Preliminary Evaluation of Don's Novato, CA earthquake prediction:

It seems that Don's forecast earthquakes are coming earlier
in the window these days, or Don is waiting till the last minute
to post these prediction now. The last two have been much more
precise than earlier versions as well. Are you using a new technique
now, Don, or is this just random fluctuations in the precision of
your method. By the way, your forecast did not include latitude
or longitude for the Orland earthquake - Don, do you want to add those
to make it official?

Anyway, Don had said in regard to an earthquake near Novata, CA:

Novato, California Earthquake Forecast Plus Two
Posted by Don In Hollister on September 04, 2001 at 14:33:38:

Magnitude: 2.5-5.0
From: 09/05/2001PDT - To: 09/10/2001PDT
Location: Novato, California
Lat: 38.1N - Long: 122.5W - Range: 40Km

Hi All. Looks like there could be a second chance for Petra to feel a quake
in her area.

This one is associated with the Rodgers Creek fault at 38.1N/122.5W. This is
about 4 miles east of Novato, California. It should occur on 09/07/2001. The
window opens on 09/05/2001 and closes on 09/10/2001.

Magnitude range is 2.5-5.0Md. In the event of a geomagnetic storm between
now and then the max magnitude could be 6.0Md. Radius on this is 40Km.

"

Two forecasts for other areas followed.

Today the following earthquake was recorded by NCSN:

01/09/05 08:10:36 38.08N 122.41W 13.0 2.6Md A* 15 km NNE of San Rafael, CA

Source:
http://quake.wr.usgs.gov/cgi-bin/quake/quake.geo.berkeley.edu/name=quake

This event is within the magnitude and time window set by Don's
forecast which was posted about 10 hours before this earthquake occurred.
The forecast epicenter was less than 10 km from the located epicenter.
That is about as good as they get, Don, Grade A.

Probability of Occurrence:

Don posted some information on the probability of this event occurring
by chance (See #9259 above).
There have been 3 events since 1-1-2001 which have occurred within
the parameters that Don chose for his prediction (40 km of 38.1N 122.5W;
Ml 2.5-5.0). The background probability at the time of the prediction was
therefore 3/247 =0.012 (p = probability the event will occur on a given
date). Since the prediction encompasses a period of 6 days (Sep5-10)
the probability the event would occur randomly in his time window is
approximately 0.07 (odds of getting it right by random guess is
therefore about 1 chance in 14).