I loaded all the pointing data from yesterday's pointing test. It is clear that the EVLA antennas are 'wiggling', just as the VLA ones are. So the problem lies in the collection of the data, or the dissemination of the results.