Low Tech cannons mortars and shells

[Anthony]

Originally Posted by Polydamas

Instead, I would start with empirical data or back-of-the-envelope calculations about hit changes at various distances down the cone, and pick numbers such that chances to hit are in the right ballpark.

The basic scaling should be that, say, 1 projectile per square yard is a fixed hit chance. Using the range/speed chart, you can get that by looking up (SM for projectile count) - (2x SM of area) (at 0, that is 1 projectile per 2 square yards), which gives us roughly:
+6: expected hits 3, hit chance 95% (15-)
+5: expected hits 2, hit chance 87% (14-)
+4: expected hits 1.5, hit chance 78% (12-)
+3: expected hits 1, hit chance 64% (11-)
+2: expected hits 0.7, hit chance 50% (10-)
+1: expected hits 0.45, hit chance 37% (9-)
+0: expected hits 0.3, hit chance 26% (8-)
-1: expected hits 0.2, hit chance 18% (7-)
-2: expected hits 0.15, hit chance 14% (7-)
-3: expected hits 0.1, hit chance 10% (6-)
-4: expected hits 0.07, hit chance 7% (6-)
-5: expected hits 0.045, hit chance 4% (5-)
-6: expected hits 0.03, hit chance 3% (5-)
-7: expected hits 0.02, hit chance 2% (4-)
-11: expected hits 0.005, hit chance 0.5% (3-)

8+(SM for shots fired) - (2x SM of area) does a decent job over most of the interesting range