When you base a prediction on something else then science and it turns out to be too alarming, that's exaggerating. When science predicts something that turns

When you base a prediction on something else then science and it turns out to be too alarming, that's exaggerating. When science predicts something that turns out to be too alarming, that's the model being perfectible. What else do you suggest we should trust, if not science?