Mike;

Ok, let's go over this again.

I have a file of every syzygy from 1798 to 2012.

I have another file of perigees, not quite as big

Then there's the NEIC produced Centennial Catalog which has quakes from 1900 to 2007.

I wrote a program to read and remember all the syzygies. It compared all the perigees to all the syzygies and remembered the ones where perigee was within 12 hours either side of a syzygy. These are the supermoons.

Then it compared each quake of mag 7+ to the supermoons and made a list of the matches.

There were 6135 syzygies in the file but only 2672 were in the date range of the quakes.

I found 274 perigees of which 217 were within range.

Then there were 1729 mag 7+ quakes to consider.

Finally, of the 217 supermoons only 60 had a quake within 3 days either way, a 6 day window.

60/217 is 27.6% of the supermoons.

Now while all this was gong on, I sorted the 1729 quakes into an array of 6 day windows and then counted how many cells had a quake. That came out to be 22.1% of them so the odds on finding a quake within a 6 day window chosen at random is 22.1%

Still with me?

Finally I used a z-binomial test to see how significant it was to find 60 hits out of 217 tries when the odds are 0.221 for a hit and that came out to be about 0.03 for a single tail test. 0.03 is 97% better than chance.

Don't trust me. I expect you to check my procedure where possible.

Roger