Doper duper

There are many myths in sports. One of the enduring ones is that the best athletes win on their talent; the dopers are the lesser ones who can't rely completely on their natural ability.

This myth persists despite Lance Armstrong, Jose Canseco and numerous other cases. News has emerged that the Tokyo marathon gold medalist & 2025 world champion has been provisionally suspended for doping (link). Oops.

The process has begun, and we'll soon learn if he has an "explanation" for the abnormality.

This story gives me a reason to review the key conclusions of Chapter 4 of Numbers Rule Your World (link).

This book came out before Lance Armstrong confessed after being ratted out by a former teammate. At the time, the establishment, including the statistics experts, was worried about false positive results in anti-doping tests. This line of thinking borrows heavily from analyses of other types of diagnostic tests, such as mammograms.

The only problem is that the analytical result was not transferable. In fact, I explained in that chapter why the opposite conclusion should have been drawn: that is to say, anti-doping testing is plagued with false negatives, not false positives.

The reasoning is simple. Can you think of any athlete from the 2024 Paris Olympics who has failed a drug test? According to this page, over 10,000 people competed in Paris. If just 1% of these athletes doped, there should have been at least 100 true positives. Add in the false positives that all these experts were worried about. Let's assume a 1% false positive rate. We apply this rate to the non-dopers, say 9,900; that should have made another 99 positive results. Where were these 199 positive findings? (How long was the list of Paris athletes you've learned to have failed a drug test?)

You can hear a pin drop.

Either you believe almost no athlete is doping, or you accept the uncomfortable truth that most dopers are getting away with doping. Most dopers have duped us.

That's just the beginning of the argument. All of the above assume that anti-doping science is perfect. The scientists know every possible way of doping, and they have an accurate detector for each doping method.

For every method for which they don't know of, or they don't have a test for, we have another pile of "false negatives". They aren't called false negatives because there are no test results to look at. They are false negatives if we define the term broadly to mean doping that goes undetected.

All this and more are covered in Chapter 4 of Numbers Rule Your World (link).