Even with 10 units dealing 1-3 damage, the chance of maximum damage will be negligible. But it's still greater than with 1000 units. Undoubtedly.

Yes, it's clear that's how it works. To simulate a normal distribution, the sum of 6 independent normal distributions is considered sufficient. I just said it as an example.

But it seems like there's just a lack of careful planning instead of thoughtful consideration of the battle plan. Plus, I also noticed that (probably due to rounding) the average is slightly higher than the arithmetic mean. I understand that no one will probably want to change such a fundamental thing, but a checkbox in HW_rules would be great.