The reason I decided to place such numbers here is because though I like forecasting for the slope 
much better than the plateau, I realize it can in some situations be easier. Particularly overnight, 
the temperature at MPO can quickly decrease 5° or so when the winds go calm - which seldom happens 
here. There is such a thing as verifying forecasts (people's opinions of how accurate forecasts are 
mean nothing to me, and are largely based on how much they like or don't like the source), and I 
think this is a little more of a challenge than for my location.

Since I began making these forecasts again, here are some verification statistics:

         MIN    bias   MAX    bias    PN   bias    PD   bias     PRE  bias ratio     PC    bias   RA
Day 1   2.200  -.933  1.767   .900   .224 -.105   .317 -.277    .175  .022  .006    .867 -.0769  .658 
Day 2   2.200 -1.000  2.500   .833   .366 -.155   .357 -.254    .139 -.024 -.006    .783  .0370  .622
Day 3   2.300  -.700  2.500   .767   .406 -.086   .404 -.258    .168  .026  .007    .733  .0769  .601
Day 4   3.100 -1.300  2.933   .667   .421 -.159   .480 -.347    .162 -.051 -.014    .517 -.1111  .481
Day 5   3.133 -1.000  4.267  1.933   .412 -.305   .466 -.390    .117 -.088 -.021    .733 -.1538  .650
Day 6   2.933  -.267  4.433  2.433   .388 -.180   .436 -.299    .186 -.034 -.009    .683 -.0741  .547

The forecasts are generally made between about 2 & 9 PM EDT, and the average time is 7:03 PM. Day 
1 refers to the 24-hour period (mean solar time) of the next day, Day 2 the day after that, etc. 
MIN & MAX are minimum & maximum temperature errors (°F), PN & PD are nighttime (00-12 Z) & daytime 
(12-00 Z) precipitation probability errors, PRE is precipitation amount error (inches), and PC 
is the percentage of times (expressed as a decimal) precipitation was correctly forecast. To ensure 
a proper scoring system (so a forecast cannot be biased to take advantage of it), squared errors 
(root mean) are used for probabilities and linear errors for others. "bias" refers to how much 
lower or higher the forecasts average compared to the observed data. "ratio" is the average per 
forecast of precipitation relative to that observed. RA is calculated as follows: If forecast 
precipitation is greater than observed, it is forecast divided by observed - if the opposite is 
so, it is observed divided by forecast (averaged for all forecasts). Thus a perfect score is 1 and 
the worst possible is 0 (as is for PC). Actually, it seems like RA can be improved overforecasting 
precipitation amount. For example - if 1 inch occurs, forecasts of a half inch and 2 inches both 
produce a score of .5. It is more difficult to forecast high amounts as accurately as low amounts,
but probably not that much. Yet precipitation amounts do tend to be logarithmically distributed.

You can see that ATTM the bias makes a rather large portion of the error - as stated my temperatures 
here on the slope do not change during the course of a day as much as may be expected - particularly 
compared to MPO on the plateau.

That is now for 30 forecasts, which is often considered a threshold for "statistical significance". 
I am happy to see the temperature errors are so low into the medium range - especially for the 
minimums (not a great amount worse at Day 6 than Day 1 - difficult even here to guess exactly what 
overnight cooling will do). Rather than make hedgecasts at the medium range, I forecast what I 
think will actually occur - which can occasionally cause some larger errors but more often than 
not provides a much better forecast.

I do not plan to present comparisons with anything else because I forecast for a different location 
and/or at different times - but do verify forecasts from other sources and the MOS guidance, and you 
should be aware that though their temperature errors are comparable for days 1 & 2, 3-5° is typical 
for days 3-6. If I feel I have to toot my proverbial horn a bit, it is because very few people have 
done that for about 18 years and quite a few have done the opposite, and it doesn't appear that will 
ever change. Precipitation is comparable (and generally a bit better) but can use more improvement.