KhanCipher, on 21 April 2013 - 04:47 PM, said:
i love it when nobody ever thinks that "That Tank Game" did capping so much better. Because if someone is shooting at you, that area you're capping isn't all that secure.
I read this, and it gave me a very interesting thought. He's exactly right. Just because a mech is standing in the little electric cap area does NOT mean that the area has been secured. I'm thinking that changing the cap mechanic like one or more of the following might make some sense...:
- If a capper in the cap area has been hit with weapons fire, from any damage-dealing range, the cap pauses (ie. stops) for X seconds (5? 10?)
- If a capper is in the cap area, but there is an enemy within X metres of the base, them the cap pauses. Sure, the capper is not being shot at, but the enemy just around the corner means that the capping mech does not have control of the area. (100m? 270m? 500m?)
- If the capper's team is down in mechs by a factor of X:Y, the capping team is not in control and the cap pauses. (2:1? 3:2?)
- Some subset or combination of the above
Something like this might better help simulate the concept of area control, and perhaps mitigate the occurrences of ninja capping, at least somewhat. You could still manage to cap a base, but this would prevent those last-minute caps where the last mech or two of a team is valiantly trying to hold onto the enemy base while being surrounded by overwhelming odds. In real life, the overwhelming odds would probably win since there is no such thing as a 'cap timer' in war.
Just my $0.02.