Not even close to radians.
Planets unrotated are on their sides... the rotate X 50 flips them 180 degrees and the rotate z 25 makes them right side up (not sure why X isn't ignored and X set to -25).
Starting Rotate X := 50
Starting Rotate Y := 0
Starting Rotate Z := 25
Since the default rotation vector is 25 (90 degrees): I decided you could rotate X away from the vertical... and then add an X rotation vector that trigonometrically pairs with the Z rotation.
This solution works near perfectly... ~10 degree tilt
Starting Rotate X := 53
Starting Rotate Z := 25
Continual Rotation Vector X := 3
Continual Rotation Vector Z := 15.1
You can vary the direction of the tilt, messing with the signs and getting Y involved... However, I see no good way of distributing different tilts throughout the planet types.
The orbital plane for satellites is harder... I can't consistently predict whether the initial position is the low or high end of the incline.
But I got this one working for the above axial tilt:
Effect 2 Offset Position X := -19.65
Effect 2 Offset Position Z := 3.75
Effect 2 Rotate X := 0.000375
Effect 2 Rotate Z := 0.001965
If you like your trig, you can quickly figure out why 3 is paired with 15.1 and 19.65 is paired with 3.75.
