A smooth bore fired in a vacuum would be propelled straight and true, friction with the atmosphere on an irregular surface would cause its path to bend and yaw,as well as turn its surface exposing new irregularities to face the frictions maximum impact on that new surface, causing the path to bend and yaw in a new direction exposing a new surface irregularity to the friction until the ball hits the mitt or the ball hopefully hits the target. It can be minimized, but will always be unpredictable. Thats why catchers drop so many pitched knuckle balls and batters, knowing they are coming, still have trouble hitting them, they are not predictable. The pitcher can throw towards the mitt, but even he doesn't know where it will impact.
With my smooth bore, out to eighty yards I was confident of only a 3" radius of my aim point, usually better than that but I always figure the worst. What I found had the biggest impact on consistence was the ball itself, patching and lube made a difference, but not as big as the ball itself. I used a .595 lee mould, because it left virtually no sprue, weighed each of them, that helped, what made a bigger difference in making a smaller predictable impact area was dimpling the balls. I made a tumbler from an old stainless steel bowl, drilled a hole in the bottom center, ran a bolt through it, covered the exposed bolt with duct tape and mounted it on a varial speed motor, which I configured in such a way, that the bowl was at around a 45 degree angle. After inspecting and weighing, I threw them in the tumbler and walked away. When done they looked like miniature golf balls. It worked for me.