Bayes’ Rule and Athletes Doping
Years ago, scientists wanted a sure-fire method of catching professional athletes who use doping to enhance their performance. However, scientists are not always able to easily detect these drugs in a given athlete’s system. Something scientists can watch out for are certain effects that performance enhancing drugs cause in the human body. Using the knowledge of these effects that the drugs can cause, and Bayes’ Rule, a technique was developed to indicate if athletes with certain blood levels were using performance enhancing drugs. This technique tests the probability of any given athlete using these drugs given their blood test (it analyzes athletes levels of hemoglobin in their blood, among other things).
Bayes’ Rule is a good, accurate convention overall for raising red flags at certain athletes’ blood levels. However, there are certain things that Bayes’ Rule simply cannot take into account, and this is why it cannot solely have 100 percent accuracy. The study tested a group of swimmers that had been training at altitude, and this naturally caused the swimmers’ bodies to produce more hemoglobin. The study incorrectly accused many of the swimmers to have been using performance enhancing drugs. Similarly, women being pregnant, and people having varying hydration levels can affect the results of this test. Therefore, experts examine these results after the test is given, and before athletes are immediately accused of doping.
http://www.newyorker.com/news/sporting-scene/using-math-to-catch-athletes-who-dope