Updating the Results for the November Earthquake Forecast
The astrological model I am experimenting with seems to be holding up quite well, though it is very early and still a small sample. Nonetheless, there is a 15-day period that can be examined and the probabilities (p) of the results can be measured.
A real test will be the 12th/13th forecast, when the model shows maximum stress.
11/14 update: The 12th & 13th forecast did not materialize at all, quite the opposite in fact. Daily EQ activity dropped from an average of 40 to 50 total EQ’s per day to 20 – 25 on the 12th & 13th. Worth noting that.
Forecast Posted 10/29, to begin 10/30:
4 hits (eq’s) out of 6 trials (days) @ (f) .366 = (p).108 or about 11 chance out of 100
8 hits (no-eq) out of 9 trials (days) @ (f).634 = (p).086 or about 9 chance out of 100
combined (joint occurrence) 12 hits in 15 days = .108 x .086 = .009 of about 9 chances out of 1,000.
Despite the 3 misses on the 11th, 12th & 13th, this forecast model remains strong with a binomial probability of 9 chances out of 1,000 and a good correlation ratio of 60%.
Please observe how closely the red lines (earthquakes) and blue lines (astro-forecast) intersect and follow each other.
As a comparison, a control group of 12 dates was randomly generated and graphed.
1 hits (eq’s) out of 6 trials (days) @ (f) .366 = (p).225 or about 23 chance out of 100
5 hits (no-eq) out of 9 trials (days) @ (f).634 = (p).232 or about 23 chance out of 100
combined (joint occurrence) 6 hits in 15 days = .225 x .232 = .052 of about 5 chances out of 100.
Please observe that the red lines (earthquakes) and blue lines (random-forecast) barely intersect and don’t follow each other for a negative correlation of -.29