All of the above 100%...
Additionally, this also might have to do with the handedness of the other team's starting pitcher, and the availability of players on the Dodgers. For instance, I'd expect to see Greg Brock starting at 1B and Kenny Landreaux in CF - but not if the SP is a lefty... And so then if other guys are not available due to injury or rest, that could be part of it... And a close look at what positions those guys actually played that season doesn't make it strange that, for example, Russell would play 2B (even though we think of him primarily as a SS; Anderson played a lot at 3B that season, and Bailor appeared at SS for a few games). Also, we might expect that against a lefty SP (if that's the case), that Candy Maldonado would appear in the lineup (in CF). But a closer look at his real-life usage vs stats makes one wonder why he got as much playing time as he did. IOW, Tommy Lasorda must've thought more of him that any other manager would, including OOTP's AI manager
I'll admit Sax at 1B seems odd. But that's probably due to a combination of factors (like some of those suggested above), and I'm guessing that unless those factors are in play long-term (like several injuries) for your 1985 OOTP Dodgers, that's a one-off/rare occurrence.
And worth noting if you are using real txns (or it's LA's roster essentially as it is if you created the game starting with 1985) is that in April Bill Madlock and Enos Cabell - two guys who spent significant time at 1B and 3B - were not yet on the Dodgers. And while OOTP rightly sees fit to put Pete Guerrero in LF instead of 3B (due to his abysmal fielding at third), Tommy was still putting Pedro at 3B in the early part of the season (giving the newly acquired Al Oliver a chance in LF). So OOTP is merely doing the best it can with lineup selection given LA's roster at the time and - likely - some limitations due to injuries or fatigue.