That's the best they can formulate at the moment, and it is probably still be too ambitious. No predictions have done better. Correctly identifying a 40% chance of an M6 event on a particular patch of a fault in 3 days would be a smashing success in this business.

You disagree?