In the first preseason game in Big Data, there were 110 plays of 4236 offensive plays that had a fumble. That's a fumble on ~2.5% of all offensive plays or 3 fumbles per team.

That's more than twice as high as the number of fumbles per team and per play as the NFL this year.

It's also pre-season? so a lot of back ups playing and a dispersal draft with a lot of players who just aren't good enough.

Like all leagues, it will take 3-5 seasons to mature and settle down. The first few years will see lots of mis-matches and blowouts. Likely exacerbated in pre-season by poor players with low levels of current attributes


Edit to add:

A dispersal draft season shouldn't (my opinion) look much like an NFL season because it doesn't have the depth of talent across 32 teams.

The best dispersal teams would likely be 0-16 against real life NFL rosters

Where are you getting 110, I checked player Stas and game logs both and came up with 63 both times???

I counted the rows in the SQL database. So there is a chance that the data is duplicated.

Update: Good call fred. Looks like the rows got duplicated in the raw_plays table because I don't unique index that table.

Must be because the total fumbles is 63 with 40 lost.

The fortunate thing is that all of the rows in the offense and defense tables are unique.