Originally Posted by deltav

But do you guys really parse out and worry about 4-day totals like this often? You'll drive yourselves crazy constantly applying significance to such short time periods - to get any accurate overall picture in my experience you really need to look at 2-3 month trends. In shorter timeframes there's just too much random variance to read too much into it. Especially four days, statistically that's such a small sample size as to be near useless.