Saturday, January 28, 2006

20 years after Challenger

It has been twenty years since the Challenger disaster. At that time, our teachers would tell us how much the Americans were ahead of the Soviet bloc because of the space shuttle.

Czechoslovakia became the 3rd country in the world whose citizen visited outer space flight back in 1977. Today, Vladimír Remek is a deputy of the European Parliament for the Communist Party. The first name Vladimír is after Mr. Lenin while his surname Remek stands for Rychle-Eeeeee-Mluvící-Eeeeee-Kosmonaut i.e. Quickly-Errrrrr-Speaking-Errrrrr-Astronaut. You can see that Mr. Remek is translated as Mr. Qesea.

Twenty years later, I personally no longer believe that the space shuttle is such an incredible technology. Feynman's conclusions about the space shuttle are here:

The main point of Feynman's observations is the same as the point of my texts about the Bayesian probability: the probabilities only have a scientific meaning if they can be determined or at least interpreted in a frequentist fashion, and they can only be trusted if the relevant experiments have actually been tried sufficiently many times to give us the result with the desired accuracy.

More concretely, the management's estimates of a mission failure - around 1 in 100,000 of Bayesian probability - were scientifically nonsensical because no one could have determined such a low probability of failure using rational methods, especially because of the top-down approach to the space shuttle design where the individual components can't be tested and evaluated separately.

Today, of course, we know that the probability of space shuttle failure was definitely much closer to 1 in 100 than to 1 in 100,000 and probably even higher. But even without the sad knowledge we have today, the management's Bayesian estimates were always a complete bogus.