Felly sgrifennodd
Norman Lynagh :
If, over a representative sample of 2/3 probability forecasts the actual
event occurred 3 times out of 3 on average then the forecasting would be
considered fairly inaccurate.

Yes Norman - I certainly wasn't directing my comment at you. You are right
of course, and a representative sample would need far more than just three
members to have any statistical validity.

But I'm sure if someone really wanted to they could trawl through the
archives and find enough samples.

The media, however, would probably think the Met Office were wonderful in
such circumstances. Until the general public understand some basic
statistics (which I doubt will ever happen given that clearly even most of
the media don't, and are happy to keep the public in the dark), we're doomed
to have this sort of discussion all too frequently.

Adrian
--
Adrian Shaw ais@
Adran Cyfrifiadureg, Prifysgol Cymru, aber.
Aberystwyth, Ceredigion, Cymru ac.
http://users.aber.ac.uk/ais uk