I know that this is obvious, but stats are meaningless as a point of comparison this early in the season. Sample size too small.
Still interesting to look at. Try changing the criteria to include all stats instead of key stats.
Especially since the only
two games for Florida include one against probably
the worst team in FBS football. Eastern Michigan
LOST to Old Dominion for goodness sake, and
almost lost to Morgan State (their only win). Consequently, Florida's stats are
dramatically inflated by their 65-0 win over the hapless Eagles.
Excluding that game from Florida's averages,
MoV drops to 6 (
now tilts to Bama)
Y/G drops to 532 (
now tilts to Bama)
Turnover Margin drops to 2.0 (still tilts to FL)
ToP drops to 33:39 (
now tied with Bama)
Opp Y/P drops to 5.77 (
now tilts to Bama)
PTD stays at 3 (still tilts to FL)
Opp PY/A increases to 8.2 (
now tilts to Bama)
Opp PC% increases to 53.3 (still tilts to FL)
Sacks drop to 2.0 (
now tilts to Bama)
RZS% increases to 100% (still tilts to FL)
Opp RZP/A stays at 3.0 (still tilts to FL)
Opp RZTDS% stays at 0 (still tilts to FL)
Opp KOR increases to 25.0 (
now tilts to Bama)
No punt returns (
no longer tilts to FL)
PuY/A increases to 48.67 (still tilts to FL)
After it's all re-tallied, Bama wins in 24 categories and Florida wins in 7 categories. Doesn't look quite as evenly split now, huh?
To be fair, Southern Miss is pretty bad too. But nowhere near as bad as Eastern Michigan. Also, Southern Miss only counts 1/3 toward Bama's averages, whereas Eastern Michigan is fully 1/2 of Florida's.
Bottom line is what you put in your post: too early in the season for stats to be meaningful - sample sizes are too small and therefore too easily corrupted by isolated events.