View Single Post
  #2   Report Post  
Old September 8th 03, 12:53 PM posted to uk.sci.weather
Dave Ludlow Dave Ludlow is offline
external usenet poster
 
First recorded activity by Weather-Banter: Jul 2003
Posts: 418
Default Opps how many bad forecasts can you make?

On Mon, 8 Sep 2003 08:53:43 +0000 (UTC), "nguk.."
wrote:

Rain for sunday nope , rain for sunday evening nope, how about rain for
sunday night
nope clear skies, lets make a forecast for rain on monday. What do i wake up
to
patchy high cloud and blue sky.

wow its a record!

Sometimes, they are their own worst enemies. I watched last night's
News 24 forecast for today carefully, sometime around midnight.

Initially, the presenter mentioned briefly that there was uncertainty
about the forecast for rain in the South. So far so good. He then went
on to talk about and show charts of rain covering Southern England and
ended up, several minutes later, by mentioning the miserable rainy day
we could all expect. If he'd again qualified this with a reminder
about the uncertainty, all would have been well. But he didn't.

The layman would have been left with an image of a miserable rainy
day... but *I* remembered what he said briefly at the start and
guessed it was a 50/50 situation. I guessed it may well stay dry until
well into the afternoon (at least) and so it has proved to be, so far.

This isn't a forecasting problem, it's one of presentation. Clearly,
THE PRESENTER should have given the confidence level at the start AND
at the end, stating the alternative scenario (e.g. the one I deduced).
Presentation presentation presentation. Will they ever get THE
PRESENTATION right should be the question. Grrrr.....

--
Dave
Fareham