The original article says that estimated probability of observed configuration of objects appearing by random chance without some unknown, common cause is 3.8 s

The original article says that estimated probability of observed configuration of objects appearing by random chance without some unknown, common cause is 3.8 sigma (0.016%). For comparison (in an apples-to-oranges kind of a way), the recently falsified diphoton anomaly detected by CERN that made news in 2015 was 2.9 sigma or about 25 times more likely to appear by chance. Although this is based on just 6 detected objects, it's a strong indication that some yet-unknown mechanism is in play.