Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

Re: Olympic data

Ludwik Kowalski wrote:
Athletes are often ranked on the basis of experimental measurements of
time intervals. My impression is that differences between some outcomes
are too small to be meaningful.

1) The Olympics aren't science.

1a) They are entertainment. That's why they're called the
Olympic Games : grown-ups playing games.

Example: In some sports a ball that hits the line is in
bounds; in other sports it's out of bounds. There's nothing
scientific about it. It's arbitrary.

1b) There's also an element of nationalism ... which is IMHO
very far from science. Instead it exploits some of the basest
human emotions.

2) If this were science, we would do it differently.

A characteristic (some would say a defining characteristic)
of scientific results is that they are predictive. Usually
the simplest type of prediction is to predict reproducibility:
we predict that if somebody reproduces the experiment they
will get the same result.

The outcome of a close race is not reproducible.

Scientists, when faced with this sort of situation, would
not focus on the results of a single race, but would focus
on the statistical description of a large ensemble of races.
They would report the mean and standard deviation of each
racer's performance. This would add greatly to the scientific
character of the enterprise ... but alas would detract from
the entertainment character.


My suggestion: laugh it off. It's just a game. Not everything
in the world is scientific.