Petra,
As Mike points out, these predictions are VERY difficult to
evaluate as significant data is missing in regard to area.
Earthquakes in Northern Italy are nearly as common as in
California (in the lower magnitude ranges) with events in
the Ml 2.0-3.5 range nearly every day. Ml>4 are much rarer.
You might want to check the local Italy catalog listing at:

http://www.dister.unige.it/geofisica/

To see a weekly listing click on the red "Seismicity" on the
left margin, then go to the earthquake list by clicking on "A List" at
LIST of recent earthquakes

in the center of the page

Farther down in the center you will find a listing of
the latest 20 events under:
A LIST of the last 20 regional and local earthquakes recorded

Click on "A LIST" to see these events.

When I look at these lists, the largest event I see is
a Md 3.0 on May 12. None of these events are large enough
to qualify as fulfilling prediction #80 (see Canie's response above).

In regard to Southern Italy, apparently the magnitude range is
3.6-3.8. The event under Sicily that Canie points to is a Ml 4.8 -
not within this magnitude range. This event was probably felt
(barely) on the surface. 200 km depth is a long ways down. Even
the largest earthquakes at this depth do not do much damage
at the surface. It seems to me highly unlikely that any process
happening at 200 km depth will have surface precursors. It takes
electromagnetic signals years to travel 200 km through the mantle,
and heat take millions of years to travel this distance. Also
the Sicily epicenter is 301 km from the prediction (360 km from
the hypocenter to the prediction), that's a long ways off to be
predicting a 3.6-3.8 event. It's like you claiming a correct
match for your Mammoth Lakes prediction with a small event that occurred
south of Los Angeles.

Here are a few statistics on seismicity in these two areas:

In Southern Italy, the Italian catalog records 1868 events of Ml>=3.5
since 1-1-1900 (it is pretty complete throughout the entire century).
This includes 54 such events since 1-1-2000. This is about 19 such
events per year over the past 100 years or about 38 per year
over the past year and a half. On a given day, then, there is
about a 10% chance that an event will occur within 300 km of the epicenter
stated in the prediction. In 11 days (the length of the prediction),
having at least one event is about 10 times that, nearly a sure
thing. It is hard to see how such a prediction could often fail.

In Northern Italy, the catalog records 1121 earthquakes of Ml>=4.2
within 300 km of the given epicenter, 45 of which have occurred
since 1-1-2000 or about 11 per year (32 per year since 1-1-2000).
On the basis of the background seismicity for the century, the
probability of a single event on a given day is about 0.03 (30% in 11
days), but on the basis of current seismicity it is about 0.06 per
day (about 66% in 11 days). That means that 2 out of 3 times
the prediction should prove right even if the signals had no
value whatsoever.

These statistics assume a very large area (300 km) about the
predicted epicenter, and probabilities would be much lower
if the predicted regions were considerably smaller. This is
why it is difficult to evalutate predictions without all the
data requested by the prediction form.