Short answer: maybe, but it’s the same as Home/Away: no global effect, and not statistically significant, but a strong enough effect for some players that it might warrant consideration.
Defenses score 0.491 more fantasy points in-division (p = 0.184, likely not statistically significant)
Quarterbacks score 0.809 fewer fantasy points in-division (p = 0.106, likely not statistically significant)
Kickers score 0.202 more fantasy points in-division (p = 0.422, very likely not statistically significant).
I also broke the results out by team and player (“But what about Derek Carr playing worse in division?!?”), and found none to be statistically significant. That’s not to say that some players don’t have fairly large in-division/out-of-division splits, it just means that there aren’t nearly enough data points to make that statement confidently. Sure enough, Derek Carr scores just over 5 points fewer during in-division games, Dak Prescott just under 5 points less, and Philip Rivers about 4.5 points less. On the other side we have Trevor Siemian scoring almost 4 points MORE in in-division games, with Brian Hoyer, Aaron Rodgers, and Matt Stafford coming in just under 3 points more in in-division games. The problem is that the p-values on these are all larger than 0.4, which means that it’s very likely that they’re within the random noise you’d expect for distributions like this.
The closest to statistical significance is HOU scoring 5 point more in-division and MIA scoring 5 points less in-division, and while these have p-values less than 0.1, when you break data out into this many subdivisions you expect to see some of those. In fact, it’s actually a little weird that I didn’t see many QBs with smaller p-values. Don’t count on those 5 point swings from MIA or HOU, they’re likely data noise.
If you want to keep believing that Derek Carr is worse in-division I won’t stop you. But just know that I also can’t back you up statistically.