Roger,

:So you'll know, I'm 66, a retired USGS
:seismologist/programmer.

Yes, I'm aware of this.

I suppose you're also familiar with the history
of the tectonophysics division at the USGS?
Did you know somebody by the name of SCOT KRUEGER?

:I *DO* know what I'm talking about and the
:evaluation and probability is not subjective.

Well, as you are aware, I don't agree. Anybody
can make statistics say what they want them to
say depending on what their motive is. I only
says this so that you know that I know.

:Your objections are valid to some degree though.

Thank you for that.

:I use the largest mag. reported by NEIC (as do
:they).

I'm aware of that which is what Alan did as well.
Actually, the QED put out by the NEIC lists several
magnitudes and from more then one location.

:Magnitudes are not an educated guess;
:they are calculated from measured amplitudes on
:seismograms.

Then you know that based on what type of crust
the seismic waves travel through, the amplitude
will be altered before they are received. So,
for a specific location, the measurements are
measured accurately. But different locations
get different measurements. These measurements
are then averaged together. Which one are right?
Which ones are way out of line? What does this
do to the average?

:Near misses are for horeshoes and people wanting
:to look better than they really are.

Tell that to the people in India that lost their
lives living 70 or more kilometers away from the
epicenter of their recent quake.

I suppose that if somebody had issued a prediction
for an M6 +/- .5 within 25k of San Jose within 3
days and an M6.6 hits 40k from San Jose within 4
days you'd consider it a miss.

You also ignored my point that some deviation
from what is expected is acceptable up to a
certain point. Subjectivity disallows this
allowable feature.

:I have a program to locate the 3 best-fitting
:quakes but no one seemed interested.

Actually I am. It just didn't fit in with the
way Michael was formulating his chart.

:If you make a prediction, I judge it by what you
:say and a quake either fits or doesn't. And your
:prediction has a measureable probability. Not
:subjective, as in "I'm doing pretty good"

By sticking exactly with the prediction, you are
being subjective.

In my example above, I'd want to know how
accurate somebody is whether or not their
predictions are right on or not.

Say somebody had issued 3 separate predictions.
Each prediction was for a 6 day window.
Each prediction was for an M5.5 +/- .5.
One was issued in March, another in June, and
another in September.
The location for one was within 50K of point A.
The location for another is within 50k of point B.
The location for another is within 50k of point C.
A quake meeting all other parameters hits within
90k of point A.
A quake meeting all other parameters hits within
60k of point B.
A quake meeting all other parameters hits within
100k of point C.

From the above I'd be able to deduce that this
particular predictor was on to something, but
left a lot to be desired in getting their location
right. I'd better get myself ready if he/she calls
for anything anywhere near me.

Again, by letting this kind of thing go by the wayside,
your are being subjective in what is considered
a hit.

Dennis