Summer Olympics. The International Olympic Committee meanwhile "fully supports" this decision targeted against the Russia Athletic Federation, which now puts further pressure on President Vladimir Putin to declare these decisions as "unfair". Cheating obviously makes the games unfair. But so too is the implementation of punishment, when it seems apparent that other nations who cry foul are surely dishonest too. One need to look no further than celebrated American cyclist and cancer activist, Lance Armstrong. Mr. Armstrong won 7 consecutive Le Tour de France races (beating standing records by four Europeans who have won 5 times each). Instead of questioning this extraordinary achievement as a statistician would, people all over the world quickly idolized Mr. Armstrong as an American role model! He was engaged to singer-songwriter Sheryl Crow, and received major sponsorships from Radio Shack and United States Postal Service. Only after all of these too-good-to-be-true attainments did we disgracefully come to terms with the true connotation of the title of his best-selling book (It's Not About the Bike). Let's take a closer look at the Olympic performance of every winter competition in history, and see if the host country's accomplishments should be considered too-good-to-be pure. For these cold Olympics, could we have had a chance for other host countries, such as Russia's Cold War adversary the United States, to have engaged in more short-sighted "Lance Armstrong" moments, or have had any of a number of other deceptive violations that have been previously overlooked by the broader public?
We begin by looking at each nation's total medal count score in each of the 22 Winter Olympics ever held, both as a non-host (in blue) and as a host (in red), if applicable. All raw data is freely available here. In the first games in (France 1924) 49 medals were awarded, but by the most recent games in (Russia 2014) the medal count had blossomed 6 fold, to 295. So each game's country medal allocation has been rescaled out of 295. Additionally nation adjustments were made to make them comparable across time without losing much impact since rarely did any of these countries host under a former break-off territory. As examples, the Soviet Union is now aggregated under Russia, and East and West Germany are consolidated under Germany.
We see that there has generally been a nearly 11 medal count gain for the host country in their performance while hosting, versus during the games on either side of their host games. For example The United States (U.S.) in 2002 won 34 medals (43 when rescaled), while in 1998 and 2006 the U.S. won rescaled scores averaging 24 between those two years. Was something mischievous afoot that allowed the U.S. to win 10 additional medals (nearly 40% more) in their host year of 2002?
And that's actually one of the least suspicious of the U.S. host game performances! In total there have been 11 host nations, and statistics decomposition allows us to see a rather shady pattern for Americans relative to Russians:
- U.S., and Norway hosted a total of 6 times. And on average earned 17 more medals during their host years.
- France, Germany, Japan, and Switzerland have hosted a total of 8 times. And on average a little less of an advantage, scoring 9 more medals during their host years.
- Austria, Canada, Italy, Yugoslavia, and lo and behold Russia have hosted a total of 8 times. And on average scored just 2 more medals during their host years.
This is a worrying pattern for some countries for sure, namely the U.S. But we'll explore the information further below so that we can see if there is other information we can learn about abnormal hosting nation advantages that augment the case for President Putin crying foul.
Now in the global heat map below we show each of the 22 host nations only, and show what the average medal "enhancement" has occurred during their hosting. On the most equitable side we have Yugoslavia, a country which generally earns 4 fewer medals while hosting, versus at about the same time when they are not hosting. Russia on the other hand generally earns 4 more medals when hosting, versus normal. But the U.S. (while contently finding fault with everyone else) somehow takes the gold, literally, with out-of-control home country bias of nearly 20 more medals when hosting, versus normal.
In this final chart below, we show the timeline of host nations with the Winter Olympics (incidentally despite often writing about financial risk, Olympics was the topic of our very first blog article a few years ago). For each nation we show the hosting performance "improvement" (additional medals scored then, versus normal). It is clear from assimilating all of these probability ideas here that generally the home country bias has petered down over time. Though not as sharply as the chart would suggest merely as fewer nations competed early on (10 then versus 26 now) and hence the amplitude of that curve may be just slightly larger. But critically, the U.S. is consistently pushing the envelope on the way down that trend. While Russia and other countries have generally operated more fairly (though obviously imperfectly) throughout the Winter Olympics' history, versus other host nations.
No comments:
Post a Comment