Roll & Rock

[ Roll & Rock ] [ Main Menu ]


  


21365


Date: April 17, 2020 at 10:06:07
From: Roger Hunter, [DNS_Address]
Subject: Shan evaluation


Hi all;

Here's my latest evaluation of Shan's "predictions"

In this effort I have skipped the probability
calculation as it tends to destroy expensive disk
drives.

I looked at all his predictions from Oct 20, 2019
to Jan 9, 2020, 272 in all.

Comparing them to the NEIC catalog of mag 4.5+
quakes
Shan had 124 hits and 148 misses.

A quake can satisfy only one prediction in this
program. Shan's predictions overlap, causing
multiple hits for a single quake.

Even this simple test shows how poorly he is doing.
124/148 is worse than coin toss results.

Roger


Responses:
[21366] [21396]


21366


Date: April 17, 2020 at 11:17:26
From: Roger Hunter, [DNS_Address]
Subject: Re: Shan evaluation


If anyone is interested I'll send a copy of the
program output for examination.

Roger


Responses:
[21396]


21396


Date: May 12, 2020 at 18:43:28
From: Roger Hunter, [DNS_Address]
Subject: Re: Shan evaluation


Another test involving 309 predictions shows the same
results; fewer hits than misses.

Shan needs to forget predictions and spend more time
with his family.

Roger


Responses:
None


[ Roll & Rock ] [ Main Menu ]

Generated by: TalkRec 1.17
    Last Updated: 30-Aug-2013 14:32:46, 80837 Bytes
    Author: Brian Steele