Weathering the Data

Okay, let's take a look at the "accu" part of AccuWeather.

High temperature for Marietta 30066 - 89° (measured with three separate thermometers--two wireless and one standard thermometer, all located outdoors but out of direct sunlight). Forecast highs:

82 - 94 - 94 - 95 - 93 - 93 - 93 - 95

So, by predicting a 13-degree range between the highest high prediction andt he lowest low, they managed to more or less miss it totally, since it was almost exactly midway between the two extremes. At no point did they ever forecast 89.° I'd say that they failed the "Accu" part today.

Now, as for their forecasts:

7/28 - 95° (3° higher than yesterday)
7/29 - 94° (2° warmer than yesterday)
7/30 - 96° (1° warmer yesterday)
7/31 - 96° (1° warmer than yesterday's forecast)
8/1 - 96° (2° higher than yesterday's forecast)
8/2 - 95° (2° warmer than yesterday's forecast!)
8/3 - 95° (5° higher than yesterday's forecast)
8/4 - 96° (16° warmer than yesterday's forecast)
8/5 - 91° (9° warmer than yesterday's forecast)
8/6 - 97° (14° warmer than yesterday's forecast)
8/7 - 93° (6° warmer than yesterday's forecast)
8/8 - 96° (8° warmer than yesterday's forecast)
8/9 - 97° (5° warmer than yesterday's forecast!)
8/10 - 88° (and once again, the coolest forecast temperature is 15 days out)

Fourteen and sixteen degree jumps in forecast temperatures? Over a 24 hour period? If that doesn't qualify as random guessing, nothing much does...