02-24-2015, 09:37 PM
6 months ago Amit proposed a test of his methods.
He would post 4 predictions each month for 6 months.
I would post 30 groups of 4 predictions chosen at random.
There would be dates when mag 6.1 quakes would happen +/- a day, so they were 3 day windows in all.
Results would be compared.
A couple of months only had one prediction due to unforseen circumstances and mine were adjusted to match.
At the end of the trial, my random predictions had a higher percentage of hits than his, showing that his method is no better than chance.
Amit disagrees. He contends I had to do much better to prove my point.
I don't thinks so.
Roger
He would post 4 predictions each month for 6 months.
I would post 30 groups of 4 predictions chosen at random.
There would be dates when mag 6.1 quakes would happen +/- a day, so they were 3 day windows in all.
Results would be compared.
A couple of months only had one prediction due to unforseen circumstances and mine were adjusted to match.
At the end of the trial, my random predictions had a higher percentage of hits than his, showing that his method is no better than chance.
Amit disagrees. He contends I had to do much better to prove my point.
I don't thinks so.
Roger