I'm not crazy. Here's some more detailed stats:
I checked all 65 Mets games this season, and found all the outfield assists for both teams. Came out to 18 OF assists by both teams in 65 games, or 0.14 Assists/team/game, which comes out to around 23/team/year, which is in the range we mentioned earlier in the thread.
I checked all 124 HBD games my team has played this year, and found only my team's outfield assists. Came out to 17 OF assists in 124 games, or 0.14 A/T/G, so the HBD rate and the RL rate seem to be well matched.
Here's how the OF assists break down:
Stretching a double: RL 5, HBD 0
Stretching a triple: RL 4, HBD 1
Out at 3B from 1B on a single: RL 2, HBD 5
Didn't get back on a fly ball out: RL 2, HBD 0
Other weird plays: RL 2, HBD 0
Out at home: RL 3, HBD 10
Now, we can't expect the programmers to program unlimited stuff in there. So yesterday when Pagan forgot to retouch 2nd base on his way back to 1st and the CF got an assist, you wouldn't expect to see that in HBD. The numbers aren't going to line up exactly, of course. But 9/18 RL assists came on players trying to stretch, where only 1/17 HBD assists came that way. And 10/17 times runners were gunned down at the plate in HBD, as opposed to 3/18 in RL.
The numbers are good for total assists. Where they come is not.
Maybe it's not worth fixing. It's still my pet peeve, and I'm not crazy for thinking so.