Injury probability increases with lower conditioning, higher fatigue, current injuries, age, and of course more playing time gives more opportunities to get injured. During pre-season I have had good success with having my starters pulled after the 1st quarter in game 1, the 2nd quarter in game 2, the 3rd quarter in game 3, and then the 1st quarter in game 4, which is close to how they do it in the NFL; that seems to get a good level of conditioning without risking too much injury. You'll also want to have a full roster of 60 active players if your cap will allow it so you have enough depth to rotate in as players fatigue. During the season, you will probably have a constant handful of probable players all season long, with some more serious injuries scattered in. If you have a lot of injuries that are more than probable, you might look at changing the fatigue threshold in your misc game planning, but it's just as likely you got some bad dice rolls. But as injuries are part of the game, you should expect to see a good amount during the season.
I normally play my probable players as if there is no injury at all, and evaluate the questionable ones as to whether I feel like I have enough depth to rest them. I'll only play doubtful players if there is really no other option. Players listed as "out" cannot be played and the AI will bench them if you don't.
