The chart on page 171 shows the 5-year average forecast-error numbers from 2001 to 2005, the latest available when the Almanac was published.

The latest numbers are now available at http://www.nhc.noaa.gov/verification/pdfs/OFCL_5-yr_averages.pdf.

Note that the milage numbers are in NAUTICAL MILES, so you'll have to multiply by 1.151 to get comparable numbers to those in the table. For example, the previous 48-hour average error was 136 miles, now it's 129. The 120-hour error went from 349 miles to 334 STATUTE MILES.

There is no guarantee that the numbers will keep going down. The progress have been due to the significant improvements in the forecast models over the last few years. The next-generation model called the H-WRF is coming on line with new, advanced physics built in. It remains to be seen if this model will make a significant improvement in track errors or not.

The chart at http://www.nhc.noaa.gov/verification/pdfs/OFCL_5-yr_averages.pdf also shows the average intensity errors in KNOTS. Notice that the NHC's intensity forecasts are off by an average of more than 20 MPH from 72 hours on, which can be the difference between an category 1 and a category 3. This is why we always need to preparte for a category higher storm than the one that is forecast.