Richard Dawkins quotes David Hume (the 18th century Scottish philosopher) in The God Delusion (2006, p. 91) concerning the likelihood of miracles:
“No testimony is sufficient to establish a miracle, unless the testimony be of such a kind, that its falsehood would be more miraculous than the fact which it endeavours to establish.”
As Dawkins goes on to argue in reference to an event at Fatima in 1917, where thousands of pilgrims supposedly saw the sun ‘crashing down’:
“It may seem improbable that seventy thousand people could simultaneously be deluded, or could simultaneously collude in a mass lie. […] Or that they all simultaneously saw a mirage [. …] But any of those apparent improbabilities is far more probable than the alternative: that the Earth was suddenly yanked sideways in its orbit, and the solar system destroyed, with nobody outside Fatima noticing. I mean, Portugal is not that isolated.”
This argument (and David Hume’s) seems to rest on a misunderstanding of what a miracle is. A miracle is not a highly unlikely event. Rather, a miracle is an event that seems highly unlikely if certain conventional methods of explanation are used. In the case of Fatima above, the obvious answer to Dawkins’ response is: there is probably something else going on, which is likely. For example, when thousands of people saw the sun ‘crashing down’, it did not involve a literal movement of the sun as Dawkins interprets any ‘miracle’ to have to involve. There is some other explanation, which does not involve very low probabilities.
Indeed, this sort of explanation is probably the explanation that people would accept once it is pointed out (whatever it might be).
For more on a way to define what miracles are, see here.