Tony
Here is the complete list of my 2009 prediction ,with 3 day window and quakes considered only those which are 6+
Prediction date------date(-actual quakes)
1)2.1.09-----------3.1(7.7),and (6.6)
2)4.1.09----------3.1.09(7.4)
3)11.1.09--------miss
4)12.1.09---------13.1(6.0)
5)15.1.09---------15.1(6.7 and 7.4)
6)18.1.09---------18.1(6.4)19.1(6.6 and 6.0)
7)25/26.1.09--------miss
8)9.2.09--------9.2 (6.0)
9)4.3.09---------miss
10)6.3.09--------6.3.09(6.5)
11)15.4.09------15.4(6.3),16.4(6.7)
12)22.4.09------21.4(6.2)
13)9.5.09-------10.5(6.1)
14)17.5.09------16.5(6.5)
15)18.5.09------miss
16)24.5.09-------24.5(6.0)
17)6.6.09--------5.6(6.4),6.6(6.0)
18)13.6.09------12.6(6.00,14.6(6.0)
1916.6.09-------16.6(6.1)
20)20.6.09-----miss
21)22.6.09-----13.6(6.7)
22)6.9.09------7.9(6.2 and 6.0)
23)3.10.09-----2.10(6.1 and 6.0),3.10(6.1),4.10(6.6)
24)4.10.09----4.10(6.6)
25)11.10.09-----10.10(6.6)11.10(6.0),12.10(6.2)
26)13.10.09---13.10(6.5,6.0,6.4),14.10(6.3 and 6.3)
27)18.10.09----19.10(6.0)
28)20.10.09----miss
29)24.12.09---23.12(6.0),24.12(6.3)
3026.12.09-----26.12(6.1)
Tony now I would like yo to evaluate the predictions
regards
Amit
25 comments:
Amit;
Here's my evaluation. Let's see if Tony agrees.
There are 122 windows in 2009 of 3 days each. The odds on a hit are 0.648 based on the number of windows containing mag 6+ quakes.
There were 113 quakes of that size in 2009.
You had 29 predictions (I must have missed one) and 22 hits but since 18.8 hits were expected that is not at all significant.
Roger
Hi Amit, For >6 quakes the probability to get it right randomly in a 3 day window is near certainty.
So I am sorry this is not a good way. You must compare >7R and perhaps 2 day window or even 1 day window to be noticed.
I am not as critical as Roger on your method, simply because you have a theory which is easy to use. But the stats show that this is not all and there is more to it which we dont know. Nice to predict SOME quakes but randomly you can do better. You need to do better than 20% in >7R 3 day window. Check it out for 2008 and 2009.
As Roger counted 133 quakes in 122 windows the chance of finding a 6+ quake in any window is 113/122=92%
That to me is a certainty.
Your theory predicts only 30 quake dates, i.e. only 30 3 day windows where earthquake will occur. Out of those 30 only 26 are hits. If you simply divide 26/30= 0.86, and you say 86% success that is wrong of course. Simply because your prediction set is not the same as the set which we have had for 2009 which was 113 quakes, you should have predicted many more dates than only 30. Now what is interesting is that you predicted 30 dates i.e. 26.5% of the quakes which occured, at a success rate of 86%. Well, with chance only I get 92% success. So this is not the whole story.
If you focus ONLY on >7R quakes in 2009 we had 13 dates for those occuring. I know some are double. If the distribution were random then 13/122 = 10% was the prob of a hit randomly. You only predicted 3 quakes over 7R in 2009,butehad 13.
Tony/Roger
How if I take the window as one day ,which means only the date for which it is predicted
I have 13 date hits out of 30 predicted dates
the odds (with 3 days window)of hits as Roger says is 64.8% and as you say it is 92%.I do not know which one is correct
Nevertheless for one day window it is 113/365 is 30%.
30% of 30 is 9. Thus any date hits above 9 is more than average
Roger and Tony are requsted to comment
regards
Amit
i think so
I think so but you dont have that many
you do actually but your quake set is too small. I agree that for those few which you predict you have a decent rate. But this is not the whole story as we get many more quakes.
The most important aspect of your theory is to show that the occurance of the quakes is not probabilistic not random but predictable. What you say is that at least a smaller set from the total set which occurs in the world is predictable and NOT by chance. That is interesting to me.
Tony
I must admit that,I restrict myself while publishing the dates.
The theory is unacceptable to many as it is against the set norms of plate tectonics.Hence ,I become very cautious and publish the dates which are very much sure to me.Missing the dates is unacceptable atleast at this stage.
regards
Amit
I have Astrology as a hobby. So you will not get criticised by me on using Astrology or the fart of a fly to predict quakes. I am not bothered of ANY method provided you predictably get things right. I think your method so far - and you are right to say it is early days and it is not fair to kill it because you only get a small set right - is very interesting to me.
I would focus on 7+ and get those right 90%. That would be interesting to me.
Here is something to think. In 2009 we had ONE and only ONE quake of 8+.
This was on 29th Sept 2009. Mercury station. Question. WHY? Why was it not Saturn or Jupiter?
Mr Tony;
113/122 is not correct. It assumes only 1 quake per window.
You have to count the number of windows containing quakes and divide that by 122.
roger
Roger
Here is answer
Quake was on 29.09.2009 at 1748 UTC
ie 2318 IST
please note Jupiter was changing direction on 13th October 2010.
Also please note on 29th September exactly at the time of quake ,Moon coincides with Jupiter.
When Jupiter is either accelerating or retarding (+ or - 15 days)momentum of magma changes
Now ,you must believe the theory
regards
Amit
Roger
Besides the Jupiter is closer ie 4.315 AU (less than 4.5 is very potential )
Note the dates on which quakes of 6+ occurred in October 2009 is as high as 21 dates.
Moreover total number of quakes (6+) during 29th Oct 2009 to 29th November 2009 is whooping 34 as against 15 average per month
These all due to Jupiter changing direction as well as closer to earth position
regards
AMit
sorry , above 34 quakes were during 29th September to 29th October 2009.
Amit
oopps-- It seems I am in hurry to go to sleep. Yet another typographical error.
In first answer to Rogers question
Jupiter changing the direction on
13th October 2009 and not 2010.
sorry again
Amit
Hello Gentlemen,
I discovered this blog searching for EQ predictors, and became intrigued by the conversation. I also am an astrolger, and very recently began applying a research model I've been working on to EQ data, and am curious as to my results.
Amit, I am looking around for a list of your predictions for June or July to compare. Is there a post?
Also, I was wondering if Tony & Roger would look at my predictions/ results & see how they hold up statistically, but I won't post my web addy unless you give permission.
Amit;
The October 13, 2009 was one of your hits already.
You're still not above chance.
Roger
Sdelapp
Thanks for visiting my blog.
You are most welcome to post your bog address on my blog for everyone who is interested.
please see my blog dated 10th January 2010(go for older posts )for probable dates till December 2010
Regards
Amit
Roger
It is unfair when you do not take in to account my multiple hits in window period and at the same time you take all quakes (including multiples quakes in a window period) while calculating averages.
You should either take in to account for both cases or do not consider in both cases
You have seen how Jupiter changing direction and closer increases the frequency of major quakes
regards
Amit
Roger
I would love to see a month(+ or - 15 days of Jupiter changing direction) as well as closer,,and at the same time having quakes (6+) less than USGS averages
Regards
Amit
Ah, great... I'm Steve DeLapp and my web addy is AstroWeather.com if you click on the EarthQuake tab you'll see my forecast /results. I am wondering if my method is useful, or just luck.
Steve
Amit;
Ok, give me a list of dates for Jupiter changing directions and it's apogees and I'll see what I can do.
Roger
Well Congratulations Amid.. I found your 2010 list in time to see a direct hit of 7.4 today. Hope it wasn't to close to home.
Steve
Post a Comment