Player Import Ratings Vary on Same Players

[Kapono24]

In the process of creating a historic league with hand-picked players I noticed that importing the exact same player (Dwight Gooden) from the exact same year (1986, in this example) leads to every version imported having slightly varied ratings despite everything being equal. Most notably, their hold ratings vary drastically from as low as 4 to as high as 93, their velocity ranges up to 6 MPH different, and occasionally a pitcher rating will be different, all of which result in different overall ratings about half the time.

I understand cloning a player will result in an exact copy, but my concern is when importing any player, I'm likely not receiving the optimal version of that player. Is there an option I'm missing that eliminates all variance and consistently creates the same, most accurate version of a player each time?

Here's a picture of the import settings I used for each import and their pitching rating. I used Neutralized Stats option with the same problem as a result, though not listed here: https://i.imgur.com/qtrb8Hw.png

[Lukas Berger]

Hold rating is random or semi-random since we don't really have the stats needed to calculate this afaik, so that will vary each time.

Velocity is tied to stuff/k ratings but may fluctuate a bit.

Unfortunately, given how the historical import process works, there are quite a few moving parts and there's probably not really anything you can do to make sure that absolutely everything is the same each time, since so much of how players import is somewhat variable.