Mary

I suspect he and the reporter crossed signals. The prediction is not wrong until tomorrow is over and no quake happens, which is about 99% likely now but not certain. He has already written and distributed a statement for reporters about his view of what to make of the failed prediction, but it hasn't failed quite yet.

And the data is not faulty or unreliable, the prediction just plain missed, which he said was a 50-50 chance from the start of the window. It may be that tweaking or overhauling his method might produce better results, but we won't know until better results can be documented.

I see the story has been picked up by the AP. What a nuisance. It is just a subtle difference in interpretation, but it's confusing, and why many scientists are reluctant to talk with the press.

John

• longer version of the story