hi
Time and again, it is debated that my prediction are below odds or random dates given by anyone.
This invitation was given earlier also . It was also tried to prove statistically ,that my prediction are below random dates,but software results are misleading.
If my predictions are below average or below random dates, it should be proved practically
One thing I know for sure is,if my predictions are below random dates,majority of people should be able to beat me (at least 7 out of 10 people )
Now here is challenge
1) Three groups of 10 people should volunteer for the work
2) All 30 people should give random dates for 6.2+ quakes,with + or - one day window period
3) The random dates should be for next six months
4) The dates should be posted on my blog,before 1st day of each month
5) I will post my dates (for that month)on 2nd day of each month
6)An independent evaluation will be done after 6 months
If at least 20 people ,from 3 groups,beat me in results , I will believe that, my predictions are below random dates or are just by fluke.
Amit
Time and again, it is debated that my prediction are below odds or random dates given by anyone.
This invitation was given earlier also . It was also tried to prove statistically ,that my prediction are below random dates,but software results are misleading.
If my predictions are below average or below random dates, it should be proved practically
One thing I know for sure is,if my predictions are below random dates,majority of people should be able to beat me (at least 7 out of 10 people )
Now here is challenge
1) Three groups of 10 people should volunteer for the work
2) All 30 people should give random dates for 6.2+ quakes,with + or - one day window period
3) The random dates should be for next six months
4) The dates should be posted on my blog,before 1st day of each month
5) I will post my dates (for that month)on 2nd day of each month
6)An independent evaluation will be done after 6 months
If at least 20 people ,from 3 groups,beat me in results , I will believe that, my predictions are below random dates or are just by fluke.
Amit
10 comments:
Amit;
I doubt you can get 30 people to cooperate for 6 months.
I'm wondering why you think 7 out of 10 must beat you.
I could predict twice a week and get most if not all the quakes.
You need to limit the number of predictions to avoid that.
Roger
Roger
Yes, You are right. There should be cap on maximum dates to be predicted
in a month. 4 dates in a month should be fixed ,otherwise complete month can go under window period.
Amit
Amit;
Four 3 day predictions in a month is about 1/3 odds.
How about this; I'll have my program pick 24 dates in the next 6 months, none to overlap. I'll make 30 sets, to be posted here. At the end we'll compare those to yours and the mag 6+ quakes and see who does best.
Fair?
Roger
Amit;
It's impossible.
In a typical month of 30 days, 4 windows take up 12 days leaving 18.
If the windows may not overlap there are only 19 total ways to have 4 different windows in the month so 30 different sets can't be found.
Roger
Roger
Why do we need all different windows? If we think that, different people are giving random dates without knowing what dates others are giving, some dates may match
Amit
Amit;
That's true and if you don't object I will continue.
OK?
Roger
ok
Roger
I would like you to post the dates for September to day itself (on this blog). After that, I will post my dates for September.
The reason for posting to day itself is, my first date is 1st September 2014 (6.6)and we are approaching window period within a day
Unfortunately the file you have mailed could not be opened. If you can post the dates on the blog (for September) it would be fine
Amit
Roger
September 2014 dates are posted
Amit
Just to let you know I post my predictions in the blog http://earthquake-predict.blogspot.gr/
the results there are obvious....
Best Wishes
Post a Comment