| Roll & Rock
|
[
Roll & Rock ] [ Main Menu ] |
|
|
|
21952 |
|
|
Date: June 23, 2021 at 12:52:25
From: Skywise, [DNS_Address]
Subject: It's probably probability time... |
|
|
Time to add probability calculations to my program.
Been reviewing all those discussions/arguments with
Roger about Jones probability. Seems I was never quite
convinced Jones p is correct. :)
I have a formula I am pretty certain is more accurate
but right now I'm not sure how I derived it.
Ughhh... Lesson. Must make more notes.
Brian
|
|
|
|
|
Responses:
[21953] [21954] |
|
21953 |
|
|
Date: June 23, 2021 at 15:40:00
From: rh5919899, [DNS_Address]
Subject: Re: It's probably probability time... |
|
|
Skywise;
It's not that hard. Odds on anything are number of ways it can happen divided by total number of possible ways.
Chances of rolling a 6 with a single die are 1/6 because a die has 6 sides and only one is 6.
With quakes the possible hits are the total number of times such a quake has happened in the past divided by the number of times it could have happened.
It's more difficult because you have more variables to deal with; location, mag, window duration, etc.
Roger
|
|
|
|
|
Responses:
[21954] |
|
21954 |
|
|
Date: June 23, 2021 at 16:59:19
From: Skywise, [DNS_Address]
Subject: Re: It's probably probability time... |
|
|
Oh, I understand it. I understand the Jones method.
I just wasn't convinced it was the right method. Close.
But not exact. (and I'm a bit OCD about that)
I don't know if you recall, but my doubt revolves around
the actual distribution of the quakes affecting the
resulting Jones probability.
I understand the idea is to eliminate the affects of
clustering. But I don't think Jones does it correctly.
I could be wrong, but I have yet to convince myself of
that.
I'm reviewing the work I did before. It seems I hadn't
quite found a solution. But it's still seems Jones is
not quite right.
Brian
|
|
|
|
|
Responses:
None |
|
[
Roll & Rock ] [ Main Menu ] |