In article ,
Will Hand writes:
Remember years ago Paul B. and I collaborated and put out joint efforts
- a lot of work but great fun! IIRC Paul got fed up because his
indicators kept out coming mild with little variability.

Which was pretty much how most of the winters turned out, IIRC, which
suggests that they were pretty good forecasts.

Going back to pattern matching, one very broad-brush technique that has
I think performed slightly better than chance in the past is to forecast
that the winter will be broadly similar to the previous one. That's
because there does seem a tendency for winters to come in "clumps", with
cold winters clustering together and likewise for mild ones. The last
real "clump" of cold winters is a long time ago now, though, between
about 1978 and 1987.
--
John Hall "Do you have cornflakes in America?"
"Well, actually, they're American."
"So what brings you to Britain then if you have cornflakes already?"
Bill Bryson: "Notes from a Small Island"