This is a potentially significant finding. The differences in y/a and dvoa are so large that it *seemingly* implies that the Hawks should substantially decrease their use of a fullback, but before jumping to such a conclusion, deeper analysis should be done.
For example, I wonder how much is explained by 3rd or 4th and very short; or when they running out the clock with a lead late in the game and everyone is pretty sure they are going to run it.
Seattle had 4.3 y/a with one RB, whch decreased to 3.7 y/a with a fullback.
Taking an example with hypothetical numbers, suppose with a fullback that 20% of the plays were 3rd or 4th and very short, and that regular plays with a FB averaged 4.4 y/a and short yardage averaged 1 y/a. Then .8(4.4) + (.2)(1.0) = 3.7 y/a with a fullback...again, those are hypothtical numbers to illustrate a potential effect.
It would be interesting to filter out these situations and then see the effect of the fullback, doing this analysis on only first and 10 situations, for example.
Also, the Seahawks running game is best at the start of a game and in the first quarter when they often run with a 2-back set, which implies using a fullback is beneficial.
There are also additional benefits from a fullback. If the defense is really biased to defend the run, then it might make your play action more effective, or the fullback might be a more effective blocker on passing plays.
...can't wait for this year's almanac to be released.