The NRC (Nuclear RegulatoryAgency) Probabilistic Risk Assessment gives the precautionary perspective, which is generally a priori, not manifested or empirical. The risk is stated to be at 1:1,000,000 chance of a nuclear accident. At the moment, there are 432reactors in the world. If we do the math, the probability of an accident is one every 2,500 years or the entire history of Western civilization back to the Greeks. Three reactor meltdowns at one time, that would be 1 million cubed giving 18 zeros; a major problem. I would argue that perhaps since it was in one place there may be a problem with the statistics of time and place, but only slight changes to 18 zeros. Prediction is one thing, and empirical is another. The evidence now shows that instead of a nuclear accident every 2,500years we could expect one every seven years, if we look over the past 35 years there have been 5 meltdowns, or one every 7 years. There were meltdowns before that, but this is giving the technology the benefit of the doubt. If we were to say there were five meltdowns over 35 years, that’s bad enough, but if we say there were three meltdowns in one month, or..over a few days at Daiichi, then the picture looks even worse.