When the weather forecaster says that “there is a twenty per cent chance of rain tomorrow” do you put on a mackintosh? Or do you assume that you’ll be ok without a coat, because a twenty per cent chance seems pretty low? Many of us view the weather forecast with scepticism but still complain vociferously when it seems as though the forecaster has got it wrong. In fact modern forecasts are remarkably accurate. If conditions don’t measure up to our expectations it’s often because of geographical variability or timing, factors that are difficult to forecast with precision. But we also don’t seem to be very good at understanding uncertainty. We expect the forecast to be just that, a definite prediction of the future. Of course, in reality, it can only be modelled on what is happening at the moment. Poor Michael Fish will always be remembered as the forecaster who denied that there was going to be a hurricane, even though he did say that it was going to be windy. When serious damage, injury and even death ensue, questions about “fault” emerge, of course. We have seen that very clearly in Italy this week. Six scientists and a government official have been found guilty of manslaughter and imprisoned because they failed to foresee a major earthquake that struck the town of l’Aquila in 2009. This is very shocking to fellow scientists who understand all too well how difficult it is to predict earthquakes with accuracy. Such predictions always carry a large margin of uncertainty. One effect of the verdict is likely to be that researchers will be unwilling to advise governments or to take part in public communications exercises in the future. But were the scientists perhaps to blame in the way they communicated? It may be that at the public meeting where they talked about risk they did sound too reassuring. Or, alternatively, was the audience predisposed to hear the reassuring words, rather than the caveats that the scientists insist were included? Just as we would rather think that a twenty per cent chance of rain means we don’t need a mackintosh after all. But in reality it would probably be wise to pick up an umbrella, just in case, and think ourselves lucky that we don't have to worry about earthquakes in the UK.