This earthquake was actually expected. As the data below indicate, an unusually strong EM Signal was detected after the first earthquake in that area occurred. All the times are UTC.

2013/11/17 09:04:55 60.30S 46.36W 10 7.8 Scotia Sea
2013/11/16 03:34:31 60.21S 47.11W 10 6.8 Scotia Sea

2013/11/14 03:56:00 Very Strong EM Signal
2013/11/13 23:45:48 60.26S 47.20W 10 6.1 Scotia Sea

And as I have said in the past, when these strong EM Signals are detected then there is a fairly high probability that there will be a powerful earthquake within the next 5 days.

It appears to me that the first earthquake caused strain to transfer in the fault zone. Then strain buildup at some other location caused the EM Signal to be generated. Then the second earthquake occurred and more strain was transferred. Then finally the main shock occurred.

The main one was if I understand things correctly, about 35 times more powerful than the second one.

I have checked my recent data. And although my program could tell that a powerful earthquake was getting ready to occur it would not have been able to identify its location based on just recent EM Signal data.

This was an especially complex situation because there appears to have been a progression of events rather than just one powerful earthquake. The 7.8 earthquake could be considered a somewhat random event as far as earthquake triggering processes are concerned. So perhaps just that one recent strong EM Signal was actually linked with it. And it could not point to the right location because the earthquake triggering time was so random.