Each year as predictions for the upcoming winter are made, we inevitably hear the latest prognostication from the Old Farmer's Almanac (Thomas, Robert, Yankee Publishing, Dublin NH,  1999), touting it's "secret formula" and unerring accuracy, and we hear anecdotally how good it did last year.  But one thing I have learned over the years is that most people don't really remember last week's forecast and certainly have little firm recollection of a forecast made last year.  And who ever goes back and checks on an old forecast anyway?

Well, as I have been meaning to do for several years now, I went back and looked at last year's forecast from the Farmer's Almanac to see how it compared with what actually happened.  Even trying to be objective and giving the benefit of the doubt to cases that were close, I found last year's forecast from the Old Farmer's 2000 Almanac for San Francisco to be laughable at best and abysmal at worst. Here's what I found:

The weather forecast section of the Almanac is divided into 16 regions; with Region 16 encompassing the southern two-thirds of California.  There is a brief summary for each region and then a monthly forecast for each of the 12 months beginning in November.  Each month has a summary of the forecast average temperature and rainfall for the month and how much above or below normal the forecast is.  Then there are general summaries for 3 to 12 day periods through the month with two or three word descriptions.  The actual Region 16 1999-2000 forecast looked like: 

To evaluate the forecast I looked at how the monthly forecast compared to the actual temperature and rainfall that occurred in San Francisco. Because the Almanac is entirely vague about the actual point(s) it is forecasting for, it is impossible to absolutely quantify their forecast vs. San Francisco.  Consequently, I have only scored those that have the wrong sign or show a significant error as being a "miss".  Furthermore, I did not evaluate the rainfall for the months of May through September as the amounts are really too small to be meaningful.

Monthly Summary:  The Old Farmer's Almanac was wrong on their monthly temperature forecast 8 out of the 12 months (67%) and wrong on their rainfall forecast 5 of the 8 months evaluated (63%).  If I needed to know the weather next January 23rd I think I'd just use climatology!

Daily Summary:  During the 12 month period from November 1999 through October 2000 this occurred 144 times.  Of those, the Farmer's Almanac was got a "hit" for only 21 days, or a percentage of 17% of the time!!

Below is the table of the Farmer's Almanac's forecast and the actual San Francisco rainfall.  Also note that there is always a change in the weather on the first of the month.  Hmmm!!